=== 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: 164be73c51fd9cb7dc5314b3417fa4c35b0a34c7 ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, # Cray-FFTW 3.3.8.10, COSMA 2.6.6, ELPA 2023.05.001, # HDF5 1.14.2, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.2.2, LIBVORI 220621, LIBXSMM 1.17, # PLUMED 2.9.0, SIRIUS 7.4.3, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Last update: 04.10.2023 # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2023.05.001 USE_HDF5 := 1.14.2 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.9.0 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a endif ifneq ($(USE_HDF5),) USE_HDF5 := $(strip $(USE_HDF5)) HDF5_INC := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/include HDF5_LIB := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/lib CFLAGS += -I$(HDF5_INC) DFLAGS += -D__HDF5 LIBS += $(HDF5_LIB)/libhdf5_fortran.a LIBS += $(HDF5_LIB)/libhdf5_hl.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/164be73c51fd9cb7dc5314b3417fa4c35b0a34c7_performance_tests/01 job id: 49946995 --- 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/164be73c51fd9cb7dc5314b3417fa4c35b0a34c7_performance_tests/02 job id: 49946997 --- 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/164be73c51fd9cb7dc5314b3417fa4c35b0a34c7_performance_tests/03 job id: 49946998 --- 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/164be73c51fd9cb7dc5314b3417fa4c35b0a34c7_performance_tests/04 job id: 49946999 --- 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/164be73c51fd9cb7dc5314b3417fa4c35b0a34c7_performance_tests/05 job id: 49947000 --- 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/164be73c51fd9cb7dc5314b3417fa4c35b0a34c7_performance_tests/06 job id: 49947001 --- 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/164be73c51fd9cb7dc5314b3417fa4c35b0a34c7_performance_tests/07 job id: 49947002 --- 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/164be73c51fd9cb7dc5314b3417fa4c35b0a34c7_performance_tests/08 job id: 49947003 --- 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/164be73c51fd9cb7dc5314b3417fa4c35b0a34c7_performance_tests/09 job id: 49947004 --- 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/164be73c51fd9cb7dc5314b3417fa4c35b0a34c7_performance_tests/10 job id: 49947006 --- 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/164be73c51fd9cb7dc5314b3417fa4c35b0a34c7_performance_tests/11 job id: 49947008 --- 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/164be73c51fd9cb7dc5314b3417fa4c35b0a34c7_performance_tests/12 job id: 49947009 --- 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/164be73c51fd9cb7dc5314b3417fa4c35b0a34c7_performance_tests/13 job id: 49947010 --- 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/164be73c51fd9cb7dc5314b3417fa4c35b0a34c7_performance_tests/14 job id: 49947011 --- 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/164be73c51fd9cb7dc5314b3417fa4c35b0a34c7_performance_tests/15 job id: 49947012 --- 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/164be73c51fd9cb7dc5314b3417fa4c35b0a34c7_performance_tests/16 job id: 49947013 --- 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/164be73c51fd9cb7dc5314b3417fa4c35b0a34c7_performance_tests/17 job id: 49947014 --- 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/164be73c51fd9cb7dc5314b3417fa4c35b0a34c7_performance_tests/18 job id: 49947015 --- 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/164be73c51fd9cb7dc5314b3417fa4c35b0a34c7_performance_tests/19 job id: 49947016 --- 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/164be73c51fd9cb7dc5314b3417fa4c35b0a34c7_performance_tests/20 job id: 49947018 --- 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/164be73c51fd9cb7dc5314b3417fa4c35b0a34c7_performance_tests/21 job id: 49947020 --- 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/164be73c51fd9cb7dc5314b3417fa4c35b0a34c7_performance_tests/22 job id: 49947021 --- 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/164be73c51fd9cb7dc5314b3417fa4c35b0a34c7_performance_tests/23 job id: 49947022 --- 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/164be73c51fd9cb7dc5314b3417fa4c35b0a34c7_performance_tests/24 job id: 49947023 --- 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/164be73c51fd9cb7dc5314b3417fa4c35b0a34c7_performance_tests/25 job id: 49947024 --- 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/164be73c51fd9cb7dc5314b3417fa4c35b0a34c7_performance_tests/26 job id: 49947027 --- Point --- name: 510 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/1r/12t) --- Point --- name: 511 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: 512 H2O (4 NVE MD steps on 64 nodes) input file: benchmarks/QS/00512_H2O/H2O-512_md.inp required files: [] output file: result.log # nodes = 64 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/164be73c51fd9cb7dc5314b3417fa4c35b0a34c7_performance_tests/27 job id: 49947028 --- Point --- name: 601 plot: h2o_512_md regex: CP2K label: (64n/12r/1t) --- Point --- name: 602 plot: h2o_512_md_mem regex: Estimated peak process memory label: (64n/12r/1t) ~~~~~~~ END TEST ~~~~~~~ === END TESTS (description) === ===== PLOTS (description) ===== ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2_mem", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md_mem", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md_mem", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md_mem", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls_mem", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_512_md", title="512 H2O (4 NVE MD steps on 64 nodes)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_512_md_mem", title="512 H2O (4 NVE MD steps on 64 nodes)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" === END PLOTS (description) === ============ RESULTS ============ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/164be73c51fd9cb7dc5314b3417fa4c35b0a34c7_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 424 8. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.023 0.037 135.726 135.726 farming_run 1 2.0 135.191 135.193 135.689 135.693 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.489900E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 230 1134128. MP_Allreduce 571 1938539. MP_Sync 25 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 10 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.099 115.706 115.766 qs_energies 1 2.0 0.000 0.000 115.492 115.494 mp2_main 1 3.0 0.000 0.000 113.297 113.299 mp2_gpw_main 1 4.0 0.028 0.034 112.232 112.234 mp2_ri_gpw_compute_in 1 5.0 0.171 0.173 92.699 93.012 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 54.642 54.953 mp2_eri_3c_integrate_gpw 272 7.0 0.151 0.168 41.027 46.027 get_2c_integrals 1 6.0 0.008 0.009 36.960 37.885 integrate_v_rspace 273 8.0 0.441 0.455 24.581 29.433 pw_transfer 6555 10.6 0.377 0.405 26.847 27.116 grid_integrate_task_list 273 9.0 20.487 25.805 20.487 25.805 fft_wrap_pw1pw2 5465 11.4 0.046 0.049 25.462 25.803 fft_wrap_pw1pw2_100 2178 12.4 1.121 1.277 23.007 23.352 rpa_ri_compute_en 1 5.0 0.019 0.021 19.425 19.571 compute_2c_integrals 1 7.0 0.002 0.002 19.317 19.318 compute_2c_integrals_loop_lm 1 8.0 0.001 0.003 18.643 18.916 mp2_eri_2c_integrate_gpw 1 9.0 2.379 2.426 18.642 18.915 cp_fm_cholesky_decompose 12 8.2 17.588 18.513 17.588 18.513 cholesky_decomp 1 7.0 0.000 0.000 16.486 17.409 fft3d_s 5443 13.4 16.150 16.408 16.173 16.430 ao_to_mo_and_store_B_mult_1 272 7.0 10.774 15.313 10.774 15.313 calculate_wavefunction 272 8.0 5.430 5.570 12.351 13.077 rpa_num_int 1 6.0 0.000 0.001 10.592 10.601 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.579 10.587 calc_mat_Q 8 8.0 0.000 0.000 9.447 9.532 contract_S_to_Q 8 9.0 0.000 0.000 8.869 8.957 calc_potential_gpw 544 9.5 0.005 0.005 8.253 8.685 parallel_gemm_fm 14 9.1 0.000 0.000 8.468 8.563 parallel_gemm_fm_cosma 14 10.1 8.467 8.563 8.467 8.563 create_integ_mat 1 6.0 0.015 0.028 8.429 8.438 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.095 8.335 potential_pw2rs 545 10.0 0.108 0.109 7.489 8.121 collocate_single_gaussian 272 10.0 0.039 0.042 7.362 7.675 array2fm 1 7.0 0.000 0.000 6.790 7.233 pw_scatter_s 2720 13.7 4.349 4.489 4.349 4.489 pw_gather_s 2722 13.2 3.432 3.793 3.432 3.793 array2fm_buffer_send 1 8.0 3.053 3.183 3.053 3.183 pw_poisson_solve 545 10.5 1.120 1.204 2.211 2.473 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.231756, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2808.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/164be73c51fd9cb7dc5314b3417fa4c35b0a34c7_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 205321. MP_Allreduce 424 9. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.027 0.040 429.147 429.149 farming_run 1 2.0 428.187 428.196 429.089 429.092 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827141120 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788822 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.229255E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 757 478553. MP_Allreduce 2021 21391. MP_Sync 37 MP_Alltoall 77 22051075. MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.037 218.087 218.089 qs_energies 1 2.0 0.000 0.000 217.827 217.839 scf_env_do_scf 1 3.0 0.000 0.000 115.379 115.379 qs_ks_update_qs_env 5 5.0 0.000 0.000 114.151 114.157 rebuild_ks_matrix 4 6.0 0.000 0.000 114.150 114.156 qs_ks_build_kohn_sham_matrix 4 7.0 0.058 0.064 114.150 114.156 hfx_ks_matrix 4 8.0 0.001 0.001 113.750 113.754 integrate_four_center 4 9.0 0.153 0.471 113.749 113.754 integrate_four_center_main 4 10.0 0.140 0.737 101.727 105.258 integrate_four_center_bin 264 11.0 101.587 105.134 101.587 105.134 mp2_main 1 3.0 0.000 0.000 102.113 102.124 mp2_gpw_main 1 4.0 0.048 0.067 100.799 100.811 init_scf_loop 1 4.0 0.000 0.000 97.010 97.010 mp2_ri_gpw_compute_in 1 5.0 0.064 0.064 74.055 75.086 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 53.810 54.838 mp2_eri_3c_integrate_gpw 91 7.0 0.143 0.159 41.614 46.478 integrate_v_rspace 95 8.0 0.398 0.566 28.111 32.869 pw_transfer 2240 10.6 0.145 0.163 29.571 29.992 fft_wrap_pw1pw2 1868 11.4 0.018 0.021 28.573 29.026 mp2_ri_gpw_compute_en 1 5.0 0.057 0.068 26.587 28.627 grid_integrate_task_list 95 9.0 23.489 28.423 23.489 28.423 ao_to_mo_and_store_B_mult_1 91 7.0 10.524 27.741 10.524 27.741 fft_wrap_pw1pw2_100 730 12.4 1.283 1.448 26.295 26.688 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.849 1.897 24.903 24.913 get_2c_integrals 1 6.0 0.000 0.000 20.107 20.177 compute_2c_integrals 1 7.0 0.002 0.003 19.064 19.080 fft3d_s 1823 13.4 18.494 18.948 18.508 18.963 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 18.682 18.894 mp2_eri_2c_integrate_gpw 1 9.0 1.740 1.937 18.681 18.893 scf_env_do_scf_inner_loop 4 4.0 0.000 0.001 18.363 18.363 calculate_wavefunction 91 8.0 2.028 2.056 9.636 9.882 potential_pw2rs 186 10.0 0.033 0.035 8.462 9.053 mp2_ri_gpw_compute_en_expansio 172 7.0 0.556 0.582 8.688 8.966 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.181 8.473 local_gemm 172 8.0 8.133 8.397 8.133 8.397 mp2_ri_gpw_compute_en_comm 22 7.0 0.503 0.519 7.971 8.336 calc_potential_gpw 182 9.5 0.002 0.002 7.930 8.149 collocate_single_gaussian 91 10.0 0.017 0.024 7.776 7.931 integrate_four_center_load 4 10.0 0.000 0.000 6.901 6.905 hfx_load_balance 1 11.0 0.000 0.000 6.901 6.905 mp_sync 37 10.5 3.857 6.480 3.857 6.480 mp2_ri_gpw_compute_en_ener 172 7.0 6.349 6.427 6.349 6.427 mp_sendrecv_dm3 2068 8.0 5.995 6.347 5.995 6.347 pw_gather_s 912 13.2 4.496 5.058 4.496 5.058 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=100.804725, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1490.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/164be73c51fd9cb7dc5314b3417fa4c35b0a34c7_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 452.460544E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62379. MP_Allreduce 10329 270. MP_Sync 530 MP_Alltoall 2083 MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.035 66.184 66.185 qs_mol_dyn_low 1 2.0 0.003 0.003 65.875 65.888 qs_forces 11 3.9 0.002 0.003 65.812 65.813 qs_energies 11 4.9 0.001 0.002 64.069 64.087 scf_env_do_scf 11 5.9 0.000 0.001 56.810 56.811 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 53.555 53.556 qs_scf_new_mos 108 7.5 0.000 0.000 40.186 40.467 qs_scf_loop_do_ot 108 8.5 0.000 0.001 40.186 40.467 dbcsr_multiply_generic 2286 12.5 0.095 0.098 38.338 38.903 ot_scf_mini 108 9.5 0.002 0.002 38.321 38.493 velocity_verlet 10 3.0 0.001 0.002 32.757 32.757 multiply_cannon 2286 13.5 0.190 0.203 28.276 30.146 multiply_cannon_loop 2286 14.5 1.840 1.953 26.984 28.928 ot_mini 108 10.5 0.001 0.001 22.827 23.084 qs_ot_get_derivative 108 11.5 0.001 0.001 19.458 19.655 mp_waitall_1 245248 16.5 10.979 17.169 10.979 17.169 multiply_cannon_metrocomm3 54864 15.5 0.073 0.080 6.808 14.459 multiply_cannon_multrec 54864 15.5 3.517 5.490 7.795 11.225 qs_ot_get_p 119 10.4 0.001 0.001 10.251 10.521 rebuild_ks_matrix 119 8.3 0.000 0.000 9.729 9.885 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 9.728 9.884 qs_ks_update_qs_env 119 7.6 0.001 0.001 8.605 8.745 mp_sum_l 7287 12.8 6.290 8.353 6.290 8.353 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 6.686 7.148 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 6.474 6.589 qs_ot_p2m_diag 50 11.0 0.004 0.006 6.406 6.477 qs_rho_update_rho_low 119 7.7 0.000 0.001 6.224 6.410 calculate_rho_elec 119 8.7 0.011 0.016 6.224 6.409 multiply_cannon_sync_h2d 54864 15.5 4.905 6.307 4.905 6.307 dbcsr_mm_accdrv_process 76910 16.1 1.862 2.901 4.193 5.953 sum_up_and_integrate 119 10.3 0.001 0.002 5.934 5.941 integrate_v_rspace 119 11.3 0.002 0.002 5.923 5.931 init_scf_run 11 5.9 0.000 0.001 5.758 5.758 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.758 5.758 cp_dbcsr_syevd 50 12.0 0.002 0.003 5.477 5.478 cp_fm_diag_elpa 50 13.0 0.000 0.000 5.006 5.007 cp_fm_redistribute_end 50 14.0 2.535 4.937 2.556 4.947 cp_fm_diag_elpa_base 50 14.0 2.373 4.773 2.384 4.793 multiply_cannon_metrocomm1 54864 15.5 0.056 0.061 2.417 4.108 density_rs2pw 119 9.7 0.004 0.004 3.908 4.014 make_m2s 4572 13.5 0.054 0.056 3.478 3.581 make_images 4572 14.5 0.134 0.139 3.395 3.497 pw_transfer 1439 11.6 0.052 0.057 3.375 3.479 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 3.299 3.404 calculate_dm_sparse 119 9.5 0.000 0.000 3.246 3.396 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.123 3.360 apply_single 119 13.6 0.000 0.000 3.123 3.359 init_scf_loop 11 6.9 0.000 0.000 3.207 3.208 potential_pw2rs 119 12.3 0.004 0.004 3.141 3.184 fft3d_ps 1201 14.6 0.376 0.484 3.020 3.123 ot_diis_step 108 11.5 0.006 0.006 3.008 3.009 jit_kernel_multiply 13 15.8 2.267 2.958 2.267 2.958 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.831 2.895 calculate_first_density_matrix 1 7.0 0.000 0.000 2.829 2.838 acc_transpose_blocks 54864 15.5 0.229 0.257 2.302 2.836 wfi_extrapolate 11 7.9 0.001 0.001 2.796 2.796 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.771 2.773 fft_wrap_pw1pw2_140 487 13.2 0.135 0.149 2.472 2.579 mp_alltoall_d11v 2130 13.8 2.327 2.504 2.327 2.504 transfer_rs2pw 487 10.6 0.005 0.006 2.240 2.331 mp_sum_d 4135 12.0 1.665 2.308 1.665 2.308 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.204 2.269 make_images_sizes 4572 15.5 0.005 0.005 1.853 2.246 mp_alltoall_i44 4572 16.5 1.849 2.242 1.849 2.242 transfer_pw2rs 487 13.2 0.005 0.006 2.078 2.123 grid_integrate_task_list 119 12.3 2.008 2.111 2.008 2.111 mp_alltoall_z22v 1201 16.6 1.864 1.941 1.864 1.941 prepare_preconditioner 11 7.9 0.000 0.000 1.855 1.887 make_preconditioner 11 8.9 0.000 0.000 1.855 1.887 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.736 1.779 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.591 1.628 mp_waitany 12084 13.8 1.451 1.599 1.451 1.599 acc_transpose_blocks_sync 164592 16.5 1.215 1.439 1.215 1.439 grid_collocate_task_list 119 9.7 1.351 1.408 1.351 1.408 make_images_data 4572 15.5 0.041 0.045 0.925 1.354 dbcsr_dot_sd 1205 11.9 0.049 0.060 0.942 1.348 arnoldi_extremal 119 11.4 0.001 0.001 1.182 1.324 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=66.185000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=430.909091, yerr=1.564059 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/164be73c51fd9cb7dc5314b3417fa4c35b0a34c7_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 487.333888E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2194560 MPI messages size (bytes): total size 310.646604E+09 min size 0.000000E+00 max size 1.145520E+06 average size 141.553031E+03 MPI breakdown and total messages size (bytes): size <= 128 724648 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 281952 4619501568 32768 < size <= 131072 494448 39143342080 131072 < size <= 4194304 440000 264807943488 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62658. MP_Allreduce 10306 303. MP_Sync 54 MP_Alltoall 2060 1068130. MP_SendRecv 16779 37093. MP_ISendRecv 16779 37093. MP_Wait 23539 MP_comm_split 50 MP_ISend 5720 128509. MP_IRecv 5720 128509. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.058 0.100 41.403 41.405 qs_mol_dyn_low 1 2.0 0.003 0.004 40.601 40.641 qs_forces 11 3.9 0.003 0.008 40.532 40.533 qs_energies 11 4.9 0.007 0.024 38.807 38.812 scf_env_do_scf 11 5.9 0.000 0.001 33.551 33.552 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 30.984 30.985 dbcsr_multiply_generic 2286 12.5 0.102 0.104 22.559 22.893 qs_scf_new_mos 108 7.5 0.001 0.001 21.721 21.952 qs_scf_loop_do_ot 108 8.5 0.001 0.001 21.720 21.952 ot_scf_mini 108 9.5 0.002 0.003 20.755 20.923 velocity_verlet 10 3.0 0.004 0.006 19.109 19.110 multiply_cannon 2286 13.5 0.210 0.217 16.951 18.245 multiply_cannon_loop 2286 14.5 1.194 1.255 15.691 17.202 ot_mini 108 10.5 0.001 0.001 12.557 12.791 mp_waitall_1 200699 16.5 5.859 10.942 5.859 10.942 qs_ot_get_derivative 108 11.5 0.001 0.001 10.155 10.323 multiply_cannon_metrocomm3 27432 15.5 0.072 0.075 3.986 9.385 multiply_cannon_multrec 27432 15.5 1.832 4.088 6.134 8.995 rebuild_ks_matrix 119 8.3 0.000 0.000 7.188 7.328 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 7.187 7.328 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.350 6.476 dbcsr_mm_accdrv_process 47894 16.0 3.495 5.763 4.223 6.285 qs_ot_get_p 119 10.4 0.001 0.001 5.183 5.412 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.735 4.644 mp_sum_l 7287 12.8 2.221 4.222 2.221 4.222 sum_up_and_integrate 119 10.3 0.001 0.003 4.061 4.067 integrate_v_rspace 119 11.3 0.002 0.003 4.046 4.052 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.933 3.963 calculate_rho_elec 119 8.7 0.021 0.024 3.932 3.962 init_scf_run 11 5.9 0.000 0.001 3.900 3.900 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.900 3.900 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.089 3.766 apply_single 119 13.6 0.000 0.000 3.088 3.766 qs_ot_p2m_diag 50 11.0 0.009 0.013 3.350 3.367 make_m2s 4572 13.5 0.051 0.053 2.961 3.289 make_images 4572 14.5 0.206 0.245 2.873 3.205 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.939 2.940 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.514 2.601 init_scf_loop 11 6.9 0.000 0.000 2.542 2.543 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.536 2.537 cp_fm_redistribute_end 50 14.0 1.290 2.493 1.299 2.497 cp_fm_diag_elpa_base 50 14.0 1.160 2.351 1.191 2.399 ot_diis_step 108 11.5 0.011 0.011 2.353 2.354 calculate_first_density_matrix 1 7.0 0.000 0.001 2.306 2.308 density_rs2pw 119 9.7 0.004 0.004 2.206 2.260 multiply_cannon_sync_h2d 27432 15.5 1.702 2.203 1.702 2.203 calculate_dm_sparse 119 9.5 0.000 0.001 2.109 2.189 pw_transfer 1439 11.6 0.065 0.069 2.042 2.077 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.950 1.987 acc_transpose_blocks 27432 15.5 0.114 0.120 1.616 1.952 grid_integrate_task_list 119 12.3 1.839 1.933 1.839 1.933 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.895 1.897 jit_kernel_multiply 10 16.3 0.670 1.885 0.670 1.885 potential_pw2rs 119 12.3 0.006 0.006 1.777 1.788 make_images_data 4572 15.5 0.047 0.054 1.336 1.734 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.674 1.713 prepare_preconditioner 11 7.9 0.000 0.000 1.607 1.633 make_preconditioner 11 8.9 0.000 0.000 1.607 1.633 fft3d_ps 1201 14.6 0.520 0.576 1.587 1.619 hybrid_alltoall_any 4725 16.4 0.054 0.116 1.193 1.567 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.504 1.563 wfi_extrapolate 11 7.9 0.001 0.001 1.527 1.527 fft_wrap_pw1pw2_140 487 13.2 0.159 0.168 1.472 1.509 mp_alltoall_d11v 2130 13.8 1.357 1.448 1.357 1.448 transfer_rs2pw 487 10.6 0.005 0.005 1.323 1.402 grid_collocate_task_list 119 9.7 1.287 1.356 1.287 1.356 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.260 1.305 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.260 1.267 transfer_pw2rs 487 13.2 0.004 0.005 1.150 1.160 mp_allgather_i34 2286 14.5 0.682 1.091 0.682 1.091 make_images_sizes 4572 15.5 0.005 0.005 0.758 1.066 mp_alltoall_i44 4572 16.5 0.753 1.061 0.753 1.061 mp_sum_d 4135 12.0 0.579 1.059 0.579 1.059 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.028 1.030 acc_transpose_blocks_sync 82296 16.5 0.820 0.948 0.820 0.948 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.898 0.912 mp_alltoall_z22v 1201 16.6 0.823 0.897 0.823 0.897 acc_transpose_blocks_kernels 27432 16.5 0.188 0.278 0.656 0.864 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=41.405000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=464.272727, yerr=1.135454 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/164be73c51fd9cb7dc5314b3417fa4c35b0a34c7_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 522.444800E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63490. MP_Allreduce 10155 305. MP_Sync 54 MP_Alltoall 1821 1607811. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.024 0.062 34.250 34.252 qs_mol_dyn_low 1 2.0 0.003 0.004 33.629 33.640 qs_forces 11 3.9 0.005 0.021 33.567 33.568 qs_energies 11 4.9 0.002 0.009 31.980 31.982 scf_env_do_scf 11 5.9 0.001 0.002 27.159 27.159 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 24.547 24.549 dbcsr_multiply_generic 2286 12.5 0.096 0.101 17.383 17.444 qs_scf_new_mos 108 7.5 0.001 0.001 16.386 16.405 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.386 16.405 velocity_verlet 10 3.0 0.001 0.002 15.863 15.865 ot_scf_mini 108 9.5 0.002 0.002 15.617 15.634 multiply_cannon 2286 13.5 0.195 0.200 13.699 14.435 multiply_cannon_loop 2286 14.5 0.859 0.886 12.859 13.624 ot_mini 108 10.5 0.001 0.001 9.429 9.450 qs_ot_get_derivative 108 11.5 0.001 0.001 7.914 7.932 multiply_cannon_multrec 18288 15.5 1.884 2.898 6.949 7.302 rebuild_ks_matrix 119 8.3 0.000 0.000 6.298 6.316 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.297 6.316 dbcsr_mm_accdrv_process 38222 16.0 4.912 5.676 4.970 5.741 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.563 5.579 mp_waitall_1 158411 16.6 2.944 4.101 2.944 4.101 qs_ot_get_p 119 10.4 0.001 0.001 3.869 3.891 sum_up_and_integrate 119 10.3 0.001 0.001 3.833 3.837 integrate_v_rspace 119 11.3 0.003 0.003 3.820 3.826 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.921 3.547 init_scf_run 11 5.9 0.000 0.001 3.533 3.533 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.533 3.533 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.503 3.509 calculate_rho_elec 119 8.7 0.031 0.031 3.503 3.508 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.612 2.617 init_scf_loop 11 6.9 0.000 0.000 2.592 2.593 multiply_cannon_metrocomm3 18288 15.5 0.047 0.049 1.439 2.399 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.049 2.370 apply_single 119 13.6 0.000 0.000 2.049 2.370 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.294 2.294 make_m2s 4572 13.5 0.044 0.045 2.123 2.271 calculate_first_density_matrix 1 7.0 0.000 0.000 2.227 2.228 make_images 4572 14.5 0.193 0.207 2.039 2.186 density_rs2pw 119 9.7 0.004 0.004 2.053 2.136 pw_transfer 1439 11.6 0.065 0.069 2.007 2.020 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.994 2.003 cp_fm_diag_elpa_base 50 14.0 1.968 1.984 1.992 2.001 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.914 1.930 grid_integrate_task_list 119 12.3 1.797 1.877 1.797 1.877 calculate_dm_sparse 119 9.5 0.000 0.001 1.856 1.864 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.854 1.863 prepare_preconditioner 11 7.9 0.000 0.000 1.807 1.809 make_preconditioner 11 8.9 0.000 0.000 1.807 1.809 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.661 1.747 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.717 1.718 acc_transpose_blocks 18288 15.5 0.080 0.082 1.631 1.673 potential_pw2rs 119 12.3 0.007 0.008 1.610 1.621 mp_sum_l 7287 12.8 1.194 1.571 1.194 1.571 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.536 1.543 ot_diis_step 108 11.5 0.011 0.012 1.499 1.499 fft3d_ps 1201 14.6 0.529 0.546 1.483 1.496 fft_wrap_pw1pw2_140 487 13.2 0.211 0.217 1.467 1.480 grid_collocate_task_list 119 9.7 1.232 1.305 1.232 1.305 transfer_rs2pw 487 10.6 0.005 0.005 1.179 1.287 wfi_extrapolate 11 7.9 0.001 0.001 1.243 1.243 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.194 1.198 multiply_cannon_sync_h2d 18288 15.5 1.029 1.146 1.029 1.146 make_images_data 4572 15.5 0.048 0.052 0.917 1.100 qs_energies_init_hamiltonians 11 5.9 0.016 0.019 1.027 1.028 transfer_pw2rs 487 13.2 0.004 0.004 1.002 1.014 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.985 1.002 hybrid_alltoall_any 4725 16.4 0.058 0.116 0.794 0.992 mp_alltoall_d11v 2130 13.8 0.768 0.860 0.768 0.860 mp_alltoall_z22v 1201 16.6 0.781 0.850 0.781 0.850 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.807 0.810 acc_transpose_blocks_kernels 18288 16.5 0.217 0.224 0.790 0.799 cp_fm_cholesky_invert 11 10.9 0.780 0.784 0.780 0.784 acc_transpose_blocks_sync 54864 16.5 0.742 0.780 0.742 0.780 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.675 0.725 multiply_cannon_metrocomm1 18288 15.5 0.030 0.031 0.432 0.697 mp_waitany 9880 13.7 0.576 0.695 0.576 0.695 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=34.252000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.818182, yerr=1.585054 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/164be73c51fd9cb7dc5314b3417fa4c35b0a34c7_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 557.469696E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63489. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.009 0.030 37.423 37.424 qs_mol_dyn_low 1 2.0 0.003 0.003 37.187 37.196 qs_forces 11 3.9 0.002 0.003 37.022 37.023 qs_energies 11 4.9 0.001 0.001 35.322 35.328 scf_env_do_scf 11 5.9 0.000 0.001 30.154 30.155 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 26.713 26.713 dbcsr_multiply_generic 2286 12.5 0.098 0.102 20.222 20.360 velocity_verlet 10 3.0 0.002 0.002 18.664 18.670 qs_scf_new_mos 108 7.5 0.001 0.001 18.427 18.487 qs_scf_loop_do_ot 108 8.5 0.001 0.001 18.426 18.487 ot_scf_mini 108 9.5 0.002 0.003 17.385 17.436 multiply_cannon 2286 13.5 0.219 0.229 16.415 17.111 multiply_cannon_loop 2286 14.5 1.527 1.622 15.423 15.977 ot_mini 108 10.5 0.001 0.001 10.488 10.552 multiply_cannon_multrec 27432 15.5 2.457 3.132 9.174 9.467 qs_ot_get_derivative 108 11.5 0.001 0.001 8.595 8.649 dbcsr_mm_accdrv_process 47916 15.9 5.949 7.438 6.614 7.930 rebuild_ks_matrix 119 8.3 0.000 0.000 6.440 6.493 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 6.440 6.493 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.727 5.773 qs_ot_get_p 119 10.4 0.001 0.001 4.002 4.075 init_scf_run 11 5.9 0.000 0.001 3.834 3.835 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.834 3.834 sum_up_and_integrate 119 10.3 0.001 0.001 3.613 3.622 integrate_v_rspace 119 11.3 0.003 0.003 3.602 3.610 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.487 3.531 calculate_rho_elec 119 8.7 0.040 0.046 3.486 3.531 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.191 3.515 init_scf_loop 11 6.9 0.000 0.000 3.420 3.421 acc_transpose_blocks 27432 15.5 0.120 0.124 2.426 2.607 prepare_preconditioner 11 7.9 0.000 0.000 2.551 2.558 make_preconditioner 11 8.9 0.000 0.000 2.551 2.558 make_m2s 4572 13.5 0.054 0.055 2.409 2.534 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.487 2.512 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.484 2.494 mp_waitall_1 137007 16.6 1.993 2.492 1.993 2.492 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.161 2.483 make_images 4572 14.5 0.274 0.335 2.303 2.428 calculate_first_density_matrix 1 7.0 0.001 0.009 2.424 2.426 calculate_dm_sparse 119 9.5 0.000 0.000 2.214 2.270 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.044 2.235 apply_single 119 13.6 0.000 0.000 2.044 2.235 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.060 2.061 density_rs2pw 119 9.7 0.004 0.004 1.921 2.029 pw_transfer 1439 11.6 0.065 0.069 1.988 2.028 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.895 1.939 grid_integrate_task_list 119 12.3 1.830 1.938 1.830 1.938 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.937 1.938 ot_diis_step 108 11.5 0.012 0.013 1.852 1.852 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.757 1.765 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.749 1.764 cp_fm_diag_elpa_base 50 14.0 1.721 1.738 1.754 1.763 acc_transpose_blocks_sync 82296 16.5 1.422 1.590 1.422 1.590 fft_wrap_pw1pw2_140 487 13.2 0.245 0.259 1.542 1.589 fft3d_ps 1201 14.6 0.555 0.605 1.415 1.441 potential_pw2rs 119 12.3 0.009 0.010 1.398 1.405 mp_sum_l 7287 12.8 0.973 1.395 0.973 1.395 multiply_cannon_metrocomm3 27432 15.5 0.041 0.042 0.851 1.388 jit_kernel_multiply 7 16.1 0.599 1.368 0.599 1.368 wfi_extrapolate 11 7.9 0.001 0.001 1.357 1.357 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.330 1.345 grid_collocate_task_list 119 9.7 1.248 1.336 1.248 1.336 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.304 1.317 cp_fm_upper_to_full 72 14.2 0.813 1.154 0.813 1.154 transfer_rs2pw 487 10.6 0.005 0.005 0.992 1.103 dbcsr_complete_redistribute 329 12.2 0.127 0.155 0.809 1.079 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.070 1.071 make_images_data 4572 15.5 0.048 0.051 0.913 1.053 mp_alltoall_d11v 2130 13.8 0.794 1.006 0.794 1.006 hybrid_alltoall_any 4725 16.4 0.065 0.155 0.785 0.979 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.800 0.876 acc_transpose_blocks_kernels 27432 16.5 0.272 0.278 0.855 0.870 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.587 0.851 transfer_pw2rs 487 13.2 0.004 0.004 0.819 0.826 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.805 0.810 cp_fm_cholesky_invert 11 10.9 0.768 0.771 0.768 0.771 mp_alltoall_z22v 1201 16.6 0.742 0.766 0.742 0.766 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=37.424000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=529.181818, yerr=2.948595 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/164be73c51fd9cb7dc5314b3417fa4c35b0a34c7_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 622.039040E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63488. