=== 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: 04dd4284328246f8f7259e2ea433868052050a67 ################# 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 1.12.0, 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: 13.09.2023 # # \ 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=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2023.05.001 USE_HDF5 := 1.12.0 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.9.0 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(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/04dd4284328246f8f7259e2ea433868052050a67_performance_tests/01 job id: 49154308 --- 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/04dd4284328246f8f7259e2ea433868052050a67_performance_tests/02 job id: 49154309 --- 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/04dd4284328246f8f7259e2ea433868052050a67_performance_tests/03 job id: 49154311 --- 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/04dd4284328246f8f7259e2ea433868052050a67_performance_tests/04 job id: 49154312 --- 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/04dd4284328246f8f7259e2ea433868052050a67_performance_tests/05 job id: 49154314 --- 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/04dd4284328246f8f7259e2ea433868052050a67_performance_tests/06 job id: 49154315 --- 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/04dd4284328246f8f7259e2ea433868052050a67_performance_tests/07 job id: 49154316 --- 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/04dd4284328246f8f7259e2ea433868052050a67_performance_tests/08 job id: 49154318 --- 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/04dd4284328246f8f7259e2ea433868052050a67_performance_tests/09 job id: 49154319 --- 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/04dd4284328246f8f7259e2ea433868052050a67_performance_tests/10 job id: 49154320 --- 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/04dd4284328246f8f7259e2ea433868052050a67_performance_tests/11 job id: 49154322 --- 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/04dd4284328246f8f7259e2ea433868052050a67_performance_tests/12 job id: 49154323 --- 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/04dd4284328246f8f7259e2ea433868052050a67_performance_tests/13 job id: 49154324 --- 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/04dd4284328246f8f7259e2ea433868052050a67_performance_tests/14 job id: 49154326 --- 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/04dd4284328246f8f7259e2ea433868052050a67_performance_tests/15 job id: 49154327 --- 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/04dd4284328246f8f7259e2ea433868052050a67_performance_tests/16 job id: 49154329 --- 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/04dd4284328246f8f7259e2ea433868052050a67_performance_tests/17 job id: 49154330 --- 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/04dd4284328246f8f7259e2ea433868052050a67_performance_tests/18 job id: 49154332 --- 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/04dd4284328246f8f7259e2ea433868052050a67_performance_tests/19 job id: 49154333 --- 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/04dd4284328246f8f7259e2ea433868052050a67_performance_tests/20 job id: 49154335 --- 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/04dd4284328246f8f7259e2ea433868052050a67_performance_tests/21 job id: 49154336 --- 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/04dd4284328246f8f7259e2ea433868052050a67_performance_tests/22 job id: 49154337 --- 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/04dd4284328246f8f7259e2ea433868052050a67_performance_tests/23 job id: 49154339 --- 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/04dd4284328246f8f7259e2ea433868052050a67_performance_tests/24 job id: 49154341 --- 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/04dd4284328246f8f7259e2ea433868052050a67_performance_tests/25 job id: 49154344 --- 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/04dd4284328246f8f7259e2ea433868052050a67_performance_tests/26 job id: 49154345 --- 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/04dd4284328246f8f7259e2ea433868052050a67_performance_tests/27 job id: 49154346 --- 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/04dd4284328246f8f7259e2ea433868052050a67_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 0.018 0.034 133.914 133.915 farming_run 1 2.0 133.411 133.412 133.884 133.887 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.482342E+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 0.062 0.083 115.163 115.164 qs_energies 1 2.0 0.000 0.000 114.741 114.752 mp2_main 1 3.0 0.000 0.000 112.570 112.581 mp2_gpw_main 1 4.0 0.055 0.061 111.510 111.520 mp2_ri_gpw_compute_in 1 5.0 0.199 0.243 92.309 92.725 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 54.698 55.114 mp2_eri_3c_integrate_gpw 272 7.0 0.152 0.163 41.090 46.216 get_2c_integrals 1 6.0 0.008 0.009 36.767 37.414 integrate_v_rspace 273 8.0 0.439 0.455 24.649 29.623 pw_transfer 6555 10.6 0.381 0.395 26.763 27.260 grid_integrate_task_list 273 9.0 20.553 25.998 20.553 25.998 fft_wrap_pw1pw2 5465 11.4 0.043 0.046 25.402 25.963 fft_wrap_pw1pw2_100 2178 12.4 1.096 1.252 22.965 23.521 rpa_ri_compute_en 1 5.0 0.037 0.042 19.058 19.273 compute_2c_integrals 1 7.0 0.020 0.021 18.942 18.942 compute_2c_integrals_loop_lm 1 8.0 0.024 0.024 18.544 18.633 mp2_eri_2c_integrate_gpw 1 9.0 2.382 2.417 18.520 18.609 cp_fm_cholesky_decompose 12 8.2 17.821 18.455 17.821 18.455 cholesky_decomp 1 7.0 0.000 0.000 16.668 17.304 fft3d_s 5443 13.4 16.097 16.561 16.121 16.584 ao_to_mo_and_store_B_mult_1 272 7.0 10.772 15.311 10.772 15.311 calculate_wavefunction 272 8.0 5.431 5.593 12.351 13.031 rpa_num_int 1 6.0 0.002 0.029 10.765 10.779 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.699 10.730 calc_mat_Q 8 8.0 0.000 0.000 9.494 9.591 contract_S_to_Q 8 9.0 0.000 0.000 8.915 9.023 calc_potential_gpw 544 9.5 0.005 0.006 8.244 8.705 parallel_gemm_fm 14 9.1 0.000 0.000 8.499 8.644 parallel_gemm_fm_cosma 14 10.1 8.499 8.644 8.499 8.644 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.038 8.276 potential_pw2rs 545 10.0 0.106 0.108 7.435 8.089 create_integ_mat 1 6.0 0.022 0.028 7.764 7.764 collocate_single_gaussian 272 10.0 0.040 0.043 7.290 7.498 array2fm 1 7.0 0.000 0.000 6.653 7.093 pw_scatter_s 2720 13.7 4.367 4.506 4.367 4.506 pw_gather_s 2722 13.2 3.443 3.847 3.443 3.847 array2fm_buffer_send 1 8.0 2.919 3.147 2.919 3.147 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.520461, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2796.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/04dd4284328246f8f7259e2ea433868052050a67_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 0.042 431.709 431.710 farming_run 1 2.0 430.710 430.731 431.665 431.669 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827128576 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788821 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.246097E+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 26246942. 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 0.015 0.034 218.231 218.232 qs_energies 1 2.0 0.003 0.021 217.749 217.791 scf_env_do_scf 1 3.0 0.000 0.000 114.999 115.000 qs_ks_update_qs_env 5 5.0 0.000 0.000 114.023 114.028 rebuild_ks_matrix 4 6.0 0.000 0.000 114.021 114.027 qs_ks_build_kohn_sham_matrix 4 7.0 0.055 0.061 114.021 114.027 hfx_ks_matrix 4 8.0 0.001 0.001 113.667 113.672 integrate_four_center 4 9.0 0.153 0.468 113.666 113.671 integrate_four_center_main 4 10.0 0.114 0.621 101.854 105.106 integrate_four_center_bin 267 11.0 101.740 104.931 101.740 104.931 mp2_main 1 3.0 0.006 0.047 102.448 102.473 mp2_gpw_main 1 4.0 0.062 0.181 101.451 101.478 init_scf_loop 1 4.0 0.000 0.000 96.960 96.961 mp2_ri_gpw_compute_in 1 5.0 0.072 0.155 74.015 75.226 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.004 53.821 55.033 mp2_eri_3c_integrate_gpw 91 7.0 0.145 0.161 41.609 46.596 integrate_v_rspace 95 8.0 0.398 0.563 28.083 32.864 pw_transfer 2240 10.6 0.146 0.167 29.501 29.948 mp2_ri_gpw_compute_en 1 5.0 0.080 0.243 27.147 29.230 fft_wrap_pw1pw2 1868 11.4 0.018 0.021 28.518 28.898 grid_integrate_task_list 95 9.0 23.497 28.464 23.497 28.464 ao_to_mo_and_store_B_mult_1 91 7.0 10.538 28.425 10.538 28.425 fft_wrap_pw1pw2_100 730 12.4 1.281 1.495 26.255 26.689 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.837 1.901 25.043 25.058 get_2c_integrals 1 6.0 0.002 0.021 20.090 20.129 compute_2c_integrals 1 7.0 0.004 0.016 19.067 19.080 fft3d_s 1823 13.4 18.495 18.997 18.510 19.012 compute_2c_integrals_loop_lm 1 8.0 0.004 0.025 18.597 18.920 mp2_eri_2c_integrate_gpw 1 9.0 1.736 1.813 18.593 18.919 scf_env_do_scf_inner_loop 4 4.0 0.000 0.001 18.037 18.037 calculate_wavefunction 91 8.0 2.015 2.058 9.620 9.834 mp2_ri_gpw_compute_en_expansio 172 7.0 0.556 0.577 8.886 9.441 potential_pw2rs 186 10.0 0.033 0.034 8.430 8.927 local_gemm 172 8.0 8.330 8.874 8.330 8.874 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.138 8.538 mp2_ri_gpw_compute_en_comm 22 7.0 0.492 0.512 7.937 8.360 calc_potential_gpw 182 9.5 0.002 0.002 7.887 8.111 collocate_single_gaussian 91 10.0 0.018 0.047 7.742 7.994 mp_sync 37 10.5 3.783 7.178 3.783 7.178 integrate_four_center_load 4 10.0 0.000 0.000 6.761 6.766 hfx_load_balance 1 11.0 0.000 0.000 6.761 6.766 mp_sendrecv_dm3 2068 8.0 5.987 6.446 5.987 6.446 mp2_ri_gpw_compute_en_ener 172 7.0 6.338 6.421 6.338 6.421 pw_gather_s 912 13.2 4.471 4.872 4.471 4.872 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.484898, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1505.