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.009 0.032 30.298 30.299 qs_mol_dyn_low 1 2.0 0.003 0.003 29.909 29.923 qs_forces 11 3.9 0.002 0.003 29.815 29.816 qs_energies 11 4.9 0.001 0.001 28.105 28.108 scf_env_do_scf 11 5.9 0.000 0.001 23.091 23.091 scf_env_do_scf_inner_loop 108 6.5 0.006 0.028 20.556 20.557 velocity_verlet 10 3.0 0.019 0.041 14.931 14.934 dbcsr_multiply_generic 2286 12.5 0.092 0.096 13.830 13.958 qs_scf_new_mos 108 7.5 0.001 0.001 12.677 12.699 qs_scf_loop_do_ot 108 8.5 0.001 0.001 12.676 12.699 ot_scf_mini 108 9.5 0.002 0.002 11.951 11.982 multiply_cannon 2286 13.5 0.223 0.230 11.054 11.523 multiply_cannon_loop 2286 14.5 0.644 0.661 10.109 10.355 ot_mini 108 10.5 0.001 0.001 6.849 6.887 multiply_cannon_multrec 9144 15.5 1.747 1.999 6.277 6.493 rebuild_ks_matrix 119 8.3 0.000 0.000 5.868 5.886 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.868 5.886 qs_ot_get_derivative 108 11.5 0.001 0.001 5.523 5.555 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.234 5.250 dbcsr_mm_accdrv_process 12550 15.8 3.566 4.436 4.418 4.495 init_scf_run 11 5.9 0.000 0.001 3.539 3.539 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.538 3.538 sum_up_and_integrate 119 10.3 0.001 0.001 3.532 3.537 integrate_v_rspace 119 11.3 0.003 0.003 3.522 3.527 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.472 3.480 calculate_rho_elec 119 8.7 0.060 0.061 3.472 3.480 qs_ot_get_p 119 10.4 0.001 0.001 3.153 3.191 init_scf_loop 11 6.9 0.000 0.000 2.510 2.513 calculate_first_density_matrix 1 7.0 0.000 0.000 2.365 2.365 mp_waitall_1 115863 16.7 1.701 2.214 1.701 2.214 pw_transfer 1439 11.6 0.066 0.068 2.084 2.095 make_m2s 4572 13.5 0.034 0.034 1.910 2.083 qs_ot_p2m_diag 50 11.0 0.022 0.023 2.017 2.020 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.991 2.002 make_images 4572 14.5 0.270 0.300 1.822 1.994 density_rs2pw 119 9.7 0.003 0.003 1.888 1.977 grid_integrate_task_list 119 12.3 1.860 1.952 1.860 1.952 calculate_dm_sparse 119 9.5 0.000 0.000 1.876 1.895 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.836 1.860 prepare_preconditioner 11 7.9 0.000 0.000 1.784 1.789 make_preconditioner 11 8.9 0.000 0.000 1.784 1.789 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.758 1.759 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.719 1.721 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.672 1.699 fft_wrap_pw1pw2_140 487 13.2 0.321 0.329 1.602 1.616 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.540 1.548 acc_transpose_blocks 9144 15.5 0.042 0.043 1.514 1.538 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.465 1.472 cp_fm_diag_elpa_base 50 14.0 1.436 1.454 1.462 1.470 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.451 1.467 fft3d_ps 1201 14.6 0.561 0.576 1.406 1.414 grid_collocate_task_list 119 9.7 1.297 1.382 1.297 1.382 potential_pw2rs 119 12.3 0.010 0.011 1.334 1.338 ot_diis_step 108 11.5 0.013 0.013 1.311 1.311 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.242 1.272 apply_single 119 13.6 0.000 0.000 1.242 1.272 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.257 1.263 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.248 1.249 jit_kernel_multiply 6 15.6 0.813 1.214 0.813 1.214 wfi_extrapolate 11 7.9 0.001 0.001 1.124 1.124 hybrid_alltoall_any 4725 16.4 0.065 0.174 0.821 1.057 make_images_data 4572 15.5 0.042 0.045 0.834 1.045 transfer_rs2pw 487 10.6 0.004 0.004 0.907 0.988 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.873 0.924 cp_fm_cholesky_invert 11 10.9 0.916 0.919 0.916 0.919 mp_alltoall_d11v 2130 13.8 0.803 0.915 0.803 0.915 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.870 0.876 multiply_cannon_metrocomm3 9144 15.5 0.020 0.021 0.442 0.777 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.760 0.762 mp_alltoall_z22v 1201 16.6 0.718 0.754 0.718 0.754 acc_transpose_blocks_sync 27432 16.5 0.723 0.748 0.723 0.748 qs_env_update_s_mstruct 11 6.9 0.000 0.001 0.694 0.741 acc_transpose_blocks_kernels 9144 16.5 0.118 0.120 0.734 0.736 transfer_pw2rs 487 13.2 0.003 0.004 0.724 0.728 mp_allgather_i34 2286 14.5 0.261 0.704 0.261 0.704 jit_kernel_transpose 5 15.6 0.616 0.618 0.616 0.618 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=30.299000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=592.181818, yerr=3.856946 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/164be73c51fd9cb7dc5314b3417fa4c35b0a34c7_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 783.659008E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63723. MP_Allreduce 10154 429. MP_Sync 54 MP_Alltoall 1582 7383731. MP_SendRecv 2499 189067. MP_ISendRecv 2499 189067. MP_Wait 6399 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.031 43.961 43.962 qs_mol_dyn_low 1 2.0 0.003 0.003 43.505 43.514 qs_forces 11 3.9 0.002 0.002 43.441 43.443 qs_energies 11 4.9 0.006 0.007 41.391 41.396 scf_env_do_scf 11 5.9 0.001 0.001 35.474 35.474 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 27.564 27.565 velocity_verlet 10 3.0 0.002 0.002 24.637 24.642 dbcsr_multiply_generic 2286 12.5 0.101 0.104 19.512 19.776 qs_scf_new_mos 108 7.5 0.001 0.001 17.903 17.990 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.903 17.989 ot_scf_mini 108 9.5 0.002 0.002 16.722 16.817 multiply_cannon 2286 13.5 0.302 0.307 15.263 16.316 multiply_cannon_loop 2286 14.5 0.862 0.882 13.925 14.987 ot_mini 108 10.5 0.001 0.001 10.091 10.204 multiply_cannon_multrec 9144 15.5 3.360 4.619 8.832 8.999 qs_ot_get_derivative 108 11.5 0.001 0.001 7.966 8.060 init_scf_loop 11 6.9 0.000 0.000 7.881 7.882 rebuild_ks_matrix 119 8.3 0.000 0.000 7.268 7.406 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.268 7.406 prepare_preconditioner 11 7.9 0.000 0.000 6.916 6.930 make_preconditioner 11 8.9 0.000 0.000 6.916 6.930 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.520 6.805 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.542 6.668 dbcsr_mm_accdrv_process 12550 15.8 4.700 6.635 5.330 6.667 cp_fm_upper_to_full 72 14.2 3.162 4.498 3.162 4.498 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.153 4.160 calculate_rho_elec 119 8.7 0.118 0.121 4.153 4.159 sum_up_and_integrate 119 10.3 0.001 0.001 3.899 3.907 integrate_v_rspace 119 11.3 0.003 0.004 3.888 3.896 init_scf_run 11 5.9 0.000 0.001 3.729 3.729 scf_env_initial_rho_setup 11 6.9 0.010 0.011 3.728 3.729 qs_ot_get_p 119 10.4 0.001 0.001 3.589 3.720 mp_waitall_1 94719 16.7 2.590 3.600 2.590 3.600 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.891 3.311 make_m2s 4572 13.5 0.037 0.038 2.586 2.792 dbcsr_complete_redistribute 329 12.2 0.286 0.295 1.985 2.774 pw_transfer 1439 11.6 0.069 0.069 2.704 2.712 make_images 4572 14.5 0.352 0.386 2.466 2.671 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.606 2.612 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.277 2.569 apply_single 119 13.6 0.000 0.000 2.277 2.569 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.644 2.437 multiply_cannon_metrocomm3 9144 15.5 0.021 0.021 1.392 2.328 calculate_dm_sparse 119 9.5 0.000 0.000 2.261 2.283 mp_alltoall_i22 627 13.8 1.372 2.207 1.372 2.207 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.149 2.198 calculate_first_density_matrix 1 7.0 0.000 0.000 2.194 2.197 fft_wrap_pw1pw2_140 487 13.2 0.576 0.580 2.170 2.178 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.392 2.175 density_rs2pw 119 9.7 0.003 0.003 2.139 2.169 grid_integrate_task_list 119 12.3 2.099 2.143 2.099 2.143 qs_ot_p2m_diag 50 11.0 0.044 0.044 2.132 2.134 ot_diis_step 108 11.5 0.014 0.015 2.090 2.091 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.968 1.969 qs_energies_init_hamiltonians 11 5.9 0.002 0.003 1.892 1.894 mp_sum_l 7287 12.8 1.119 1.846 1.119 1.846 acc_transpose_blocks 9144 15.5 0.044 0.045 1.809 1.836 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.764 1.765 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.695 1.738 fft3d_ps 1201 14.6 0.595 0.603 1.701 1.705 grid_collocate_task_list 119 9.7 1.535 1.562 1.535 1.562 cp_fm_cholesky_invert 11 10.9 1.548 1.554 1.548 1.554 hybrid_alltoall_any 4725 16.4 0.090 0.149 1.182 1.473 wfi_extrapolate 11 7.9 0.001 0.001 1.459 1.459 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.456 1.457 cp_fm_diag_elpa_base 50 14.0 1.312 1.363 1.455 1.456 make_images_data 4572 15.5 0.046 0.049 1.155 1.420 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.379 1.402 potential_pw2rs 119 12.3 0.014 0.015 1.363 1.366 mp_alltoall_d11v 2130 13.8 1.256 1.303 1.256 1.303 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.248 1.257 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.214 1.227 acc_transpose_blocks_sync 27432 16.5 1.095 1.123 1.095 1.123 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 0.996 1.041 mp_alltoall_z22v 1201 16.6 0.970 0.987 0.970 0.987 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.939 0.952 qs_create_task_list 11 7.9 0.000 0.001 0.933 0.946 generate_qs_task_list 11 8.9 0.368 0.387 0.933 0.946 transfer_rs2pw 487 10.6 0.005 0.005 0.824 0.883 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=43.962000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=737.363636, yerr=12.864841 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/164be73c51fd9cb7dc5314b3417fa4c35b0a34c7_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 499.855360E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66212. MP_Allreduce 9776 488. MP_Sync 52 MP_Alltoall 1938 1383689. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.008 0.026 85.106 85.107 qs_mol_dyn_low 1 2.0 0.003 0.003 84.698 84.708 qs_forces 11 3.9 0.011 0.014 84.624 84.625 qs_energies 11 4.9 0.001 0.003 81.649 81.664 scf_env_do_scf 11 5.9 0.000 0.001 72.607 72.609 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 66.622 66.623 dbcsr_multiply_generic 2055 12.4 0.107 0.111 52.147 52.470 qs_scf_new_mos 99 7.5 0.000 0.001 48.970 49.078 qs_scf_loop_do_ot 99 8.5 0.000 0.001 48.969 49.077 ot_scf_mini 99 9.5 0.002 0.002 46.469 46.585 velocity_verlet 10 3.0 0.001 0.002 45.514 45.514 multiply_cannon 2055 13.4 0.183 0.188 42.896 43.954 multiply_cannon_loop 2055 14.4 1.799 1.843 41.825 42.912 ot_mini 99 10.5 0.001 0.001 27.356 27.485 qs_ot_get_derivative 99 11.5 0.001 0.001 20.435 20.563 multiply_cannon_multrec 49320 15.4 11.308 12.031 17.345 17.982 rebuild_ks_matrix 110 8.3 0.000 0.000 14.977 15.183 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.013 14.977 15.182 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.105 13.292 mp_waitall_1 220248 16.4 11.420 12.147 11.420 12.147 qs_ot_get_p 110 10.4 0.001 0.001 10.378 10.502 multiply_cannon_sync_h2d 49320 15.4 9.513 10.228 9.513 10.228 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 8.037 8.558 multiply_cannon_metrocomm3 49320 15.4 0.083 0.088 6.660 7.931 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.242 7.816 apply_single 110 13.6 0.000 0.000 7.242 7.816 sum_up_and_integrate 110 10.3 0.002 0.003 7.333 7.347 integrate_v_rspace 110 11.3 0.003 0.003 7.307 7.328 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.946 6.965 init_scf_run 11 5.9 0.000 0.001 6.907 6.907 scf_env_initial_rho_setup 11 6.9 0.001 0.002 6.907 6.907 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.639 6.791 calculate_rho_elec 110 8.6 0.020 0.024 6.639 6.790 ot_diis_step 99 11.5 0.006 0.006 6.674 6.675 dbcsr_mm_accdrv_process 87628 16.1 3.073 3.154 5.907 6.158 cp_dbcsr_syevd 48 12.0 0.003 0.003 6.082 6.083 init_scf_loop 11 6.9 0.000 0.000 5.951 5.952 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.443 5.512 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.462 5.496 cp_fm_diag_elpa_base 48 14.0 5.449 5.484 5.461 5.494 mp_sum_l 6594 12.7 4.496 5.363 4.496 5.363 make_m2s 4110 13.4 0.060 0.066 4.294 4.414 make_images 4110 14.4 0.178 0.190 4.199 4.321 wfi_extrapolate 11 7.9 0.001 0.001 4.197 4.197 calculate_dm_sparse 110 9.5 0.001 0.001 3.890 3.984 density_rs2pw 110 9.6 0.004 0.005 3.624 3.906 multiply_cannon_metrocomm1 49320 15.4 0.065 0.068 2.624 3.812 prepare_preconditioner 11 7.9 0.000 0.000 3.719 3.750 make_preconditioner 11 8.9 0.000 0.000 3.719 3.750 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.594 3.598 pw_transfer 1331 11.6 0.055 0.067 3.433 3.531 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.487 3.528 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.421 3.481 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.344 3.446 grid_integrate_task_list 110 12.3 3.260 3.396 3.260 3.396 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.226 3.277 fft_wrap_pw1pw2_140 451 13.1 0.389 0.431 2.844 2.951 potential_pw2rs 110 12.3 0.005 0.006 2.852 2.879 mp_alltoall_d11v 2046 13.8 2.252 2.801 2.252 2.801 fft3d_ps 1111 14.6 0.792 0.879 2.640 2.714 acc_transpose_blocks 49320 15.4 0.213 0.220 2.574 2.638 calculate_first_density_matrix 1 7.0 0.000 0.001 2.602 2.610 jit_kernel_multiply 13 15.9 2.554 2.575 2.554 2.575 transfer_rs2pw 451 10.6 0.005 0.006 2.041 2.322 grid_collocate_task_list 110 9.6 2.161 2.277 2.161 2.277 mp_waitany 14300 13.8 1.880 2.159 1.880 2.159 make_images_data 4110 15.4 0.044 0.047 1.920 2.085 cp_fm_cholesky_invert 11 10.9 2.026 2.031 2.026 2.031 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.003 2.031 mp_sum_d 3889 11.9 1.462 1.992 1.462 1.992 transfer_pw2rs 451 13.1 0.006 0.007 1.918 1.941 hybrid_alltoall_any 4261 16.3 0.084 0.483 1.651 1.925 mp_alltoall_z22v 1111 16.6 1.556 1.718 1.556 1.718 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=85.107000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=475.636364, yerr=2.267266 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/164be73c51fd9cb7dc5314b3417fa4c35b0a34c7_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 591.581184E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1972800 MPI messages size (bytes): total size 1.077520E+12 min size 0.000000E+00 max size 4.537280E+06 average size 546.188250E+03 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 222984 1826684928 8192 < size <= 32768 520356 13399818240 32768 < size <= 131072 372336 35386294272 131072 < size <= 4194304 787758 788321309808 4194304 < size <= 16777216 54450 238588003280 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66430. MP_Allreduce 9775 566. MP_Sync 52 MP_Alltoall 1717 1628374. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.041 72.727 72.728 qs_mol_dyn_low 1 2.0 0.003 0.006 72.145 72.174 qs_forces 11 3.9 0.003 0.004 72.070 72.071 qs_energies 11 4.9 0.004 0.011 68.700 68.704 scf_env_do_scf 11 5.9 0.001 0.005 59.756 59.759 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 51.397 51.398 dbcsr_multiply_generic 2055 12.4 0.123 0.127 38.983 39.261 velocity_verlet 10 3.0 0.001 0.002 38.310 38.311 qs_scf_new_mos 99 7.5 0.001 0.001 34.864 34.983 qs_scf_loop_do_ot 99 8.5 0.001 0.001 34.864 34.982 ot_scf_mini 99 9.5 0.003 0.004 33.133 33.239 multiply_cannon 2055 13.4 0.224 0.243 31.625 32.591 multiply_cannon_loop 2055 14.4 1.166 1.196 30.194 31.141 ot_mini 99 10.5 0.001 0.001 19.245 19.379 multiply_cannon_multrec 24660 15.4 6.918 8.202 13.981 15.228 rebuild_ks_matrix 110 8.3 0.000 0.000 13.904 14.013 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 13.904 14.012 qs_ot_get_derivative 99 11.5 0.001 0.001 13.352 13.461 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.242 12.329 mp_waitall_1 176588 16.5 8.145 10.972 8.145 10.972 init_scf_loop 11 6.9 0.001 0.004 8.320 8.321 multiply_cannon_metrocomm3 24660 15.4 0.073 0.076 5.409 8.321 multiply_cannon_sync_h2d 24660 15.4 6.263 7.504 6.263 7.504 dbcsr_mm_accdrv_process 52282 16.1 5.540 6.475 6.896 7.250 qs_ot_get_p 110 10.4 0.001 0.001 7.097 7.238 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.551 7.077 apply_single 110 13.6 0.000 0.001 6.550 7.077 sum_up_and_integrate 110 10.3 0.001 0.002 6.679 6.693 integrate_v_rspace 110 11.3 0.003 0.003 6.652 6.662 init_scf_run 11 5.9 0.000 0.001 6.336 6.336 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.335 6.336 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.306 6.328 calculate_rho_elec 110 8.6 0.039 0.047 6.305 6.328 prepare_preconditioner 11 7.9 0.000 0.000 6.255 6.272 make_preconditioner 11 8.9 0.001 0.001 6.255 6.272 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.844 6.003 ot_diis_step 99 11.5 0.010 0.010 5.821 5.821 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.995 5.751 make_m2s 4110 13.4 0.057 0.060 4.638 5.020 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.992 5.017 make_images 4110 14.4 0.409 0.467 4.529 4.907 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.508 4.509 pw_transfer 1331 11.6 0.067 0.075 3.870 4.021 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.763 3.915 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.880 3.896 cp_fm_diag_elpa_base 48 14.0 3.830 3.848 3.877 3.893 density_rs2pw 110 9.6 0.004 0.004 3.481 3.700 wfi_extrapolate 11 7.9 0.001 0.001 3.692 3.693 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.481 3.549 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.372 3.373 grid_integrate_task_list 110 12.3 3.149 3.308 3.149 3.308 fft_wrap_pw1pw2_140 451 13.1 0.460 0.474 3.091 3.245 calculate_dm_sparse 110 9.5 0.001 0.001 3.036 3.070 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.986 3.038 cp_fm_cholesky_invert 11 10.9 3.029 3.037 3.029 3.037 fft3d_ps 1111 14.6 1.111 1.350 2.859 3.021 make_images_data 4110 15.4 0.049 0.052 2.523 2.927 hybrid_alltoall_any 4261 16.3 0.106 0.450 2.229 2.898 mp_sum_l 6594 12.7 2.088 2.875 2.088 2.875 calculate_first_density_matrix 1 7.0 0.001 0.004 2.545 2.547 potential_pw2rs 110 12.3 0.008 0.009 2.509 2.520 mp_alltoall_d11v 2046 13.8 2.010 2.367 2.010 2.367 grid_collocate_task_list 110 9.6 2.168 2.331 2.168 2.331 qs_ot_get_orbitals 99 10.5 0.001 0.001 2.079 2.101 acc_transpose_blocks 24660 15.4 0.116 0.118 1.979 2.008 qs_energies_init_hamiltonians 11 5.9 0.007 0.013 2.004 2.005 transfer_rs2pw 451 10.6 0.006 0.007 1.719 1.972 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.825 1.841 mp_allgather_i34 2055 14.4 0.737 1.827 0.737 1.827 multiply_cannon_metrocomm4 22605 15.4 0.079 0.084 0.798 1.777 jit_kernel_multiply 9 16.2 0.991 1.729 0.991 1.729 mp_waitany 10164 13.8 1.330 1.644 1.330 1.644 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.567 1.579 mp_irecv_dv 57340 16.2 0.670 1.556 0.670 1.556 cp_fm_cholesky_decompose 22 10.9 1.545 1.551 1.545 1.551 mp_alltoall_z22v 1111 16.6 1.475 1.548 1.475 1.548 dbcsr_complete_redistribute 325 12.2 0.232 0.297 1.256 1.533 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.386 1.487 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=72.728000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=557.818182, yerr=6.575688 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/164be73c51fd9cb7dc5314b3417fa4c35b0a34c7_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 665.948160E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66421. MP_Allreduce 9774 562. MP_Sync 52 MP_Alltoall 1496 4511006. MP_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.196 0.219 63.817 63.819 qs_mol_dyn_low 1 2.0 0.003 0.004 63.102 63.112 qs_forces 11 3.9 0.003 0.008 63.029 63.030 qs_energies 11 4.9 0.006 0.036 59.777 59.787 scf_env_do_scf 11 5.9 0.001 0.001 51.417 51.417 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 41.900 41.900 velocity_verlet 10 3.0 0.001 0.002 35.101 35.103 dbcsr_multiply_generic 2055 12.4 0.109 0.115 29.840 30.197 qs_scf_new_mos 99 7.5 0.001 0.001 26.677 26.779 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.677 26.778 ot_scf_mini 99 9.5 0.002 0.003 25.372 25.487 multiply_cannon 2055 13.4 0.212 0.221 23.013 24.296 multiply_cannon_loop 2055 14.4 0.816 0.845 21.709 23.177 ot_mini 99 10.5 0.001 0.001 14.560 14.688 rebuild_ks_matrix 110 8.3 0.000 0.000 12.543 12.702 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.543 12.701 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.053 11.192 mp_waitall_1 139946 16.5 7.743 10.760 7.743 10.760 multiply_cannon_multrec 16440 15.4 3.521 4.575 9.713 10.735 qs_ot_get_derivative 99 11.5 0.001 0.001 9.947 10.065 init_scf_loop 11 6.9 0.001 0.004 9.478 9.479 multiply_cannon_metrocomm3 16440 15.4 0.046 0.048 4.844 7.934 prepare_preconditioner 11 7.9 0.000 0.000 7.626 7.645 make_preconditioner 11 8.9 0.000 0.001 7.626 7.645 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.937 7.305 sum_up_and_integrate 110 10.3 0.002 0.003 6.579 6.593 integrate_v_rspace 110 11.3 0.003 0.004 6.552 6.567 dbcsr_mm_accdrv_process 34862 16.1 5.359 5.649 6.040 6.180 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.063 6.086 calculate_rho_elec 110 8.6 0.058 0.059 6.063 6.085 qs_ot_get_p 110 10.4 0.001 0.001 5.832 5.986 init_scf_run 11 5.9 0.000 0.001 5.762 5.762 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.762 5.762 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.063 5.583 apply_single 110 13.6 0.000 0.000 5.063 5.583 make_m2s 4110 13.4 0.049 0.051 4.454 4.825 make_images 4110 14.4 0.400 0.524 4.340 4.712 ot_diis_step 99 11.5 0.011 0.011 4.558 4.560 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.448 4.102 qs_ot_p2m_diag 48 11.0 0.042 0.044 4.044 4.050 pw_transfer 1331 11.6 0.066 0.074 3.798 3.808 multiply_cannon_sync_h2d 16440 15.4 3.191 3.735 3.191 3.735 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.690 3.704 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.658 3.658 density_rs2pw 110 9.6 0.004 0.004 3.238 3.419 grid_integrate_task_list 110 12.3 3.195 3.401 3.195 3.401 wfi_extrapolate 11 7.9 0.001 0.001 3.208 3.208 fft_wrap_pw1pw2_140 451 13.1 0.580 0.587 3.108 3.127 cp_fm_cholesky_invert 11 10.9 3.103 3.112 3.103 3.112 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.073 3.083 cp_fm_diag_elpa_base 48 14.0 3.003 3.039 3.071 3.081 hybrid_alltoall_any 4261 16.3 0.109 0.383 2.296 2.967 make_images_data 4110 15.4 0.045 0.049 2.487 2.949 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.896 2.897 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.582 2.646 fft3d_ps 1111 14.6 1.092 1.105 2.631 2.641 calculate_dm_sparse 110 9.5 0.001 0.001 2.588 2.626 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.502 2.567 multiply_cannon_metrocomm4 14385 15.4 0.049 0.053 0.914 2.547 mp_sum_l 6594 12.7 1.756 2.524 1.756 2.524 calculate_first_density_matrix 1 7.0 0.001 0.004 2.451 2.453 mp_irecv_dv 48980 15.7 0.838 2.410 0.838 2.410 grid_collocate_task_list 110 9.6 2.223 2.403 2.223 2.403 potential_pw2rs 110 12.3 0.011 0.011 2.385 2.395 mp_alltoall_d11v 2046 13.8 2.041 2.266 2.041 2.266 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 2.077 2.079 dbcsr_complete_redistribute 325 12.2 0.320 0.390 1.583 2.057 cp_fm_cholesky_decompose 22 10.9 1.873 1.898 1.873 1.898 cp_fm_upper_to_full 70 14.2 1.433 1.808 1.433 1.808 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.688 1.710 mp_allgather_i34 2055 14.4 0.615 1.708 0.615 1.708 transfer_rs2pw 451 10.6 0.005 0.005 1.478 1.651 acc_transpose_blocks 16440 15.4 0.077 0.079 1.585 1.642 copy_fm_to_dbcsr 174 11.2 0.001 0.002 1.042 1.515 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.476 1.491 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.370 1.471 mp_waitany 17072 13.8 1.226 1.409 1.226 1.409 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.379 1.389 transfer_pw2rs 451 13.1 0.005 0.005 1.342 1.347 mp_alltoall_z22v 1111 16.6 1.324 1.345 1.324 1.345 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=63.819000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=631.000000, yerr=8.655004 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/164be73c51fd9cb7dc5314b3417fa4c35b0a34c7_performance_tests/12/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 601.317074E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 4916280 0.0% 0.0% 100.0% average stack size 0.0 0.0 200.2 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 731.656192E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66419. MP_Allreduce 9774 603. MP_Sync 52 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.027 0.049 67.141 67.143 qs_mol_dyn_low 1 2.0 0.003 0.003 66.774 66.785 qs_forces 11 3.9 0.003 0.003 66.707 66.708 qs_energies 11 4.9 0.001 0.001 63.258 63.261 scf_env_do_scf 11 5.9 0.000 0.001 54.917 54.920 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 42.837 42.837 velocity_verlet 10 3.0 0.001 0.002 38.124 38.126 dbcsr_multiply_generic 2055 12.4 0.115 0.119 30.984 31.236 qs_scf_new_mos 99 7.5 0.001 0.001 28.168 28.273 qs_scf_loop_do_ot 99 8.5 0.001 0.001 28.167 28.272 ot_scf_mini 99 9.5 0.002 0.003 26.449 26.549 multiply_cannon 2055 13.4 0.238 0.255 23.696 24.856 multiply_cannon_loop 2055 14.4 1.403 1.454 22.313 23.036 ot_mini 99 10.5 0.001 0.001 15.263 15.383 multiply_cannon_multrec 24660 15.4 4.100 6.663 13.027 14.108 rebuild_ks_matrix 110 8.3 0.000 0.000 12.034 12.160 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.034 12.160 init_scf_loop 11 6.9 0.000 0.000 12.037 12.038 qs_ot_get_derivative 99 11.5 0.001 0.001 11.047 11.153 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.610 10.720 prepare_preconditioner 11 7.9 0.000 0.000 10.283 10.296 make_preconditioner 11 8.9 0.000 0.000 10.283 10.296 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.497 9.976 dbcsr_mm_accdrv_process 52304 16.0 7.615 8.959 8.771 9.740 mp_waitall_1 121746 16.5 4.548 6.411 4.548 6.411 sum_up_and_integrate 110 10.3 0.001 0.002 6.306 6.317 integrate_v_rspace 110 11.3 0.003 0.003 6.280 6.293 qs_ot_get_p 110 10.4 0.001 0.001 6.045 6.168 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.886 5.894 calculate_rho_elec 110 8.6 0.078 0.081 5.886 5.893 make_m2s 4110 13.4 0.059 0.061 5.583 5.833 make_images 4110 14.4 0.580 0.705 5.444 5.690 init_scf_run 11 5.9 0.000 0.001 5.578 5.578 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.577 5.578 cp_fm_upper_to_full 70 14.2 3.354 4.873 3.354 4.873 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.083 4.194 apply_single 110 13.6 0.000 0.000 4.083 4.193 ot_diis_step 99 11.5 0.011 0.011 4.172 4.173 qs_ot_p2m_diag 48 11.0 0.055 0.064 4.129 4.146 dbcsr_complete_redistribute 325 12.2 0.418 0.458 2.767 3.925 pw_transfer 1331 11.6 0.066 0.075 3.728 3.762 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.621 3.657 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.580 3.623 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.566 3.566 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.396 3.459 grid_integrate_task_list 110 12.3 3.271 3.448 3.271 3.448 multiply_cannon_metrocomm3 24660 15.4 0.038 0.040 1.547 3.426 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.165 3.302 density_rs2pw 110 9.6 0.004 0.004 2.996 3.211 fft_wrap_pw1pw2_140 451 13.1 0.610 0.631 3.094 3.131 make_images_data 4110 15.4 0.049 0.053 2.791 3.072 wfi_extrapolate 11 7.9 0.001 0.001 3.068 3.068 calculate_dm_sparse 110 9.5 0.001 0.001 3.020 3.055 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.996 3.009 cp_fm_diag_elpa_base 48 14.0 2.848 2.910 2.993 3.007 hybrid_alltoall_any 4261 16.3 0.123 0.460 2.428 3.001 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.777 2.899 cp_fm_cholesky_invert 11 10.9 2.871 2.880 2.871 2.880 mp_alltoall_i22 605 13.7 1.669 2.872 1.669 2.872 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.817 2.819 multiply_cannon_sync_h2d 24660 15.4 2.343 2.620 2.343 2.620 fft3d_ps 1111 14.6 1.082 1.112 2.502 2.517 acc_transpose_blocks 24660 15.4 0.114 0.118 2.395 2.489 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.427 2.474 grid_collocate_task_list 110 9.6 2.268 2.418 2.268 2.418 calculate_first_density_matrix 1 7.0 0.000 0.000 2.413 2.417 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.245 2.245 potential_pw2rs 110 12.3 0.013 0.013 2.122 2.132 mp_alltoall_d11v 2046 13.8 1.847 1.992 1.847 1.992 cp_fm_cholesky_decompose 22 10.9 1.822 1.879 1.822 1.879 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.831 1.864 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.610 1.702 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.667 1.680 mp_sum_l 6594 12.7 1.021 1.630 1.021 1.630 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.547 1.563 transfer_rs2pw 451 10.6 0.005 0.006 1.234 1.474 jit_kernel_multiply 8 15.9 0.810 1.468 0.810 1.468 acc_transpose_blocks_sync 73980 16.4 1.385 1.467 1.385 1.467 mp_waitany 13376 13.8 1.061 1.372 1.061 1.372 multiply_cannon_metrocomm4 20550 15.4 0.063 0.066 0.875 1.364 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=67.143000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=694.545455, yerr=8.392244 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/164be73c51fd9cb7dc5314b3417fa4c35b0a34c7_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 858.480640E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66417. MP_Allreduce 9774 644. MP_Sync 52 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.048 55.458 55.459 qs_mol_dyn_low 1 2.0 0.003 0.003 55.185 55.194 qs_forces 11 3.9 0.003 0.003 55.120 55.121 qs_energies 11 4.9 0.001 0.001 51.492 51.497 scf_env_do_scf 11 5.9 0.000 0.001 43.289 43.290 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 35.469 35.470 velocity_verlet 10 3.0 0.001 0.002 31.332 31.335 dbcsr_multiply_generic 2055 12.4 0.105 0.109 23.325 23.482 qs_scf_new_mos 99 7.5 0.001 0.001 20.975 21.028 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.975 21.027 ot_scf_mini 99 9.5 0.002 0.002 19.723 19.770 multiply_cannon 2055 13.4 0.238 0.252 17.744 19.010 multiply_cannon_loop 2055 14.4 0.601 0.628 16.473 16.734 rebuild_ks_matrix 110 8.3 0.000 0.000 11.566 11.612 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 11.566 11.611 ot_mini 99 10.5 0.001 0.001 10.837 10.879 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.294 10.335 multiply_cannon_multrec 8220 15.4 3.183 4.688 7.590 8.906 init_scf_loop 11 6.9 0.000 0.000 7.773 7.774 mp_waitall_1 103326 16.6 5.987 7.645 5.987 7.645 qs_ot_get_derivative 99 11.5 0.001 0.001 7.094 7.142 sum_up_and_integrate 110 10.3 0.001 0.002 6.179 6.193 integrate_v_rspace 110 11.3 0.003 0.003 6.152 6.166 