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/04dd4284328246f8f7259e2ea433868052050a67_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 452.616192E+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 0.027 0.057 55.409 55.412 qs_mol_dyn_low 1 2.0 0.003 0.006 54.994 55.054 qs_forces 11 3.9 0.012 0.081 54.845 54.882 qs_energies 11 4.9 0.016 0.051 53.258 53.281 scf_env_do_scf 11 5.9 0.002 0.015 45.673 45.674 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 43.497 43.498 dbcsr_multiply_generic 2286 12.5 0.096 0.101 34.305 35.020 qs_scf_new_mos 108 7.5 0.000 0.001 33.170 33.434 qs_scf_loop_do_ot 108 8.5 0.001 0.001 33.169 33.433 ot_scf_mini 108 9.5 0.002 0.003 31.535 31.701 multiply_cannon 2286 13.5 0.190 0.197 27.238 29.381 multiply_cannon_loop 2286 14.5 1.813 1.948 26.518 28.640 velocity_verlet 10 3.0 0.006 0.012 26.566 26.569 ot_mini 108 10.5 0.001 0.002 18.919 19.171 qs_ot_get_derivative 108 11.5 0.001 0.001 15.956 16.145 mp_waitall_1 245248 16.5 8.626 16.075 8.626 16.075 multiply_cannon_metrocomm3 54864 15.5 0.073 0.082 6.246 14.374 multiply_cannon_multrec 54864 15.5 3.688 5.921 7.587 10.967 qs_ot_get_p 119 10.4 0.001 0.003 8.094 8.360 rebuild_ks_matrix 119 8.3 0.000 0.000 8.090 8.252 qs_ks_build_kohn_sham_matrix 119 9.3 0.015 0.028 8.090 8.252 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.119 7.260 mp_sum_l 7287 12.8 4.953 6.910 4.953 6.910 init_scf_run 11 5.9 0.000 0.001 6.165 6.165 scf_env_initial_rho_setup 11 6.9 0.001 0.003 6.165 6.165 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.611 6.036 multiply_cannon_sync_h2d 54864 15.5 5.172 5.844 5.172 5.844 dbcsr_mm_accdrv_process 76910 16.1 1.809 2.913 3.812 5.521 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.193 5.228 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.037 5.125 sum_up_and_integrate 119 10.3 0.002 0.006 4.641 4.661 integrate_v_rspace 119 11.3 0.002 0.004 4.629 4.651 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.502 4.503 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.300 4.440 calculate_rho_elec 119 8.7 0.011 0.017 4.299 4.439 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.284 4.285 cp_fm_redistribute_end 50 14.0 2.186 4.258 2.194 4.261 acc_transpose_blocks 54864 15.5 0.225 0.256 3.174 4.239 cp_fm_diag_elpa_base 50 14.0 2.061 4.141 2.066 4.151 calculate_first_density_matrix 1 7.0 0.002 0.006 3.947 3.952 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.707 3.710 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.419 3.483 apply_preconditioner_dbcsr 119 12.6 0.000 0.001 2.933 3.139 apply_single 119 13.6 0.000 0.000 2.932 3.139 calculate_dm_sparse 119 9.5 0.000 0.001 2.854 3.002 ot_diis_step 108 11.5 0.006 0.009 2.703 2.704 multiply_cannon_metrocomm1 54864 15.5 0.055 0.061 1.593 2.671 jit_kernel_multiply 13 15.8 1.941 2.535 1.941 2.535 acc_transpose_blocks_kernels 54864 16.5 0.243 0.383 1.731 2.491 density_rs2pw 119 9.7 0.004 0.004 2.328 2.425 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.333 2.398 init_scf_loop 11 6.9 0.004 0.013 2.144 2.147 wfi_extrapolate 11 7.9 0.001 0.003 2.131 2.131 jit_kernel_transpose 5 15.5 1.488 2.116 1.488 2.116 mp_sum_d 4135 12.0 1.363 2.115 1.363 2.115 grid_integrate_task_list 119 12.3 2.018 2.114 2.018 2.114 potential_pw2rs 119 12.3 0.004 0.004 1.983 2.011 pw_transfer 1439 11.6 0.052 0.056 1.826 1.905 make_m2s 4572 13.5 0.054 0.056 1.815 1.864 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.750 1.829 make_images 4572 14.5 0.134 0.140 1.732 1.780 transfer_rs2pw 487 10.6 0.005 0.006 1.522 1.640 mp_alltoall_d11v 2130 13.8 1.363 1.561 1.363 1.561 fft3d_ps 1201 14.6 0.371 0.477 1.465 1.541 mp_waitany 12084 13.8 1.308 1.516 1.308 1.516 fft_wrap_pw1pw2_140 487 13.2 0.141 0.154 1.368 1.445 acc_transpose_blocks_sync 164592 16.5 1.174 1.443 1.174 1.443 transfer_pw2rs 487 13.2 0.006 0.006 1.421 1.438 grid_collocate_task_list 119 9.7 1.356 1.411 1.356 1.411 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.384 1.409 dbcsr_dot_sd 1205 11.9 0.048 0.059 0.734 1.201 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=55.412000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=431.090909, yerr=1.239835 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/04dd4284328246f8f7259e2ea433868052050a67_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 488.857600E+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 789605. 