prepare_preconditioner 11 7.9 0.000 0.000 6.129 6.138 make_preconditioner 11 8.9 0.000 0.000 6.129 6.137 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.962 5.977 calculate_rho_elec 110 8.6 0.114 0.114 5.961 5.976 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.697 5.772 dbcsr_mm_accdrv_process 17442 15.9 3.116 4.154 4.266 5.194 init_scf_run 11 5.9 0.000 0.001 5.066 5.066 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.066 5.066 qs_ot_get_p 110 10.4 0.001 0.001 4.898 4.963 multiply_cannon_metrocomm3 8220 15.4 0.019 0.020 3.082 4.709 make_m2s 4110 13.4 0.038 0.039 4.235 4.479 make_images 4110 14.4 0.651 0.711 4.106 4.346 pw_transfer 1331 11.6 0.066 0.070 3.833 3.846 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.776 3.825 apply_single 110 13.6 0.000 0.000 3.776 3.825 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.725 3.743 ot_diis_step 99 11.5 0.012 0.012 3.719 3.719 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.494 3.499 grid_integrate_task_list 110 12.3 3.394 3.467 3.394 3.467 fft_wrap_pw1pw2_140 451 13.1 0.777 0.788 3.218 3.246 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.143 3.144 density_rs2pw 110 9.6 0.004 0.004 2.930 3.082 cp_fm_cholesky_invert 11 10.9 2.970 2.974 2.970 2.974 wfi_extrapolate 11 7.9 0.001 0.001 2.739 2.739 hybrid_alltoall_any 4261 16.3 0.200 0.852 2.300 2.720 qs_energies_init_hamiltonians 11 5.9 0.003 0.009 2.666 2.667 make_images_data 4110 15.4 0.040 0.046 2.267 2.594 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.554 2.561 cp_fm_diag_elpa_base 48 14.0 2.497 2.521 2.553 2.560 calculate_dm_sparse 110 9.5 0.001 0.001 2.477 2.520 multiply_cannon_sync_h2d 8220 15.4 2.349 2.514 2.349 2.514 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.505 2.506 grid_collocate_task_list 110 9.6 2.364 2.462 2.364 2.462 fft3d_ps 1111 14.6 1.135 1.163 2.390 2.406 calculate_first_density_matrix 1 7.0 0.000 0.000 2.233 2.234 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.125 2.151 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.030 2.065 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.787 1.992 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.958 1.983 potential_pw2rs 110 12.3 0.016 0.016 1.966 1.968 mp_alltoall_d11v 2046 13.8 1.801 1.916 1.801 1.916 cp_fm_cholesky_decompose 22 10.9 1.684 1.708 1.684 1.708 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.602 1.606 qs_env_update_s_mstruct 11 6.9 0.000 0.001 1.487 1.603 dbcsr_complete_redistribute 325 12.2 0.561 0.588 1.474 1.575 mp_allgather_i34 2055 14.4 0.440 1.525 0.440 1.525 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.432 1.443 acc_transpose_blocks 8220 15.4 0.039 0.040 1.361 1.401 multiply_cannon_metrocomm4 6165 15.4 0.019 0.020 0.478 1.382 mp_irecv_dv 24056 15.7 0.451 1.334 0.451 1.334 qs_create_task_list 11 7.9 0.000 0.001 1.207 1.307 generate_qs_task_list 11 8.9 0.373 0.440 1.206 1.307 transfer_rs2pw 451 10.6 0.005 0.005 1.126 1.288 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.241 1.262 mp_waitany 9240 13.8 1.035 1.217 1.035 1.217 multiply_cannon_metrocomm1 8220 15.4 0.022 0.023 0.800 1.214 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=55.459000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=807.363636, yerr=15.812956 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/164be73c51fd9cb7dc5314b3417fa4c35b0a34c7_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 1.375261E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67098. MP_Allreduce 9752 812. MP_Sync 52 MP_Alltoall 1474 16505187. MP_SendRecv 2310 360267. MP_ISendRecv 2310 360267. MP_Wait 5214 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.036 88.263 88.264 qs_mol_dyn_low 1 2.0 0.003 0.003 87.828 87.837 qs_forces 11 3.9 0.003 0.003 87.546 87.547 qs_energies 11 4.9 0.001 0.001 83.398 83.400 scf_env_do_scf 11 5.9 0.001 0.001 73.283 73.283 velocity_verlet 10 3.0 0.005 0.006 56.547 56.580 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 44.722 44.724 dbcsr_multiply_generic 2055 12.4 0.125 0.134 29.950 30.095 init_scf_loop 11 6.9 0.000 0.000 28.483 28.486 qs_scf_new_mos 99 7.5 0.001 0.001 27.325 27.403 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.325 27.403 prepare_preconditioner 11 7.9 0.000 0.000 26.464 26.477 make_preconditioner 11 8.9 0.000 0.000 26.464 26.477 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.788 25.934 ot_scf_mini 99 9.5 0.002 0.002 25.513 25.587 multiply_cannon 2055 13.4 0.332 0.352 22.618 23.303 multiply_cannon_loop 2055 14.4 0.819 0.832 20.783 21.135 cp_fm_upper_to_full 70 14.2 12.621 18.015 12.621 18.015 ot_mini 99 10.5 0.001 0.001 14.449 14.516 rebuild_ks_matrix 110 8.3 0.000 0.001 13.684 13.772 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 13.683 13.772 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.410 12.488 dbcsr_complete_redistribute 325 12.2 1.017 1.069 7.297 10.397 multiply_cannon_multrec 8220 15.4 4.060 4.228 9.741 9.849 qs_ot_get_derivative 99 11.5 0.001 0.001 9.667 9.741 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.233 9.334 mp_waitall_1 84994 16.7 7.786 8.837 7.786 8.837 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.661 8.732 mp_alltoall_i22 605 13.7 5.309 8.448 5.309 8.448 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.150 7.177 calculate_rho_elec 110 8.6 0.225 0.225 7.150 7.176 sum_up_and_integrate 110 10.3 0.002 0.002 6.797 6.812 integrate_v_rspace 110 11.3 0.004 0.004 6.769 6.784 make_m2s 4110 13.4 0.043 0.043 5.478 6.006 init_scf_run 11 5.9 0.000 0.001 5.820 5.820 scf_env_initial_rho_setup 11 6.9 0.010 0.013 5.820 5.820 make_images 4110 14.4 0.886 0.941 5.290 5.818 qs_ot_get_p 110 10.4 0.001 0.001 5.716 5.806 dbcsr_mm_accdrv_process 11614 15.7 3.858 4.056 5.534 5.750 cp_fm_cholesky_invert 11 10.9 5.538 5.544 5.538 5.544 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.888 5.399 apply_single 110 13.6 0.000 0.000 4.888 5.398 multiply_cannon_metrocomm3 8220 15.4 0.019 0.020 5.021 5.340 pw_transfer 1331 11.6 0.075 0.075 4.963 4.967 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 4.846 4.849 ot_diis_step 99 11.5 0.015 0.016 4.763 4.763 fft_wrap_pw1pw2_140 451 13.1 1.278 1.281 4.226 4.231 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.060 4.066 grid_integrate_task_list 110 12.3 3.687 3.743 3.687 3.743 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.675 3.676 hybrid_alltoall_any 4261 16.3 0.263 0.564 2.898 3.654 density_rs2pw 110 9.6 0.004 0.004 3.581 3.603 make_images_data 4110 15.4 0.045 0.049 2.904 3.567 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.077 3.555 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.544 3.544 wfi_extrapolate 11 7.9 0.001 0.001 3.397 3.398 calculate_dm_sparse 110 9.5 0.001 0.001 3.179 3.208 multiply_cannon_sync_h2d 8220 15.4 3.135 3.157 3.135 3.157 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.980 2.982 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.947 2.948 cp_fm_diag_elpa_base 48 14.0 2.415 2.609 2.946 2.946 fft3d_ps 1111 14.6 1.296 1.303 2.901 2.907 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.700 2.737 grid_collocate_task_list 110 9.6 2.678 2.712 2.678 2.712 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.437 2.469 calculate_first_density_matrix 1 7.0 0.000 0.000 2.295 2.298 potential_pw2rs 110 12.3 0.022 0.022 2.279 2.285 qs_env_update_s_mstruct 11 6.9 0.001 0.001 2.193 2.257 mp_alltoall_d11v 2046 13.8 2.145 2.205 2.145 2.205 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.102 2.177 cp_fm_cholesky_decompose 22 10.9 2.075 2.097 2.075 2.097 qs_create_task_list 11 7.9 0.001 0.001 1.881 1.926 generate_qs_task_list 11 8.9 0.730 0.783 1.880 1.926 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.837 1.895 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.875 1.885 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=88.264000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1251.545455, yerr=49.670484 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/164be73c51fd9cb7dc5314b3417fa4c35b0a34c7_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420243808256 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 11528908111872 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 93.514772E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755942624 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 631.410688E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009031E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175955597216 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4020 57505. MP_Allreduce 11129 795. MP_Sync 87 MP_Alltoall 2226 2125975. MP_SendRecv 24320 18752. MP_ISendRecv 24320 18752. MP_Wait 42476 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.089 0.165 210.046 210.049 qs_mol_dyn_low 1 2.0 0.003 0.004 208.977 209.036 qs_forces 11 3.9 0.005 0.005 208.882 208.883 qs_energies 11 4.9 0.003 0.012 203.164 203.181 scf_env_do_scf 11 5.9 0.001 0.003 186.563 186.567 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 165.971 165.973 dbcsr_multiply_generic 2507 12.6 0.181 0.189 126.348 127.247 qs_scf_new_mos 117 7.6 0.001 0.001 126.082 126.304 qs_scf_loop_do_ot 117 8.6 0.001 0.001 126.082 126.303 velocity_verlet 10 3.0 0.001 0.002 124.914 124.916 ot_scf_mini 117 9.6 0.003 0.004 119.329 119.527 multiply_cannon 2507 13.6 0.239 0.248 101.818 103.760 multiply_cannon_loop 2507 14.6 2.391 2.445 99.624 101.473 ot_mini 117 10.6 0.001 0.001 66.641 66.892 multiply_cannon_multrec 60168 15.6 31.810 33.628 41.547 43.374 qs_ot_get_derivative 117 11.6 0.001 0.001 41.791 42.000 rebuild_ks_matrix 128 8.3 0.001 0.001 33.831 34.149 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.018 33.830 34.148 mp_waitall_1 267128 16.5 29.341 32.484 29.341 32.484 qs_ot_get_p 128 10.4 0.001 0.002 30.422 30.771 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.298 30.554 multiply_cannon_sync_h2d 60168 15.6 26.341 27.949 26.341 27.949 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.381 25.048 apply_single 128 13.6 0.001 0.001 24.381 25.047 ot_diis_step 117 11.6 0.008 0.008 24.626 24.627 qs_ot_p2m_diag 83 11.4 0.079 0.091 23.759 23.835 cp_dbcsr_syevd 83 12.4 0.005 0.005 20.754 20.755 init_scf_loop 11 6.9 0.001 0.007 20.502 20.505 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 20.110 20.269 multiply_cannon_metrocomm3 60168 15.6 0.119 0.124 16.442 18.513 cp_fm_diag_elpa 83 13.4 0.000 0.000 17.675 17.703 cp_fm_diag_elpa_base 83 14.4 17.603 17.634 17.671 17.699 prepare_preconditioner 11 7.9 0.000 0.000 15.888 15.925 make_preconditioner 11 8.9 0.000 0.002 15.888 15.924 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.138 15.318 make_m2s 5014 13.6 0.105 0.114 14.031 14.369 sum_up_and_integrate 128 10.3 0.002 0.004 14.237 14.265 integrate_v_rspace 128 11.3 0.004 0.004 14.178 14.207 make_images 5014 14.6 0.401 0.431 13.848 14.194 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.479 13.709 calculate_rho_elec 128 8.7 0.045 0.064 13.478 13.708 init_scf_run 11 5.9 0.000 0.001 12.386 12.387 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.386 12.387 mp_sum_l 7950 12.9 9.256 10.653 9.256 10.653 dbcsr_mm_accdrv_process 124484 16.2 4.693 4.821 9.315 9.858 wfi_extrapolate 11 7.9 0.001 0.001 9.138 9.138 cp_fm_cholesky_invert 11 10.9 9.035 9.042 9.035 9.042 calculate_dm_sparse 128 9.5 0.001 0.001 8.559 8.643 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.150 8.265 qs_ot_get_orbitals 117 10.6 0.001 0.001 8.127 8.227 pw_transfer 1547 11.6 0.074 0.086 7.813 8.024 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 7.611 7.820 make_images_data 5014 15.6 0.066 0.072 6.839 7.800 multiply_cannon_metrocomm1 60168 15.6 0.094 0.098 6.222 7.787 density_rs2pw 128 9.7 0.006 0.007 7.036 7.703 grid_integrate_task_list 128 12.3 7.058 7.482 7.058 7.482 hybrid_alltoall_any 5200 16.5 0.296 2.274 5.987 7.227 fft_wrap_pw1pw2_140 523 13.2 1.125 1.176 6.616 6.802 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.669 6.679 fft3d_ps 1291 14.7 2.194 2.851 5.616 5.981 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 5.790 5.886 mp_alltoall_d11v 2415 14.1 4.434 5.681 4.434 5.681 grid_collocate_task_list 128 9.7 4.842 5.192 4.842 5.192 potential_pw2rs 128 12.3 0.009 0.010 4.739 4.775 cp_fm_cholesky_decompose 22 10.9 4.687 4.699 4.687 4.699 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=210.049000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=597.909091, yerr=6.921042 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/164be73c51fd9cb7dc5314b3417fa4c35b0a34c7_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 2.199914E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 6022464 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.2 marketing flops 145.647559E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 839.520256E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2425920 MPI messages size (bytes): total size 4.132350E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703416E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 71436 2336489472 32768 < size <= 131072 728832 55956209664 131072 < size <= 4194304 1386864 1409906900992 4194304 < size <= 16777216 155760 1473826772352 16777216 < size 68112 1190343475200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4024 57903. MP_Allreduce 11138 958. MP_Sync 87 MP_Alltoall 1983 6335681. MP_SendRecv 12126 47072. MP_ISendRecv 12126 47072. MP_Wait 26114 MP_ISend 11836 212447. MP_IRecv 11836 212447. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.035 0.145 193.098 193.100 qs_mol_dyn_low 1 2.0 0.003 0.004 192.597 192.620 qs_forces 11 3.9 0.004 0.005 192.513 192.517 qs_energies 11 4.9 0.005 0.030 185.717 185.731 scf_env_do_scf 11 5.9 0.001 0.005 169.235 169.245 scf_env_do_scf_inner_loop 118 6.6 0.003 0.010 135.991 135.994 velocity_verlet 10 3.0 0.004 0.005 122.265 122.266 dbcsr_multiply_generic 2527 12.6 0.191 0.198 98.634 100.101 qs_scf_new_mos 118 7.6 0.001 0.001 96.755 97.279 qs_scf_loop_do_ot 118 8.6 0.001 0.001 96.754 97.278 ot_scf_mini 118 9.6 0.004 0.005 91.880 92.470 multiply_cannon 2527 13.6 0.507 0.568 77.815 82.159 multiply_cannon_loop 2527 14.6 1.584 1.655 74.404 77.764 ot_mini 118 10.6 0.001 0.001 51.175 51.693 mp_waitall_1 216438 16.6 24.855 38.746 24.855 38.746 multiply_cannon_multrec 30324 15.6 20.968 26.225 31.596 37.339 rebuild_ks_matrix 129 8.3 0.001 0.001 32.736 33.465 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.021 32.736 33.464 init_scf_loop 11 6.9 0.001 0.003 33.148 33.149 qs_ks_update_qs_env 129 7.6 0.001 0.001 29.449 30.103 qs_ot_get_derivative 118 11.6 0.001 0.002 28.784 29.365 multiply_cannon_metrocomm3 30324 15.6 0.098 0.103 16.117 29.289 prepare_preconditioner 11 7.9 0.000 0.000 28.797 28.877 make_preconditioner 11 8.9 0.000 0.001 28.797 28.877 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.488 28.033 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 22.461 23.704 apply_single 129 13.6 0.001 0.001 22.460 23.704 qs_ot_get_p 129 10.4 0.001 0.001 22.488 23.191 ot_diis_step 118 11.6 0.014 0.015 22.205 22.207 multiply_cannon_sync_h2d 30324 15.6 17.946 20.274 17.946 20.274 qs_ot_p2m_diag 83 11.4 0.189 0.217 17.456 17.497 cp_fm_cholesky_invert 11 10.9 16.779 16.792 16.779 16.792 make_m2s 5054 13.6 0.090 0.096 14.616 16.211 cp_dbcsr_syevd 83 12.4 0.005 0.006 16.209 16.210 make_images 5054 14.6 1.188 1.385 14.402 15.997 sum_up_and_integrate 129 10.3 0.002 0.004 14.277 14.302 integrate_v_rspace 129 11.3 0.003 0.004 14.216 14.245 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.677 13.717 calculate_rho_elec 129 8.7 0.088 0.104 13.676 13.717 cp_fm_diag_elpa 83 13.4 0.000 0.000 12.897 12.925 cp_fm_diag_elpa_base 83 14.4 12.638 12.748 12.891 12.916 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 11.390 11.841 init_scf_run 11 5.9 0.000 0.001 11.572 11.574 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.572 11.574 multiply_cannon_metrocomm4 27797 15.6 0.106 0.120 3.924 11.328 mp_irecv_dv 70031 16.3 3.716 10.922 3.716 10.922 dbcsr_mm_accdrv_process 62734 16.2 5.500 6.065 10.086 10.668 make_images_data 5054 15.6 0.067 0.077 8.609 10.528 hybrid_alltoall_any 5240 16.5 0.354 1.536 7.350 10.016 pw_transfer 1559 11.6 0.086 0.096 8.799 8.849 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 8.574 8.628 wfi_extrapolate 11 7.9 0.001 0.001 8.454 8.454 density_rs2pw 129 9.7 0.006 0.007 7.301 7.708 fft_wrap_pw1pw2_140 527 13.2 1.223 1.244 7.507 7.571 grid_integrate_task_list 129 12.3 7.238 7.541 7.238 7.541 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 6.454 7.277 cp_fm_cholesky_decompose 22 10.9 7.095 7.168 7.095 7.168 calculate_dm_sparse 129 9.5 0.001 0.001 6.537 6.693 mp_sum_l 8010 12.9 4.450 6.660 4.450 6.660 fft3d_ps 1301 14.7 2.862 3.027 6.115 6.162 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.100 6.109 qs_ot_get_orbitals 118 10.6 0.001 0.001 5.494 5.584 grid_collocate_task_list 129 9.7 5.041 5.399 5.041 5.399 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.171 5.379 mp_allgather_i34 2527 14.6 1.918 5.168 1.918 5.168 mp_alltoall_d11v 2423 14.1 4.291 4.891 4.291 4.891 potential_pw2rs 129 12.3 0.015 0.018 4.681 4.699 mp_sum_d 4499 12.1 2.722 4.202 2.722 4.202 dbcsr_complete_redistribute 395 12.7 0.775 0.872 3.241 4.063 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=193.100000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=799.727273, yerr=4.024717 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/164be73c51fd9cb7dc5314b3417fa4c35b0a34c7_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 981.630976E+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 3992 58351. MP_Allreduce 11057 1000. MP_Sync 87 MP_Alltoall 1712 9388896. MP_SendRecv 7936 75008. MP_ISendRecv 7936 75008. MP_Wait 21820 MP_ISend 11748 275205. MP_IRecv 11748 275205. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.029 178.106 178.107 qs_mol_dyn_low 1 2.0 0.003 0.003 177.485 177.502 qs_forces 11 3.9 0.004 0.004 177.397 177.400 qs_energies 11 4.9 0.001 0.002 170.835 170.844 scf_env_do_scf 11 5.9 0.001 0.001 155.155 155.156 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 119.507 119.508 velocity_verlet 10 3.0 0.003 0.004 114.893 114.895 dbcsr_multiply_generic 2507 12.6 0.180 0.187 82.735 83.779 qs_scf_new_mos 117 7.6 0.001 0.001 82.089 82.386 qs_scf_loop_do_ot 117 8.6 0.001 0.001 82.088 82.385 ot_scf_mini 117 9.6 0.003 0.004 77.795 78.126 multiply_cannon 2507 13.6 0.507 0.525 62.530 67.153 multiply_cannon_loop 2507 14.6 1.131 1.184 59.408 61.905 ot_mini 117 10.6 0.001 0.001 43.151 43.506 init_scf_loop 11 6.9 0.000 0.000 35.542 35.543 mp_waitall_1 170520 16.6 25.487 35.273 25.487 35.273 prepare_preconditioner 11 7.9 0.000 0.000 31.527 31.570 make_preconditioner 11 8.9 0.000 0.000 31.527 31.570 rebuild_ks_matrix 128 8.3 0.001 0.001 30.486 30.930 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.017 30.485 30.930 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.113 30.506 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.469 27.869 multiply_cannon_multrec 20056 15.6 12.995 16.075 22.329 25.408 multiply_cannon_metrocomm3 20056 15.6 0.063 0.066 15.744 25.241 qs_ot_get_derivative 117 11.6 0.002 0.002 23.268 23.592 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.994 21.080 apply_single 128 13.6 0.001 0.001 19.994 21.080 qs_ot_get_p 128 10.4 0.001 0.001 19.835 20.267 ot_diis_step 117 11.6 0.018 0.018 19.768 19.769 make_m2s 5014 13.6 0.079 0.083 15.047 16.278 make_images 5014 14.6 1.159 1.246 14.815 16.036 qs_ot_p2m_diag 83 11.4 0.265 0.272 15.554 15.565 multiply_cannon_sync_h2d 20056 15.6 13.489 14.887 13.489 14.887 cp_fm_cholesky_invert 11 10.9 14.759 14.768 14.759 14.768 cp_dbcsr_syevd 83 12.4 0.005 0.006 14.477 14.478 sum_up_and_integrate 128 10.3 0.002 0.002 14.145 14.169 integrate_v_rspace 128 11.3 0.003 0.004 14.085 14.110 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.683 13.716 calculate_rho_elec 128 8.7 0.130 0.145 13.682 13.715 cp_fm_diag_elpa 83 13.4 0.000 0.000 11.204 11.220 cp_fm_diag_elpa_base 83 14.4 10.786 10.953 11.200 11.217 make_images_data 5014 15.6 0.063 0.072 9.077 10.681 init_scf_run 11 5.9 0.000 0.001 10.576 10.576 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.576 10.576 hybrid_alltoall_any 5200 16.5 0.450 2.046 7.854 10.094 multiply_cannon_metrocomm4 17549 15.6 0.067 0.077 3.518 9.566 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.250 9.491 mp_irecv_dv 50230 16.2 3.390 9.306 3.390 9.306 dbcsr_mm_accdrv_process 41502 16.2 5.619 5.987 8.806 9.012 pw_transfer 1547 11.6 0.085 0.101 8.789 8.893 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.565 8.675 grid_integrate_task_list 128 12.3 7.288 7.747 7.288 7.747 fft_wrap_pw1pw2_140 523 13.2 1.305 1.340 7.554 7.672 wfi_extrapolate 11 7.9 0.001 0.001 7.548 7.548 cp_fm_cholesky_decompose 22 10.9 7.442 7.516 7.442 7.516 density_rs2pw 128 9.7 0.006 0.006 7.015 7.450 cp_fm_upper_to_full 105 14.8 5.833 7.369 5.833 7.369 dbcsr_complete_redistribute 395 12.7 1.182 1.251 4.766 6.586 fft3d_ps 1291 14.7 2.764 2.997 5.938 6.021 calculate_dm_sparse 128 9.5 0.001 0.001 5.894 5.994 grid_collocate_task_list 128 9.7 5.167 5.578 5.167 5.578 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.533 5.537 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.566 5.381 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.575 5.275 mp_alltoall_d11v 2415 14.1 4.481 5.028 4.481 5.028 mp_allgather_i34 2507 14.6 1.613 4.897 1.613 4.897 mp_sum_l 7950 12.9 3.326 4.763 3.326 4.763 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.581 4.704 potential_pw2rs 128 12.3 0.021 0.023 4.558 4.566 transfer_fm_to_dbcsr 11 9.9 0.019 0.023 2.395 4.189 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.010 4.040 mp_alltoall_i22 716 14.1 1.952 3.943 1.952 3.943 qs_energies_init_hamiltonians 11 5.9 0.012 0.014 3.766 3.769 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.517 3.562 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=178.107000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=921.181818, yerr=14.552839 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/164be73c51fd9cb7dc5314b3417fa4c35b0a34c7_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 4.353788E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 5977344 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.168191E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1143192 MPI messages size (bytes): total size 2.023815E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.770319E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 319024 36042702848 131072 < size <= 4194304 715736 785529176064 4194304 < size <= 16777216 70320 665379241840 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58348. MP_Allreduce 11057 1083. MP_Sync 87 MP_Alltoall 1712 12503107. MP_SendRecv 5888 75008. MP_ISendRecv 5888 75008. MP_Wait 22442 MP_ISend 14952 244818. MP_IRecv 14952 244818. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.027 202.409 202.410 qs_mol_dyn_low 1 2.0 0.003 0.003 202.030 202.043 qs_forces 11 3.9 0.004 0.005 201.923 201.928 qs_energies 11 4.9 0.002 0.002 194.678 194.690 scf_env_do_scf 11 5.9 0.001 0.001 177.413 177.427 velocity_verlet 10 3.0 0.002 0.002 133.686 133.689 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 126.412 126.413 qs_scf_new_mos 117 7.6 0.001 0.001 88.078 88.410 qs_scf_loop_do_ot 117 8.6 0.001 0.001 88.077 88.409 ot_scf_mini 117 9.6 0.003 0.003 83.453 83.776 dbcsr_multiply_generic 2507 12.6 0.188 0.195 82.535 83.199 multiply_cannon 2507 13.6 0.545 0.585 56.017 58.991 multiply_cannon_loop 2507 14.6 1.812 1.907 52.102 53.862 init_scf_loop 11 6.9 0.000 0.000 50.867 50.868 prepare_preconditioner 11 7.9 0.000 0.000 46.700 46.731 make_preconditioner 11 8.9 0.000 0.000 46.700 46.731 make_full_inverse_cholesky 11 9.9 0.011 0.023 40.267 45.283 ot_mini 117 10.6 0.001 0.001 44.520 44.859 multiply_cannon_multrec 30084 15.6 13.723 18.863 26.590 31.435 rebuild_ks_matrix 128 8.3 0.001 0.001 30.514 30.844 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.019 30.513 30.843 mp_waitall_1 147882 16.7 18.740 28.633 18.740 28.633 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.534 27.821 qs_ot_get_derivative 117 11.6 0.002 0.002 24.821 25.160 qs_ot_get_p 128 10.4 0.001 0.001 24.720 25.077 make_m2s 5014 13.6 0.095 0.098 21.851 23.001 make_images 5014 14.6 1.965 2.295 21.551 22.700 qs_ot_p2m_diag 83 11.4 0.343 0.390 20.175 20.229 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 19.099 19.673 apply_single 128 13.6 0.001 0.001 19.099 19.672 ot_diis_step 117 11.6 0.017 0.018 19.572 19.574 cp_fm_cholesky_invert 11 10.9 18.881 18.891 18.881 18.891 cp_dbcsr_syevd 83 12.4 0.005 0.006 18.625 18.628 cp_fm_upper_to_full 105 14.8 11.614 17.055 11.614 17.055 sum_up_and_integrate 128 10.3 0.002 0.003 15.020 15.040 integrate_v_rspace 128 11.3 0.003 0.004 14.960 14.981 cp_fm_diag_elpa 83 13.4 0.000 0.000 14.883 14.900 cp_fm_diag_elpa_base 83 14.4 13.845 14.176 14.876 14.892 multiply_cannon_metrocomm3 30084 15.6 0.050 0.053 6.244 14.860 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.822 14.853 calculate_rho_elec 128 8.7 0.173 0.188 14.822 14.852 make_images_data 5014 15.6 0.066 0.074 11.857 13.493 dbcsr_complete_redistribute 395 12.7 1.503 1.617 9.452 13.209 dbcsr_mm_accdrv_process 62264 16.2 8.353 9.201 12.441 12.979 hybrid_alltoall_any 5200 16.5 0.533 2.218 10.409 12.205 copy_fm_to_dbcsr 209 11.7 0.001 0.002 7.919 11.667 multiply_cannon_sync_h2d 30084 15.6 10.451 11.339 10.451 11.339 init_scf_run 11 5.9 0.000 0.001 11.219 11.221 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.219 11.220 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 10.577 10.819 transfer_fm_to_dbcsr 11 9.9 0.001 0.011 6.411 10.083 pw_transfer 1547 11.6 0.086 0.101 9.586 9.644 mp_alltoall_i22 716 14.1 5.644 9.522 5.644 9.522 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 9.360 9.424 cp_fm_cholesky_decompose 22 10.9 9.166 9.252 9.166 9.252 fft_wrap_pw1pw2_140 523 13.2 1.446 1.476 8.084 8.158 wfi_extrapolate 11 7.9 0.001 0.001 8.142 8.142 density_rs2pw 128 9.7 0.006 0.006 7.740 7.993 grid_integrate_task_list 128 12.3 7.515 7.857 7.515 7.857 multiply_cannon_metrocomm4 25070 15.6 0.085 0.094 2.780 7.288 mp_irecv_dv 76098 16.2 2.626 7.016 2.626 7.016 fft3d_ps 1291 14.7 2.848 2.900 6.539 6.591 calculate_dm_sparse 128 9.5 0.001 0.001 6.314 6.400 mp_alltoall_d11v 2415 14.1 5.513 5.936 5.513 5.936 grid_collocate_task_list 128 9.7 5.287 5.672 5.287 5.672 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.600 5.648 potential_pw2rs 128 12.3 0.024 0.024 5.177 5.214 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.745 4.848 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.458 4.532 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.418 4.511 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 4.430 4.432 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=202.410000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1100.545455, yerr=24.440516 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/164be73c51fd9cb7dc5314b3417fa4c35b0a34c7_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410024443904 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 11444712984576 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 92.796586E+12 0.0% 0.0% 100.0% flops max/rank 5.820059E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705502176 0.0% 0.0% 100.0% number of processed stacks 1944496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3448.5 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.560752E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 238560 MPI messages size (bytes): total size 1.321104E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.537828E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 112800 59139686400 4194304 < size <= 16777216 104112 545846722560 16777216 < size 20064 716108700816 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8852 52. MP_Alltoall 9584 804353. MP_ISend 39716 2104723. MP_IRecv 39716 2103824. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4003 58187. MP_Allreduce 11085 1167. MP_Sync 86 MP_Alltoall 1700 18828148. MP_SendRecv 3810 122880. MP_ISendRecv 3810 122880. MP_Wait 16000 MP_ISend 10600 423612. MP_IRecv 10600 423612. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.042 0.075 183.702 183.723 qs_mol_dyn_low 1 2.0 0.003 0.003 182.949 182.965 qs_forces 11 3.9 0.004 0.005 182.846 182.849 qs_energies 11 4.9 0.001 0.002 175.305 175.310 scf_env_do_scf 11 5.9 0.001 0.001 157.337 157.346 velocity_verlet 10 3.0 0.004 0.005 118.655 118.659 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 117.449 117.450 qs_scf_new_mos 116 7.6 0.001 0.001 79.042 79.123 qs_scf_loop_do_ot 116 8.6 0.001 0.001 79.041 79.123 ot_scf_mini 116 9.6 0.003 0.004 74.571 74.634 dbcsr_multiply_generic 2485 12.5 0.181 0.190 73.427 73.763 multiply_cannon 2485 13.5 0.553 0.578 53.611 57.269 multiply_cannon_loop 2485 14.5 0.801 0.835 50.464 51.158 init_scf_loop 11 6.9 0.000 0.000 39.728 39.730 ot_mini 116 10.6 0.001 0.001 38.997 39.051 prepare_preconditioner 11 7.9 0.000 0.000 35.779 35.809 make_preconditioner 11 8.9 0.000 0.000 35.779 35.809 make_full_inverse_cholesky 11 9.9 0.013 0.024 33.565 33.907 mp_waitall_1 124680 16.7 25.915 31.766 25.915 31.766 rebuild_ks_matrix 127 8.3 0.001 0.001 29.908 29.968 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.018 29.907 29.968 qs_ks_update_qs_env 127 7.6 0.001 0.001 27.219 27.279 qs_ot_get_p 127 10.4 0.001 0.001 22.801 22.902 qs_ot_get_derivative 116 11.6 0.001 0.002 22.211 22.273 multiply_cannon_multrec 9940 15.5 10.358 14.212 17.921 21.835 cp_fm_cholesky_invert 11 10.9 20.379 20.388 20.379 20.388 multiply_cannon_metrocomm3 9940 15.5 0.025 0.027 12.037 18.806 qs_ot_p2m_diag 82 11.4 0.490 0.496 18.560 18.582 cp_dbcsr_syevd 82 12.4 0.005 0.006 17.194 17.196 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 16.868 17.039 apply_single 127 13.6 0.001 0.001 16.868 17.039 ot_diis_step 116 11.6 0.020 0.021 16.714 16.714 make_m2s 4970 13.5 0.065 0.071 15.713 16.554 make_images 4970 14.5 2.162 2.587 15.410 16.250 qs_rho_update_rho_low 127 7.7 0.001 0.001 15.113 15.141 calculate_rho_elec 127 8.7 0.254 0.265 15.112 15.140 sum_up_and_integrate 127 10.3 0.002 0.002 15.077 15.121 integrate_v_rspace 127 11.3 0.004 0.004 15.017 15.063 cp_fm_diag_elpa 82 13.4 0.000 0.000 13.831 13.841 cp_fm_diag_elpa_base 82 14.4 13.535 13.670 13.826 13.837 multiply_cannon_sync_h2d 9940 15.5 10.737 11.274 10.737 11.274 init_scf_run 11 5.9 0.000 0.001 10.934 10.935 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.934 10.935 make_images_data 4970 15.5 0.055 0.065 8.883 10.408 pw_transfer 1535 11.6 0.085 0.093 10.062 10.095 hybrid_alltoall_any 5155 16.4 0.837 3.758 8.558 9.915 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 9.841 9.879 cp_fm_cholesky_decompose 22 10.9 9.573 9.694 9.573 9.694 qs_ot_get_derivative_diag 76 12.4 0.002 0.003 8.880 8.922 fft_wrap_pw1pw2_140 519 13.2 1.788 1.828 8.460 8.498 grid_integrate_task_list 127 12.3 7.716 8.095 7.716 8.095 density_rs2pw 127 9.7 0.005 0.006 7.690 8.007 multiply_cannon_metrocomm1 9940 15.5 0.030 0.030 5.233 7.948 dbcsr_mm_accdrv_process 20590 16.1 3.137 4.218 7.222 7.882 wfi_extrapolate 11 7.9 0.001 0.001 7.867 7.867 fft3d_ps 1281 14.7 2.748 2.820 6.599 6.641 calculate_dm_sparse 127 9.5 0.001 0.001 6.146 6.208 mp_alltoall_d11v 2401 14.1 5.391 6.129 5.391 6.129 dbcsr_complete_redistribute 393 12.7 2.094 2.176 5.686 6.109 grid_collocate_task_list 127 9.7 5.524 5.796 5.524 5.796 qs_energies_init_hamiltonians 11 5.9 0.004 0.022 5.330 5.331 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.243 5.255 potential_pw2rs 127 12.3 0.027 0.028 5.081 5.106 multiply_cannon_metrocomm4 7455 15.5 0.026 0.029 1.783 4.998 mp_irecv_dv 28618 15.9 1.745 4.913 1.745 4.913 mp_allgather_i34 2485 14.5 1.346 4.618 1.346 4.618 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.034 4.084 copy_fm_to_dbcsr 208 11.6 0.001 0.002 3.645 4.022 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.901 3.987 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.595 3.891 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.747 3.772 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.688 3.762 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=183.723000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1483.090909, yerr=18.686186 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/164be73c51fd9cb7dc5314b3417fa4c35b0a34c7_performance_tests/20/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 11.696233E+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 1964048 0.0% 0.0% 100.0% average stack size 0.0 0.0 3439.8 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 3.116380E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 100280 MPI messages size (bytes): total size 1.136195E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.330227E+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 45208 35089547264 4194304 < size <= 16777216 44352 379752284160 16777216 < size 10104 721350092304 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 59279. MP_Allreduce 11057 1504. MP_Sync 87 MP_Alltoall 1712 36974159. MP_SendRecv 1792 218624. MP_ISendRecv 1792 218624. MP_Wait 9802 MP_ISend 6408 1080322. MP_IRecv 6408 1080322. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.020 0.053 315.876 315.885 qs_mol_dyn_low 1 2.0 0.003 0.003 315.314 315.331 qs_forces 11 3.9 0.004 0.004 315.205 315.208 qs_energies 11 4.9 0.002 0.002 306.007 306.012 scf_env_do_scf 11 5.9 0.001 0.001 282.630 282.633 velocity_verlet 10 3.0 0.002 0.002 227.184 227.194 scf_env_do_scf_inner_loop 117 6.6 0.004 0.009 148.385 148.387 init_scf_loop 11 6.9 0.000 0.000 133.945 133.947 prepare_preconditioner 11 7.9 0.000 0.000 128.918 128.945 make_preconditioner 11 8.9 0.000 0.000 128.918 128.945 make_full_inverse_cholesky 11 9.9 0.038 0.039 104.266 126.017 qs_scf_new_mos 117 7.6 0.001 0.001 101.192 101.312 qs_scf_loop_do_ot 117 8.6 0.001 0.001 101.191 101.311 ot_scf_mini 117 9.6 0.003 0.004 96.172 96.252 dbcsr_multiply_generic 2507 12.6 0.212 0.223 85.991 86.271 cp_fm_upper_to_full 105 14.8 53.032 76.761 53.032 76.761 multiply_cannon 2507 13.6 0.683 0.728 60.086 63.203 multiply_cannon_loop 2507 14.6 1.053 1.078 55.758 56.312 ot_mini 117 10.6 0.001 0.001 47.730 47.880 dbcsr_complete_redistribute 395 12.7 3.984 4.015 31.044 44.520 copy_fm_to_dbcsr 209 11.7 0.001 0.001 27.278 40.744 cp_fm_cholesky_invert 11 10.9 39.463 39.470 39.463 39.470 transfer_fm_to_dbcsr 11 9.9 0.030 0.030 24.609 37.900 rebuild_ks_matrix 128 8.3 0.001 0.001 36.309 36.493 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.017 36.309 36.493 mp_alltoall_i22 716 14.1 22.394 35.962 22.394 35.962 mp_waitall_1 103674 16.8 30.701 34.128 30.701 34.128 qs_ks_update_qs_env 128 7.6 0.001 0.001 33.557 33.719 qs_ot_get_p 128 10.4 0.001 0.001 32.909 33.129 qs_ot_p2m_diag 83 11.4 0.880 0.885 28.131 28.140 qs_ot_get_derivative 117 11.6 0.002 0.002 27.048 27.121 cp_dbcsr_syevd 83 12.4 0.006 0.006 26.177 26.180 cp_fm_diag_elpa 83 13.4 0.000 0.000 22.295 22.297 cp_fm_diag_elpa_base 83 14.4 17.705 19.451 22.290 22.290 make_m2s 5014 13.6 0.076 0.079 20.523 22.162 make_images 5014 14.6 3.077 3.256 20.042 21.677 ot_diis_step 117 11.6 0.022 0.022 20.569 20.569 multiply_cannon_metrocomm3 10028 15.6 0.025 0.027 19.565 20.195 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.602 20.038 apply_single 128 13.6 0.001 0.001 19.602 20.038 qs_rho_update_rho_low 128 7.7 0.001 0.001 18.495 18.505 calculate_rho_elec 128 8.7 0.479 0.480 18.494 18.504 multiply_cannon_multrec 10028 15.6 10.405 12.230 18.189 18.260 sum_up_and_integrate 128 10.3 0.002 0.002 17.773 17.860 integrate_v_rspace 128 11.3 0.004 0.004 17.711 17.798 make_images_data 5014 15.6 0.064 0.070 11.891 14.562 hybrid_alltoall_any 5200 16.5 1.310 3.048 11.869 14.532 multiply_cannon_sync_h2d 10028 15.6 14.301 14.356 14.301 14.356 pw_transfer 1547 11.6 0.096 0.097 13.232 13.256 fft_wrap_pw1pw2 1291 12.7 0.012 0.012 12.993 13.014 init_scf_run 11 5.9 0.000 0.001 12.926 12.926 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.925 12.926 cp_fm_cholesky_decompose 22 10.9 11.657 11.680 11.657 11.680 fft_wrap_pw1pw2_140 523 13.2 3.037 3.065 11.402 11.430 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 10.771 10.816 wfi_extrapolate 11 7.9 0.001 0.001 9.769 9.769 dbcsr_mm_accdrv_process 20762 16.1 4.345 6.166 7.535 9.435 density_rs2pw 128 9.7 0.005 0.005 9.009 9.057 grid_integrate_task_list 128 12.3 8.573 8.767 8.573 8.767 mp_alltoall_d11v 2415 14.1 8.568 8.736 8.568 8.736 fft3d_ps 1291 14.7 2.859 2.873 8.248 8.303 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 8.192 8.193 calculate_dm_sparse 128 9.5 0.001 0.001 6.900 6.953 copy_dbcsr_to_fm 186 11.8 0.004 0.004 6.652 6.751 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.621 6.656 grid_collocate_task_list 128 9.7 6.430 6.502 6.430 6.502 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=315.885000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2849.090909, yerr=134.689174 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/164be73c51fd9cb7dc5314b3417fa4c35b0a34c7_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 1.260003E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255646. MP_Allreduce 3139 6114. MP_Sync 4 MP_Alltoall 54 6098861. MP_SendRecv 285 19200. MP_ISendRecv 285 19200. MP_Wait 1017 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.088 0.131 85.681 85.683 qs_energies 1 2.0 0.000 0.000 84.700 84.714 ls_scf 1 3.0 0.000 0.000 83.792 83.806 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.543 72.765 multiply_cannon 111 7.7 0.018 0.021 55.894 57.097 multiply_cannon_loop 111 8.7 0.228 0.241 52.476 54.032 ls_scf_main 1 4.0 0.000 0.000 52.385 52.395 density_matrix_trs4 2 5.0 0.002 0.003 46.838 46.916 ls_scf_init_scf 1 4.0 0.000 0.000 28.348 28.349 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.206 27.266 mp_waitall_1 11031 10.9 22.095 25.553 22.095 25.553 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.119 25.141 multiply_cannon_multrec 2664 9.7 8.145 8.884 15.638 17.367 multiply_cannon_sync_h2d 2664 9.7 13.698 16.159 13.698 16.159 make_m2s 222 7.7 0.009 0.011 12.969 13.530 make_images 222 8.7 0.099 0.108 12.947 13.510 multiply_cannon_metrocomm1 2664 9.7 0.010 0.010 9.443 13.288 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.427 8.848 make_images_data 222 9.7 0.004 0.005 7.549 8.235 dbcsr_mm_accdrv_process 4760 10.4 0.596 0.730 7.108 8.066 hybrid_alltoall_any 227 10.6 0.217 1.851 6.435 7.565 dbcsr_mm_accdrv_process_sort 4760 11.4 6.313 7.208 6.313 7.208 calculate_norms 4752 9.8 5.540 6.253 5.540 6.253 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.981 5.183 mp_sum_l 887 5.1 3.160 4.639 3.160 4.639 multiply_cannon_metrocomm4 2442 9.7 0.012 0.016 2.072 4.056 mp_irecv_dv 6231 10.9 2.056 4.022 2.056 4.022 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.337 3.512 arnoldi_extremal 4 6.8 0.000 0.000 3.466 3.495 arnoldi_normal_ev 4 7.8 0.001 0.003 3.466 3.495 build_subspace 16 8.4 0.009 0.012 3.362 3.364 make_images_sizes 222 9.7 0.000 0.000 0.767 3.313 mp_alltoall_i44 222 10.7 0.767 3.312 0.767 3.312 ls_scf_post 1 4.0 0.000 0.000 3.059 3.073 ls_scf_store_result 1 5.0 0.000 0.000 2.883 2.920 dbcsr_special_finalize 555 9.7 0.005 0.006 2.302 2.700 dbcsr_merge_single_wm 555 10.7 0.453 0.594 2.293 2.691 make_images_pack 222 9.7 2.203 2.625 2.205 2.626 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.372 2.600 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.348 2.458 dbcsr_matrix_vector_mult_local 304 10.0 2.062 2.454 2.064 2.456 dbcsr_sort_data 658 11.4 2.098 2.408 2.098 2.408 buffer_matrices_ensure_size 222 8.7 1.749 2.121 1.749 2.121 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.784 1.785 rebuild_ks_matrix 3 7.3 0.000 0.000 1.775 1.776 qs_ks_build_kohn_sham_matrix 3 8.3 0.009 0.012 1.775 1.776 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.683000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1131.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/164be73c51fd9cb7dc5314b3417fa4c35b0a34c7_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 2.157711E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266673. MP_Allreduce 3138 10075. MP_Sync 4 MP_Alltoall 47 15335933. MP_SendRecv 141 57600. MP_ISendRecv 141 57600. MP_Wait 687 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.027 0.042 90.780 90.781 qs_energies 1 2.0 0.000 0.002 90.264 90.268 ls_scf 1 3.0 0.002 0.015 88.942 88.945 dbcsr_multiply_generic 111 6.7 0.015 0.016 74.661 74.968 multiply_cannon 111 7.7 0.027 0.040 53.009 58.062 ls_scf_main 1 4.0 0.001 0.021 54.989 54.995 multiply_cannon_loop 111 8.7 0.135 0.146 50.082 53.595 density_matrix_trs4 2 5.0 0.003 0.011 49.233 49.472 ls_scf_init_scf 1 4.0 0.000 0.002 30.291 30.293 mp_waitall_1 9105 10.9 20.645 29.944 20.645 29.944 ls_scf_init_matrix_S 1 5.0 0.000 0.001 28.959 29.072 multiply_cannon_multrec 1332 9.7 13.294 16.728 22.657 27.180 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.602 26.614 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.675 20.502 make_m2s 222 7.7 0.006 0.007 14.766 15.426 make_images 222 8.7 1.364 1.706 14.735 15.398 dbcsr_mm_accdrv_process 4041 10.4 0.354 0.527 8.959 10.499 dbcsr_mm_accdrv_process_sort 4041 11.4 8.478 9.972 8.478 9.972 make_images_data 222 9.7 0.004 0.005 8.447 9.368 hybrid_alltoall_any 227 10.6 0.543 2.595 7.759 9.192 mp_sum_l 887 5.1 5.430 8.736 5.430 8.736 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.206 7.571 mp_irecv_dv 3311 11.0 3.186 7.516 3.186 7.516 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.130 6.896 calculate_norms 2376 9.8 6.055 6.709 6.055 6.709 multiply_cannon_sync_h2d 1332 9.7 4.847 6.210 4.847 6.210 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.968 5.194 arnoldi_extremal 4 6.8 0.000 0.000 4.947 4.964 arnoldi_normal_ev 4 7.8 0.001 0.004 4.947 4.964 build_subspace 16 8.4 0.014 0.020 4.688 4.691 ls_scf_post 1 4.0 0.000 0.001 3.660 3.662 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.296 3.506 ls_scf_store_result 1 5.0 0.000 0.000 3.398 3.481 dbcsr_matrix_vector_mult_local 304 10.0 2.749 3.234 2.751 3.236 mp_allgather_i34 111 8.7 0.920 3.139 0.920 3.139 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.621 2.715 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.149 2.683 dbcsr_data_new 4174 10.1 2.110 2.411 2.110 2.411 make_images_pack 222 9.7 1.813 2.118 1.816 2.120 dbcsr_sort_data 436 11.2 1.830 2.120 1.830 2.120 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.984 1.986 rebuild_ks_matrix 3 7.3 0.000 0.000 1.971 1.973 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.971 1.973 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=90.781000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1774.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/164be73c51fd9cb7dc5314b3417fa4c35b0a34c7_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 2.895409E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265448. MP_Allreduce 3138 10896. MP_Sync 4 MP_Alltoall 47 23526250. MP_SendRecv 93 57600. MP_ISendRecv 93 57600. MP_Wait 639 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.142 0.222 91.525 91.525 qs_energies 1 2.0 0.000 0.000 90.647 90.653 ls_scf 1 3.0 0.000 0.000 89.235 89.241 dbcsr_multiply_generic 111 6.7 0.016 0.016 73.474 73.754 multiply_cannon 111 7.7 0.032 0.069 51.218 55.469 ls_scf_main 1 4.0 0.000 0.000 55.230 55.234 multiply_cannon_loop 111 8.7 0.117 0.128 48.539 52.128 density_matrix_trs4 2 5.0 0.002 0.003 49.370 49.498 mp_waitall_1 7281 11.0 21.783 31.584 21.783 31.584 ls_scf_init_scf 1 4.0 0.000 0.000 30.325 30.327 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.091 29.195 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.790 26.802 multiply_cannon_multrec 888 9.7 12.794 15.249 21.464 24.903 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.106 21.463 make_m2s 222 7.7 0.006 0.007 15.778 16.428 make_images 222 8.7 1.587 1.848 15.739 16.391 make_images_data 222 9.7 0.004 0.004 8.735 9.904 hybrid_alltoall_any 227 10.6 0.644 2.967 8.268 9.564 dbcsr_mm_accdrv_process 3754 10.4 0.317 0.490 8.186 9.457 dbcsr_mm_accdrv_process_sort 3754 11.4 7.752 8.967 7.752 8.967 mp_sum_l 887 5.1 4.775 8.088 4.775 8.088 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.584 7.232 multiply_cannon_sync_h2d 888 9.7 6.079 7.131 6.079 7.131 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.480 6.950 mp_irecv_dv 2335 11.1 2.465 6.910 2.465 6.910 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.644 6.468 arnoldi_extremal 4 6.8 0.000 0.000 5.558 5.572 arnoldi_normal_ev 4 7.8 0.001 0.005 5.558 5.572 build_subspace 16 8.4 0.014 0.020 5.253 5.258 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.672 4.888 calculate_norms 1584 9.8 4.364 4.778 4.364 4.778 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.677 3.960 ls_scf_post 1 4.0 0.000 0.000 3.680 3.688 dbcsr_matrix_vector_mult_local 304 10.0 3.034 3.627 3.036 3.629 ls_scf_store_result 1 5.0 0.000 0.000 3.406 3.498 mp_allgather_i34 111 8.7 0.796 3.322 0.796 3.322 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.842 2.949 dbcsr_data_new 4116 9.9 2.109 2.462 2.109 2.462 make_images_sizes 222 9.7 0.000 0.000 1.273 2.311 mp_alltoall_i44 222 10.7 1.272 2.311 1.272 2.311 dbcsr_sort_data 325 11.1 1.822 2.103 1.822 2.103 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.903 1.905 rebuild_ks_matrix 3 7.3 0.000 0.000 1.886 1.887 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.886 1.887 make_images_pack 222 9.7 1.618 1.878 1.621 1.881 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=91.525000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2257.