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 0.007 0.027 41.611 41.612 qs_mol_dyn_low 1 2.0 0.003 0.003 41.422 41.431 qs_forces 11 3.9 0.002 0.003 40.888 40.889 qs_energies 11 4.9 0.002 0.002 39.192 39.197 scf_env_do_scf 11 5.9 0.000 0.001 32.590 32.590 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 30.037 30.037 dbcsr_multiply_generic 2286 12.5 0.102 0.105 23.806 24.142 qs_scf_new_mos 108 7.5 0.001 0.001 21.022 21.250 qs_scf_loop_do_ot 108 8.5 0.001 0.001 21.022 21.249 ot_scf_mini 108 9.5 0.002 0.003 20.103 20.265 multiply_cannon 2286 13.5 0.209 0.218 18.379 19.773 velocity_verlet 10 3.0 0.001 0.002 19.413 19.424 multiply_cannon_loop 2286 14.5 1.196 1.254 17.167 18.695 ot_mini 108 10.5 0.001 0.001 12.378 12.610 mp_waitall_1 200699 16.5 6.202 12.306 6.202 12.306 multiply_cannon_metrocomm3 27432 15.5 0.072 0.075 4.474 10.893 qs_ot_get_derivative 108 11.5 0.001 0.001 9.991 10.156 multiply_cannon_multrec 27432 15.5 1.834 4.229 6.164 8.952 rebuild_ks_matrix 119 8.3 0.000 0.000 7.015 7.156 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.015 7.156 dbcsr_mm_accdrv_process 47894 16.0 3.556 5.874 4.250 6.341 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.175 6.303 init_scf_run 11 5.9 0.000 0.003 5.395 5.395 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.395 5.395 qs_ot_get_p 119 10.4 0.002 0.005 4.765 4.985 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.741 4.659 mp_sum_l 7287 12.8 2.138 4.149 2.138 4.149 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.089 4.075 apply_single 119 13.6 0.000 0.000 3.089 4.074 sum_up_and_integrate 119 10.3 0.001 0.001 3.916 3.925 integrate_v_rspace 119 11.3 0.002 0.003 3.902 3.911 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.840 3.873 calculate_rho_elec 119 8.7 0.021 0.024 3.840 3.873 calculate_first_density_matrix 1 7.0 0.000 0.001 3.803 3.806 acc_transpose_blocks 27432 15.5 0.114 0.120 2.594 3.420 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.386 3.387 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.161 3.204 make_m2s 4572 13.5 0.053 0.054 2.834 3.168 make_images 4572 14.5 0.207 0.245 2.745 3.081 qs_ot_p2m_diag 50 11.0 0.009 0.013 3.025 3.044 cp_dbcsr_syevd 50 12.0 0.003 0.005 2.619 2.620 init_scf_loop 11 6.9 0.000 0.000 2.530 2.530 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.387 2.472 ot_diis_step 108 11.5 0.011 0.011 2.338 2.339 acc_transpose_blocks_kernels 27432 16.5 0.188 0.276 1.636 2.337 multiply_cannon_sync_h2d 27432 15.5 1.685 2.306 1.685 2.306 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.258 2.259 cp_fm_redistribute_end 50 14.0 1.144 2.230 1.147 2.232 density_rs2pw 119 9.7 0.004 0.004 2.116 2.199 cp_fm_diag_elpa_base 50 14.0 1.054 2.141 1.081 2.175 calculate_dm_sparse 119 9.5 0.000 0.001 2.068 2.156 jit_kernel_transpose 5 15.5 1.448 2.065 1.448 2.065 pw_transfer 1439 11.6 0.065 0.069 1.974 2.010 grid_integrate_task_list 119 12.3 1.836 1.929 1.836 1.929 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.883 1.921 make_images_data 4572 15.5 0.048 0.054 1.314 1.705 potential_pw2rs 119 12.3 0.006 0.006 1.644 1.659 jit_kernel_multiply 8 16.5 0.636 1.632 0.636 1.632 prepare_preconditioner 11 7.9 0.000 0.000 1.599 1.625 make_preconditioner 11 8.9 0.000 0.000 1.599 1.625 hybrid_alltoall_any 4725 16.4 0.054 0.117 1.164 1.556 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.491 1.552 fft3d_ps 1201 14.6 0.520 0.575 1.518 1.551 wfi_extrapolate 11 7.9 0.001 0.001 1.520 1.520 fft_wrap_pw1pw2_140 487 13.2 0.160 0.168 1.457 1.494 mp_alltoall_d11v 2130 13.8 1.306 1.427 1.306 1.427 grid_collocate_task_list 119 9.7 1.290 1.353 1.290 1.353 transfer_rs2pw 487 10.6 0.005 0.005 1.230 1.310 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.267 1.277 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.213 1.256 mp_allgather_i34 2286 14.5 0.633 1.069 0.633 1.069 transfer_pw2rs 487 13.2 0.005 0.006 1.040 1.050 mp_sum_d 4135 12.0 0.549 1.032 0.549 1.032 acc_transpose_blocks_sync 82296 16.5 0.820 0.950 0.820 0.950 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.931 0.932 make_images_sizes 4572 15.5 0.005 0.005 0.651 0.926 mp_alltoall_i44 4572 16.5 0.646 0.922 0.646 0.922 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.897 0.910 mp_alltoall_z22v 1201 16.6 0.770 0.843 0.770 0.843 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=41.612000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=465.363636, yerr=1.493789 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/04dd4284328246f8f7259e2ea433868052050a67_performance_tests/05/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 59.051995E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3143552 0.0% 0.0% 100.0% average stack size 0.0 0.0 46.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 519.540736E+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 0.019 0.077 34.408 34.410 qs_mol_dyn_low 1 2.0 0.003 0.004 34.025 34.036 qs_forces 11 3.9 0.004 0.018 33.949 33.950 qs_energies 11 4.9 0.003 0.013 32.363 32.366 scf_env_do_scf 11 5.9 0.001 0.002 26.046 26.046 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 23.489 23.489 dbcsr_multiply_generic 2286 12.5 0.099 0.105 18.572 18.652 multiply_cannon 2286 13.5 0.196 0.207 15.241 15.871 qs_scf_new_mos 108 7.5 0.001 0.001 15.628 15.640 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.628 15.640 velocity_verlet 10 3.0 0.001 0.002 15.624 15.626 multiply_cannon_loop 2286 14.5 0.861 0.912 14.436 15.105 ot_scf_mini 108 9.5 0.002 0.003 14.865 14.876 ot_mini 108 10.5 0.001 0.001 9.196 9.206 qs_ot_get_derivative 108 11.5 0.001 0.001 7.709 7.718 multiply_cannon_multrec 18288 15.5 1.873 2.812 6.960 7.231 rebuild_ks_matrix 119 8.3 0.000 0.000 6.115 6.127 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 6.115 6.126 dbcsr_mm_accdrv_process 38222 16.0 4.899 5.836 4.993 5.901 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.379 5.391 init_scf_run 11 5.9 0.000 0.002 5.071 5.071 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.071 5.071 mp_waitall_1 158411 16.6 2.854 4.000 2.854 4.000 calculate_first_density_matrix 1 7.0 0.001 0.004 3.769 3.770 sum_up_and_integrate 119 10.3 0.001 0.001 3.671 3.676 integrate_v_rspace 119 11.3 0.003 0.005 3.658 3.665 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.993 3.518 qs_ot_get_p 119 10.4 0.001 0.002 3.419 3.440 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.366 3.374 calculate_rho_elec 119 8.7 0.031 0.031 3.365 3.373 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.243 3.245 acc_transpose_blocks 18288 15.5 0.080 0.081 3.074 3.160 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.027 3.035 init_scf_loop 11 6.9 0.001 0.003 2.538 2.539 multiply_cannon_metrocomm3 18288 15.5 0.047 0.049 1.508 2.498 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.936 2.355 apply_single 119 13.6 0.000 0.000 1.936 2.355 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.265 2.272 acc_transpose_blocks_kernels 18288 16.5 0.217 0.225 2.232 2.243 make_m2s 4572 13.5 0.045 0.046 1.927 2.076 jit_kernel_transpose 5 15.6 2.014 2.025 2.014 2.025 make_images 4572 14.5 0.194 0.206 1.841 1.988 density_rs2pw 119 9.7 0.004 0.004 1.919 1.980 cp_dbcsr_syevd 50 12.0 0.003 0.004 1.963 1.965 pw_transfer 1439 11.6 0.065 0.069 1.888 1.899 grid_integrate_task_list 119 12.3 1.805 1.877 1.805 1.877 calculate_dm_sparse 119 9.5 0.000 0.001 1.856 1.866 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.796 1.808 prepare_preconditioner 11 7.9 0.000 0.000 1.763 1.767 make_preconditioner 11 8.9 0.000 0.002 1.763 1.767 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.732 1.739 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.711 1.722 cp_fm_diag_elpa_base 50 14.0 1.688 1.700 1.709 1.721 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.612 1.702 ot_diis_step 108 11.5 0.011 0.012 1.473 1.473 potential_pw2rs 119 12.3 0.007 0.009 1.461 1.467 mp_sum_l 7287 12.8 1.023 1.458 1.023 1.458 fft_wrap_pw1pw2_140 487 13.2 0.211 0.217 1.425 1.436 fft3d_ps 1201 14.6 0.526 0.544 1.365 1.377 grid_collocate_task_list 119 9.7 1.235 1.333 1.235 1.333 wfi_extrapolate 11 7.9 0.001 0.001 1.244 1.244 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.190 1.194 multiply_cannon_sync_h2d 18288 15.5 1.045 1.192 1.045 1.192 transfer_rs2pw 487 10.6 0.005 0.005 1.112 1.182 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.008 1.010 make_images_data 4572 15.5 0.048 0.052 0.826 1.001 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.954 0.971 transfer_pw2rs 487 13.2 0.004 0.004 0.902 0.906 hybrid_alltoall_any 4725 16.4 0.058 0.118 0.717 0.899 mp_alltoall_d11v 2130 13.8 0.719 0.852 0.719 0.852 acc_transpose_blocks_sync 54864 16.5 0.743 0.818 0.743 0.818 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.809 0.810 jit_kernel_multiply 4 15.7 0.036 0.778 0.036 0.778 mp_alltoall_z22v 1201 16.6 0.681 0.748 0.681 0.748 cp_fm_cholesky_invert 11 10.9 0.742 0.747 0.742 0.747 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.003 0.674 0.739 multiply_cannon_metrocomm1 18288 15.5 0.029 0.031 0.462 0.708 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=34.410000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=494.727273, yerr=1.052349 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/04dd4284328246f8f7259e2ea433868052050a67_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 557.539328E+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 0.011 0.040 40.795 40.798 qs_mol_dyn_low 1 2.0 0.003 0.003 40.520 40.529 qs_forces 11 3.9 0.002 0.003 40.457 40.458 qs_energies 11 4.9 0.001 0.002 38.684 38.692 scf_env_do_scf 11 5.9 0.000 0.001 31.842 31.843 scf_env_do_scf_inner_loop 108 6.5 0.003 0.008 28.155 28.155 dbcsr_multiply_generic 2286 12.5 0.102 0.105 22.403 22.577 velocity_verlet 10 3.0 0.001 0.002 19.840 19.842 qs_scf_new_mos 108 7.5 0.001 0.001 19.268 19.324 qs_scf_loop_do_ot 108 8.5 0.001 0.001 19.267 19.323 multiply_cannon 2286 13.5 0.219 0.225 17.992 18.431 ot_scf_mini 108 9.5 0.002 0.002 18.188 18.246 multiply_cannon_loop 2286 14.5 1.490 1.578 16.912 17.300 ot_mini 108 10.5 0.001 0.001 11.030 11.083 multiply_cannon_multrec 27432 15.5 2.446 3.115 9.087 9.371 qs_ot_get_derivative 108 11.5 0.001 0.001 9.054 9.113 dbcsr_mm_accdrv_process 47916 15.9 6.128 7.788 6.538 7.848 rebuild_ks_matrix 119 8.3 0.000 0.000 6.797 6.854 qs_ks_build_kohn_sham_matrix 119 9.3 0.015 0.034 6.797 6.853 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.020 6.075 init_scf_run 11 5.9 0.000 0.002 5.460 5.461 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.460 5.460 qs_ot_get_p 119 10.4 0.002 0.005 