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/164be73c51fd9cb7dc5314b3417fa4c35b0a34c7_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 3.407802E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266673. MP_Allreduce 3138 13030. MP_Sync 4 MP_Alltoall 47 30278988. MP_SendRecv 69 86400. MP_ISendRecv 69 86400. MP_Wait 531 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.028 0.048 96.762 96.763 qs_energies 1 2.0 0.000 0.000 96.172 96.177 ls_scf 1 3.0 0.000 0.000 94.522 94.526 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.029 78.255 ls_scf_main 1 4.0 0.000 0.000 58.630 58.631 multiply_cannon 111 7.7 0.041 0.108 51.551 56.028 density_matrix_trs4 2 5.0 0.002 0.003 52.433 52.560 multiply_cannon_loop 111 8.7 0.151 0.166 46.445 49.264 ls_scf_init_scf 1 4.0 0.000 0.000 32.653 32.656 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.441 31.504 mp_waitall_1 6369 11.0 22.176 29.264 22.176 29.264 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.938 28.950 multiply_cannon_multrec 1332 9.7 14.266 17.290 22.252 25.082 make_m2s 222 7.7 0.007 0.008 20.936 22.409 make_images 222 8.7 3.140 3.615 20.885 22.362 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 8.890 16.118 make_images_data 222 9.7 0.004 0.004 11.623 13.282 hybrid_alltoall_any 227 10.6 0.797 3.774 10.970 12.865 dbcsr_mm_accdrv_process 3641 10.4 0.277 0.467 7.621 9.177 dbcsr_mm_accdrv_process_sort 3641 11.4 7.183 8.682 7.183 8.682 mp_sum_l 887 5.1 3.944 7.013 3.944 7.013 multiply_cannon_sync_h2d 1332 9.7 5.451 6.223 5.451 6.223 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.089 6.189 mp_irecv_dv 3229 10.9 2.065 6.102 2.065 6.102 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.999 5.450 arnoldi_extremal 4 6.8 0.000 0.000 5.406 5.418 arnoldi_normal_ev 4 7.8 0.001 0.005 5.406 5.418 build_subspace 16 8.4 0.014 0.021 5.072 5.079 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.589 4.792 mp_allgather_i34 111 8.7 2.221 4.707 2.221 4.707 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.487 4.552 calculate_norms 2376 9.8 4.182 4.538 4.182 4.538 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.698 4.007 dbcsr_matrix_vector_mult_local 304 10.0 3.209 3.703 3.211 3.705 dbcsr_sort_data 658 11.4 3.070 3.473 3.070 3.473 dbcsr_special_finalize 555 9.7 0.006 0.007 2.811 3.255 dbcsr_merge_single_wm 555 10.7 0.540 0.672 2.802 3.247 ls_scf_post 1 4.0 0.000 0.000 3.239 3.244 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.038 3.105 ls_scf_store_result 1 5.0 0.000 0.000 2.986 3.043 dbcsr_data_release 10477 10.7 1.593 2.389 1.593 2.389 dbcsr_finalize 304 7.8 0.049 0.061 1.814 1.985 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.939 1.941 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=96.763000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2799.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/164be73c51fd9cb7dc5314b3417fa4c35b0a34c7_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 4.752175E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265536. MP_Allreduce 3129 15263. MP_Sync 4 MP_Alltoall 47 46208988. MP_SendRecv 45 115200. MP_ISendRecv 45 115200. MP_Wait 528 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.036 0.055 91.897 91.898 qs_energies 1 2.0 0.000 0.000 91.154 91.158 ls_scf 1 3.0 0.000 0.000 89.216 89.221 dbcsr_multiply_generic 111 6.7 0.017 0.018 70.784 70.947 ls_scf_main 1 4.0 0.000 0.000 56.283 56.283 multiply_cannon 111 7.7 0.083 0.194 52.693 55.675 multiply_cannon_loop 111 8.7 0.089 0.098 50.129 51.382 density_matrix_trs4 2 5.0 0.002 0.003 49.290 49.369 ls_scf_init_scf 1 4.0 0.000 0.000 29.539 29.541 mp_waitall_1 5436 11.0 24.328 28.865 24.328 28.865 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.289 28.325 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.188 26.207 multiply_cannon_multrec 444 9.7 13.696 16.362 20.859 22.088 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.736 15.487 make_m2s 222 7.7 0.005 0.005 13.365 14.289 make_images 222 8.7 2.040 2.481 13.297 14.220 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.265 13.896 make_images_data 222 9.7 0.003 0.004 8.142 9.638 hybrid_alltoall_any 227 10.6 0.799 3.807 8.101 9.419 multiply_cannon_sync_h2d 444 9.7 6.737 8.166 6.737 8.166 dbcsr_mm_accdrv_process 3003 10.4 0.357 0.429 6.861 7.967 dbcsr_mm_accdrv_process_sort 3003 11.4 6.491 7.568 6.491 7.568 arnoldi_extremal 4 6.8 0.000 0.000 5.856 5.875 arnoldi_normal_ev 4 7.8 0.002 0.005 5.856 5.875 build_subspace 16 8.4 0.015 0.020 5.457 5.466 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.449 4.687 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.212 4.394 mp_sum_l 887 5.1 2.723 4.252 2.723 4.252 dbcsr_matrix_vector_mult_local 304 10.0 3.727 4.192 3.729 4.194 calculate_norms 792 9.8 3.620 3.735 3.620 3.735 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.537 3.729 mp_irecv_dv 1241 11.2 1.523 3.695 1.523 3.695 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.512 3.557 mp_allgather_i34 111 8.7 1.138 3.523 1.138 3.523 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.919 3.454 ls_scf_post 1 4.0 0.000 0.000 3.394 3.399 make_images_sizes 222 9.7 0.000 0.000 0.840 3.255 mp_alltoall_i44 222 10.7 0.840 3.254 0.840 3.254 ls_scf_store_result 1 5.0 0.000 0.000 3.183 3.218 dbcsr_data_new 4608 9.7 1.782 2.319 1.782 2.319 dbcsr_finalize 304 7.8 0.062 0.077 2.199 2.302 dbcsr_merge_all 275 8.9 0.479 0.532 2.058 2.140 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.024 2.025 rebuild_ks_matrix 3 7.3 0.000 0.000 1.992 1.993 qs_ks_build_kohn_sham_matrix 3 8.3 0.011 0.015 1.992 1.993 qs_energies_init_hamiltonians 1 3.0 0.001 0.002 1.920 1.921 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=91.898000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3699.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/164be73c51fd9cb7dc5314b3417fa4c35b0a34c7_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 8.935457E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284089. MP_Allreduce 3123 21388. MP_Sync 4 MP_Alltoall 47 88727262. MP_SendRecv 42 732600. MP_ISendRecv 42 732600. MP_Wait 267 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.074 0.097 106.448 106.448 qs_energies 1 2.0 0.000 0.000 105.029 105.036 ls_scf 1 3.0 0.000 0.000 102.090 102.096 dbcsr_multiply_generic 111 6.7 0.024 0.027 75.253 75.406 ls_scf_main 1 4.0 0.000 0.000 64.195 64.196 density_matrix_trs4 2 5.0 0.002 0.003 55.126 55.196 multiply_cannon 111 7.7 0.109 0.201 48.530 50.947 multiply_cannon_loop 111 8.7 0.097 0.099 45.397 46.062 ls_scf_init_scf 1 4.0 0.000 0.000 34.118 34.119 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.553 32.591 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.848 29.857 mp_waitall_1 4527 11.1 21.941 26.151 21.941 26.151 make_m2s 222 7.7 0.005 0.005 22.929 24.141 make_images 222 8.7 3.572 3.891 22.821 24.032 multiply_cannon_multrec 444 9.7 17.856 18.511 22.558 23.251 hybrid_alltoall_any 227 10.6 1.654 3.623 13.021 15.971 make_images_data 222 9.7 0.003 0.004 13.230 15.609 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.104 10.476 multiply_cannon_sync_h2d 444 9.7 8.788 8.832 8.788 8.832 arnoldi_extremal 4 6.8 0.000 0.000 7.650 7.661 arnoldi_normal_ev 4 7.8 0.003 0.011 7.650 7.661 build_subspace 16 8.4 0.026 0.036 7.094 7.108 dbcsr_matrix_vector_mult 304 9.0 0.017 0.033 5.607 5.738 dbcsr_matrix_vector_mult_local 304 10.0 5.102 5.397 5.105 5.400 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.174 5.261 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.875 5.144 dbcsr_mm_accdrv_process 1814 10.4 0.267 0.357 4.512 4.640 dbcsr_mm_accdrv_process_sort 1814 11.4 4.175 4.310 4.175 4.310 mp_allgather_i34 111 8.7 1.249 3.976 1.249 3.976 make_images_sizes 222 9.7 0.000 0.000 1.574 3.839 mp_alltoall_i44 222 10.7 1.574 3.838 1.574 3.838 ls_scf_post 1 4.0 0.000 0.000 3.776 3.783 ls_scf_store_result 1 5.0 0.000 0.000 3.494 3.532 calculate_norms 792 9.8 3.234 3.272 3.234 3.272 dbcsr_finalize 304 7.8 0.082 0.089 3.059 3.120 qs_energies_init_hamiltonians 1 3.0 0.001 0.002 2.908 2.909 dbcsr_merge_all 275 8.9 0.885 0.921 2.844 2.899 dbcsr_complete_redistribute 5 7.6 1.432 1.475 2.751 2.891 matrix_ls_to_qs 2 6.0 0.000 0.000 2.414 2.561 dbcsr_sort_data 325 11.1 2.432 2.488 2.432 2.488 dbcsr_new_transposed 4 7.5 0.241 0.252 2.408 2.432 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.429 2.431 dbcsr_data_new 6591 9.6 1.884 2.370 1.884 2.370 rebuild_ks_matrix 3 7.3 0.000 0.000 2.362 2.364 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 2.362 2.364 mp_alltoall_d11v 48 9.2 2.173 2.243 2.173 2.243 dbcsr_frobenius_norm 74 6.6 2.056 2.133 2.191 2.220 dbcsr_add_d 103 6.2 0.000 0.000 2.125 2.203 dbcsr_add_anytype 103 7.2 0.858 0.890 2.125 2.203 dbcsr_data_release 12724 10.6 1.963 2.190 1.963 2.190 dbcsr_redistribute 4 8.5 1.372 1.451 2.139 2.152 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=106.448000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=7042.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/164be73c51fd9cb7dc5314b3417fa4c35b0a34c7_performance_tests/27/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 7009386627072 0.0% 0.0% 100.0% flops 9 x 9 x 32 7335108845568 0.0% 0.0% 100.0% flops 9 x 22 x 32 9866241589248 0.0% 0.0% 100.0% flops 22 x 9 x 32 9884108906496 0.0% 0.0% 100.0% flops 22 x 22 x 32 13354440523776 0.0% 0.0% 100.0% flops 32 x 32 x 9 20607185977344 0.0% 0.0% 100.0% flops 32 x 32 x 22 25186560638976 0.0% 0.0% 100.0% flops 9 x 32 x 32 28458319085568 0.0% 0.0% 100.0% flops 22 x 32 x 32 34782389993472 0.0% 0.0% 100.0% flops 9 x 32 x 9 42881542373376 0.0% 0.0% 100.0% flops 22 x 32 x 9 55680402235392 0.0% 0.0% 100.0% flops 9 x 32 x 22 55680402235392 0.0% 0.0% 100.0% flops 22 x 32 x 22 72328573419520 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 383.054662E+12 0.0% 0.0% 100.0% flops max/rank 733.641090E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 26899403712 0.0% 0.0% 100.0% number of processed stacks 118860288 0.0% 0.0% 100.0% average stack size 0.0 0.0 226.3 marketing flops 780.439111E+12 ------------------------------------------------------------------------------- # multiplications 1445 max memory usage/rank 588.718080E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 102097920 MPI messages size (bytes): total size 37.227590E+12 min size 0.000000E+00 max size 4.551360E+06 average size 364.626312E+03 MPI breakdown and total messages size (bytes): size <= 128 731472 0 128 < size <= 8192 11922720 97670922240 8192 < size <= 32768 24718992 614677610496 32768 < size <= 131072 20000256 1970081366016 131072 < size <= 4194304 42515668 24886801223040 4194304 < size <= 16777216 2208812 9656099886720 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4640 78072. MP_Allreduce 13232 2081. MP_Sync 1064 MP_Alltoall 2588 1145522522. MP_SendRecv 168740 11136. MP_ISendRecv 92040 11136. MP_Wait 102830 MP_comm_split 40 MP_ISend 26090 85106. MP_IRecv 37890 59644. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.042 0.149 274.955 274.957 qs_mol_dyn_low 1 2.0 0.004 0.045 273.673 273.752 qs_forces 5 3.8 0.005 0.046 273.525 273.527 qs_energies 5 4.8 0.003 0.038 270.187 270.209 scf_env_do_scf 5 5.8 0.000 0.002 254.124 254.126 scf_env_do_scf_inner_loop 105 6.6 0.002 0.013 218.675 218.684 qs_scf_new_mos 105 7.6 0.000 0.001 170.096 170.487 qs_scf_loop_do_ot 105 8.6 0.001 0.001 170.095 170.486 ot_scf_mini 105 9.6 0.003 0.005 159.857 160.187 dbcsr_multiply_generic 1445 12.2 0.127 0.134 139.369 140.413 multiply_cannon 1445 13.2 0.278 0.289 117.600 120.036 velocity_verlet 4 3.0 0.002 0.020 118.634 118.636 multiply_cannon_loop 1445 14.2 2.851 2.991 115.048 117.675 qs_ot_get_p 112 10.4 0.001 0.001 65.486 65.764 ot_mini 105 10.6 0.001 0.002 63.550 63.693 qs_ot_p2m_diag 40 11.0 0.020 0.030 53.330 53.424 cp_dbcsr_syevd 40 12.0 0.002 0.002 49.651 49.653 mp_waitall_1 488190 16.1 39.254 48.774 39.254 48.774 multiply_cannon_multrec 69360 15.2 29.474 34.297 39.979 45.075 cp_fm_syevd 40 13.0 0.000 0.000 43.584 43.715 qs_ot_get_derivative 55 11.6 0.001 0.003 41.619 41.792 cp_fm_redistribute_end 40 14.0 19.103 38.091 19.111 38.096 cp_fm_syevd_base 40 14.0 18.968 37.963 18.968 37.963 multiply_cannon_metrocomm3 69360 15.2 0.202 0.216 26.996 37.799 init_scf_loop 7 6.6 0.000 0.005 35.407 35.410 rebuild_ks_matrix 110 8.4 0.000 0.000 32.408 32.587 qs_ks_build_kohn_sham_matrix 110 9.4 0.012 0.024 32.407 32.586 multiply_cannon_sync_h2d 69360 15.2 28.802 32.355 28.802 32.355 qs_ks_update_qs_env 112 7.6 0.001 0.001 29.796 29.954 prepare_preconditioner 7 7.6 0.000 0.000 29.856 29.892 make_preconditioner 7 8.6 0.000 0.002 29.856 29.892 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 23.167 23.375 apply_single 62 13.6 0.000 0.000 23.167 23.375 qs_rho_update_rho_low 110 7.6 0.001 0.001 22.890 23.268 calculate_rho_elec 110 8.6 0.030 0.033 22.889 23.267 ot_new_cg_direction 55 11.6 0.001 0.003 21.186 21.188 make_full_inverse_cholesky 7 9.6 0.000 0.000 20.795 20.850 density_rs2pw 110 9.6 0.005 0.007 16.678 17.329 qs_ot_get_orbitals 105 10.6 0.001 0.001 15.405 15.703 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 14.746 14.856 cp_fm_cholesky_invert 7 10.6 14.089 14.102 14.089 14.102 mp_sum_l 4764 12.2 12.901 13.635 12.901 13.635 transfer_rs2pw 445 10.6 0.007 0.008 12.796 13.547 pw_transfer 1645 12.4 0.080 0.105 13.052 13.340 fft_wrap_pw1pw2 1425 13.5 0.012 0.016 12.912 13.206 init_scf_run 5 5.8 0.000 0.001 13.189 13.190 scf_env_initial_rho_setup 5 6.8 0.002 0.003 13.189 13.190 calculate_dm_sparse 110 9.5 0.000 0.001 11.894 12.101 dbcsr_mm_accdrv_process 154766 15.8 6.222 6.415 10.375 11.494 fft_wrap_pw1pw2_240 915 15.0 1.139 1.230 11.069 11.321 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 11.181 11.275 sum_up_and_integrate 60 10.3 0.001 0.003 11.212 11.226 integrate_v_rspace 60 11.3 0.002 0.002 11.195 11.210 qs_vxc_create 110 10.4 0.002 0.005 11.020 11.074 check_diag 80 13.5 8.706 8.935 10.840 10.967 multiply_cannon_metrocomm1 69360 15.2 0.096 0.103 5.252 9.766 fft3d_pb 915 16.0 2.383 2.639 9.273 9.617 make_m2s 2890 13.2 0.078 0.084 8.458 9.027 make_images 2890 14.2 0.239 0.259 8.352 8.920 make_full_single_inverse 7 9.6 0.001 0.001 8.732 8.769 acc_transpose_blocks 69360 15.2 0.351 0.375 8.058 8.614 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.002 8.134 8.151 xc_rho_set_and_dset_create 110 12.4 0.077 0.097 7.668 7.930 calculate_first_density_matrix 1 7.0 0.000 0.004 7.659 7.672 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 7.583 7.657 xc_vxc_pw_create 60 11.3 0.039 0.052 7.373 7.428 potential_pw2rs 60 12.3 0.003 0.003 7.147 7.221 transfer_rs2pw_30 110 11.6 1.267 1.321 6.491 7.084 mp_sendrecv_dv 168740 12.6 6.772 6.985 6.772 6.985 xc_pw_derive 510 13.4 0.005 0.006 6.701 6.771 mp_alltoall_z22v 2340 17.7 6.384 6.758 6.384 6.758 cp_fm_cholesky_decompose 14 10.2 6.119 6.128 6.119 6.128 transfer_pw2rs 245 13.2 0.003 0.004 5.910 5.982 acc_transpose_blocks_kernels 69360 16.2 0.853 0.891 5.302 5.759 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=274.957000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=560.600000, yerr=1.959592 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 164be73c51fd9cb7dc5314b3417fa4c35b0a34c7 Summary: empty Status: OK