4.142 4.205 acc_transpose_blocks 27432 15.5 0.119 0.123 3.867 3.963 sum_up_and_integrate 119 10.3 0.001 0.001 3.900 3.909 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.865 3.909 calculate_rho_elec 119 8.7 0.040 0.046 3.864 3.908 calculate_first_density_matrix 1 7.0 0.000 0.001 3.905 3.907 integrate_v_rspace 119 11.3 0.003 0.003 3.889 3.898 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.268 3.671 init_scf_loop 11 6.9 0.000 0.000 3.660 3.660 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.456 3.459 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.235 3.245 mp_waitall_1 137007 16.6 2.546 3.115 2.546 3.115 make_m2s 4572 13.5 0.055 0.057 2.762 2.866 make_images 4572 14.5 0.274 0.333 2.652 2.753 prepare_preconditioner 11 7.9 0.000 0.000 2.742 2.751 make_preconditioner 11 8.9 0.000 0.000 2.742 2.751 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.354 2.676 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.604 2.634 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.531 2.545 density_rs2pw 119 9.7 0.004 0.004 2.286 2.398 acc_transpose_blocks_kernels 27432 16.5 0.270 0.278 2.333 2.351 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.175 2.296 apply_single 119 13.6 0.000 0.000 2.175 2.296 calculate_dm_sparse 119 9.5 0.000 0.000 2.215 2.269 pw_transfer 1439 11.6 0.065 0.069 2.220 2.265 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.128 2.176 cp_dbcsr_syevd 50 12.0 0.004 0.005 2.102 2.105 jit_kernel_transpose 5 15.6 2.063 2.082 2.063 2.082 ot_diis_step 108 11.5 0.012 0.012 1.915 1.916 grid_integrate_task_list 119 12.3 1.831 1.913 1.831 1.913 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.776 1.787 cp_fm_diag_elpa_base 50 14.0 1.738 1.757 1.772 1.782 fft_wrap_pw1pw2_140 487 13.2 0.245 0.259 1.675 1.724 fft3d_ps 1201 14.6 0.556 0.605 1.648 1.679 potential_pw2rs 119 12.3 0.009 0.009 1.655 1.660 mp_sum_l 7287 12.8 1.249 1.606 1.249 1.606 multiply_cannon_metrocomm3 27432 15.5 0.040 0.041 0.922 1.544 wfi_extrapolate 11 7.9 0.001 0.001 1.491 1.491 acc_transpose_blocks_sync 82296 16.5 1.387 1.486 1.387 1.486 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.392 1.404 transfer_rs2pw 487 10.6 0.004 0.005 1.255 1.373 grid_collocate_task_list 119 9.7 1.249 1.325 1.249 1.325 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.273 1.285 make_images_data 4572 15.5 0.048 0.052 1.149 1.277 hybrid_alltoall_any 4725 16.4 0.065 0.155 1.013 1.173 jit_kernel_multiply 6 15.8 0.345 1.155 0.345 1.155 cp_fm_upper_to_full 72 14.2 0.815 1.154 0.815 1.154 dbcsr_complete_redistribute 329 12.2 0.137 0.172 0.854 1.125 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.099 1.101 mp_alltoall_d11v 2130 13.8 0.919 1.071 0.919 1.071 mp_alltoall_z22v 1201 16.6 0.974 1.008 0.974 1.008 transfer_pw2rs 487 13.2 0.004 0.004 0.979 0.984 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.615 0.879 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.797 0.875 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.868 0.874 cp_fm_cholesky_invert 11 10.9 0.850 0.854 0.850 0.854 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=40.798000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=530.000000, yerr=2.256304 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/04dd4284328246f8f7259e2ea433868052050a67_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 616.488960E+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 0.026 31.865 31.866 qs_mol_dyn_low 1 2.0 0.003 0.003 31.678 31.686 qs_forces 11 3.9 0.002 0.004 31.590 31.593 qs_energies 11 4.9 0.002 0.003 29.862 29.865 scf_env_do_scf 11 5.9 0.000 0.001 23.370 23.370 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 20.681 20.682 velocity_verlet 10 3.0 0.001 0.002 15.563 15.566 dbcsr_multiply_generic 2286 12.5 0.094 0.099 15.352 15.489 multiply_cannon 2286 13.5 0.223 0.228 12.495 12.960 qs_scf_new_mos 108 7.5 0.001 0.001 12.732 12.757 qs_scf_loop_do_ot 108 8.5 0.001 0.001 12.732 12.756 ot_scf_mini 108 9.5 0.002 0.002 11.990 12.019 multiply_cannon_loop 2286 14.5 0.643 0.666 11.529 11.734 ot_mini 108 10.5 0.001 0.001 6.881 6.916 multiply_cannon_multrec 9144 15.5 1.723 1.892 6.144 6.339 rebuild_ks_matrix 119 8.3 0.000 0.000 5.874 5.895 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 5.874 5.894 qs_ot_get_derivative 108 11.5 0.001 0.001 5.550 5.578 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.226 5.244 init_scf_run 11 5.9 0.000 0.001 4.995 4.996 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.995 4.995 dbcsr_mm_accdrv_process 12550 15.8 3.784 4.347 4.311 4.390 calculate_first_density_matrix 1 7.0 0.001 0.001 3.777 3.779 sum_up_and_integrate 119 10.3 0.001 0.001 3.525 3.530 integrate_v_rspace 119 11.3 0.003 0.003 3.515 3.520 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.460 3.466 calculate_rho_elec 119 8.7 0.060 0.061 3.459 3.465 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.223 3.224 qs_ot_get_p 119 10.4 0.002 0.003 3.136 3.185 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.033 3.041 acc_transpose_blocks 9144 15.5 0.041 0.042 3.002 3.040 init_scf_loop 11 6.9 0.000 0.000 2.667 2.668 mp_waitall_1 115863 16.7 1.784 2.319 1.784 2.319 acc_transpose_blocks_kernels 9144 16.5 0.118 0.120 2.223 2.226 make_m2s 4572 13.5 0.035 0.036 1.977 2.157 jit_kernel_transpose 5 15.6 2.105 2.109 2.105 2.109 pw_transfer 1439 11.6 0.065 0.068 2.055 2.067 make_images 4572 14.5 0.271 0.302 1.887 2.065 qs_ot_p2m_diag 50 11.0 0.022 0.024 2.057 2.060 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.962 1.975 grid_integrate_task_list 119 12.3 1.875 1.933 1.875 1.933 density_rs2pw 119 9.7 0.004 0.004 1.832 1.928 prepare_preconditioner 11 7.9 0.000 0.000 1.905 1.910 make_preconditioner 11 8.9 0.000 0.000 1.905 1.910 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.887 1.907 calculate_dm_sparse 119 9.5 0.000 0.000 1.829 1.849 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.786 1.816 cp_dbcsr_syevd 50 12.0 0.005 0.007 1.790 1.791 fft_wrap_pw1pw2_140 487 13.2 0.322 0.331 1.589 1.604 jit_kernel_multiply 8 15.7 0.487 1.544 0.487 1.544 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.461 1.471 cp_fm_diag_elpa_base 50 14.0 1.431 1.450 1.457 1.465 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.426 1.440 fft3d_ps 1201 14.6 0.566 0.575 1.376 1.385 grid_collocate_task_list 119 9.7 1.298 1.379 1.298 1.379 potential_pw2rs 119 12.3 0.010 0.011 1.322 1.325 ot_diis_step 108 11.5 0.013 0.013 1.313 1.313 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.258 1.295 apply_single 119 13.6 0.000 0.000 1.258 1.295 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.246 1.249 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.189 1.197 wfi_extrapolate 11 7.9 0.001 0.001 1.167 1.167 hybrid_alltoall_any 4725 16.4 0.065 0.176 0.865 1.112 make_images_data 4572 15.5 0.042 0.046 0.870 1.086 transfer_rs2pw 487 10.6 0.004 0.005 0.885 0.982 cp_fm_cholesky_invert 11 10.9 0.967 0.971 0.967 0.971 mp_alltoall_d11v 2130 13.8 0.867 0.956 0.867 0.956 multiply_cannon_metrocomm3 9144 15.5 0.020 0.021 0.464 0.928 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.872 0.919 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.893 0.903 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.771 0.773 acc_transpose_blocks_sync 27432 16.5 0.722 0.760 0.722 0.760 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.695 0.746 mp_allgather_i34 2286 14.5 0.283 0.744 0.283 0.744 mp_alltoall_z22v 1201 16.6 0.684 0.715 0.684 0.715 transfer_pw2rs 487 13.2 0.003 0.004 0.711 0.715 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=31.866000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=586.090909, yerr=3.449757 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/04dd4284328246f8f7259e2ea433868052050a67_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 782.372864E+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 0.013 0.032 44.967 44.968 qs_mol_dyn_low 1 2.0 0.003 0.003 44.681 44.690 qs_forces 11 3.9 0.002 0.003 44.072 44.073 qs_energies 11 4.9 0.002 0.003 42.084 42.087 scf_env_do_scf 11 5.9 0.001 0.001 34.652 34.652 scf_env_do_scf_inner_loop 108 6.5 0.008 0.011 26.777 26.778 velocity_verlet 10 3.0 0.003 0.005 24.530 24.604 dbcsr_multiply_generic 2286 12.5 0.102 0.106 20.654 20.845 multiply_cannon 2286 13.5 0.299 0.306 16.712 17.590 qs_scf_new_mos 108 7.5 0.001 0.001 17.332 17.426 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.331 17.425 multiply_cannon_loop 2286 14.5 0.871 0.892 15.458 16.373 ot_scf_mini 108 9.5 0.002 0.002 16.162 16.260 ot_mini 108 10.5 0.001 0.001 9.785 9.899 multiply_cannon_multrec 9144 15.5 3.423 4.898 8.865 9.021 init_scf_loop 11 6.9 0.000 0.000 7.845 7.847 qs_ot_get_derivative 108 11.5 0.001 0.001 7.747 7.845 rebuild_ks_matrix 119 8.3 0.000 0.000 7.129 7.268 qs_ks_build_kohn_sham_matrix 119 9.3 0.014 0.015 7.129 7.267 prepare_preconditioner 11 7.9 0.000 0.000 6.862 6.875 make_preconditioner 11 8.9 0.000 0.000 6.862 6.875 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.379 6.755 dbcsr_mm_accdrv_process 12550 15.8 4.570 6.587 5.304 6.691 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.434 6.560 init_scf_run 11 5.9 0.000 0.001 5.371 5.371 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.371 5.371 cp_fm_upper_to_full 72 14.2 3.161 4.592 3.161 4.592 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.137 4.206 calculate_rho_elec 119 8.7 0.118 0.121 4.137 4.205 calculate_first_density_matrix 1 7.0 0.001 0.001 3.830 3.897 sum_up_and_integrate 119 10.3 0.001 0.001 3.797 3.804 integrate_v_rspace 119 11.3 0.003 0.003 3.787 3.793 qs_ot_get_p 119 10.4 0.003 0.003 3.443 3.581 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.455 3.456 acc_transpose_blocks 9144 15.5 0.045 0.045 3.362 3.398 mp_waitall_1 94719 16.7 2.379 3.356 2.379 3.356 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.194 3.244 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.848 3.244 dbcsr_complete_redistribute 329 12.2 0.286 0.288 2.028 2.872 pw_transfer 1439 11.6 0.069 0.069 2.618 2.624 make_m2s 4572 13.5 0.038 0.039 2.409 2.556 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.716 2.538 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.520 2.526 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.185 2.494 apply_single 119 13.6 0.000 0.000 2.185 2.493 make_images 4572 14.5 0.356 0.391 2.287 2.434 calculate_dm_sparse 119 9.5 0.000 0.001 2.333 2.396 mp_alltoall_i22 627 13.8 1.460 2.338 1.460 2.338 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.477 2.293 multiply_cannon_metrocomm3 9144 15.5 0.021 0.022 1.358 2.241 acc_transpose_blocks_kernels 9144 16.5 0.121 0.123 2.187 2.194 fft_wrap_pw1pw2_140 487 13.2 0.576 0.580 2.125 2.131 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.074 2.124 density_rs2pw 119 9.7 0.003 0.004 2.089 2.107 grid_integrate_task_list 119 12.3 2.078 2.099 2.078 2.099 jit_kernel_transpose 5 15.6 2.067 2.073 2.067 2.073 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.039 2.042 ot_diis_step 108 11.5 0.014 0.014 2.013 2.014 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.777 1.779 mp_sum_l 7287 12.8 0.988 1.755 0.988 1.755 cp_dbcsr_syevd 50 12.0 0.004 0.005 1.679 1.680 fft3d_ps 1201 14.6 0.594 0.603 1.616 1.619 grid_collocate_task_list 119 9.7 1.521 1.538 1.521 1.538 cp_fm_cholesky_invert 11 10.9 1.429 1.432 1.429 1.432 wfi_extrapolate 11 7.9 0.001 0.001 1.408 1.408 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.327 1.391 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.382 1.383 cp_fm_diag_elpa_base 50 14.0 1.233 1.290 1.380 1.380 hybrid_alltoall_any 4725 16.4 0.090 0.149 1.094 1.336 mp_alltoall_d11v 2130 13.8 1.246 1.331 1.246 1.331 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.310 1.329 potential_pw2rs 119 12.3 0.014 0.014 1.305 1.306 make_images_data 4572 15.5 0.046 0.049 1.031 1.242 acc_transpose_blocks_sync 27432 16.5 1.113 1.149 1.113 1.149 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.105 1.119 jit_kernel_multiply 6 15.8 0.704 1.097 0.704 1.097 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.991 1.041 qs_create_task_list 11 7.9 0.001 0.003 0.938 0.950 generate_qs_task_list 11 8.9 0.372 0.391 0.937 0.949 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.905 0.918 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=44.968000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=734.272727, yerr=16.722825 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/04dd4284328246f8f7259e2ea433868052050a67_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 501.473280E+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 2521294. 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 0.051 0.087 86.673 86.690 qs_mol_dyn_low 1 2.0 0.003 0.007 85.998 86.042 qs_forces 11 3.9 0.003 0.004 85.875 85.878 qs_energies 11 4.9 0.002 0.002 83.002 83.024 scf_env_do_scf 11 5.9 0.000 0.001 72.595 72.597 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 66.744 66.744 dbcsr_multiply_generic 2055 12.4 0.109 0.111 53.374 53.755 qs_scf_new_mos 99 7.5 0.000 0.001 49.041 49.194 qs_scf_loop_do_ot 99 8.5 0.000 0.001 49.040 49.193 ot_scf_mini 99 9.5 0.002 0.002 46.607 46.716 multiply_cannon 2055 13.4 0.187 0.192 44.294 45.392 velocity_verlet 10 3.0 0.002 0.004 44.850 44.851 multiply_cannon_loop 2055 14.4 1.790 1.849 43.180 44.353 ot_mini 99 10.5 0.001 0.001 27.408 27.546 qs_ot_get_derivative 99 11.5 0.001 0.001 20.545 20.675 multiply_cannon_multrec 49320 15.4 11.329 12.019 17.385 18.058 rebuild_ks_matrix 110 8.3 0.000 0.000 14.836 15.036 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.013 14.835 15.036 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.036 13.212 mp_waitall_1 220248 16.4 11.482 12.476 11.482 12.476 qs_ot_get_p 110 10.4 0.001 0.002 10.532 10.647 multiply_cannon_sync_h2d 49320 15.4 9.507 10.152 9.507 10.152 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.923 8.512 init_scf_run 11 5.9 0.000 0.001 8.285 8.285 scf_env_initial_rho_setup 11 6.9 0.001 0.001 8.285 8.285 multiply_cannon_metrocomm3 49320 15.4 0.083 0.087 6.718 8.205 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.292 7.773 apply_single 110 13.6 0.000 0.001 7.292 7.773 sum_up_and_integrate 110 10.3 0.002 0.003 7.277 7.289 integrate_v_rspace 110 11.3 0.003 0.003 7.252 7.269 qs_ot_p2m_diag 48 11.0 0.012 0.019 7.097 7.121 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.753 6.918 calculate_rho_elec 110 8.6 0.020 0.024 6.753 6.917 ot_diis_step 99 11.5 0.006 0.006 6.642 6.642 dbcsr_mm_accdrv_process 87628 16.1 3.064 3.168 5.925 6.206 cp_dbcsr_syevd 48 12.0 0.003 0.004 6.167 6.167 init_scf_loop 11 6.9 0.000 0.000 5.822 5.822 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.605 5.669 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.503 5.530 cp_fm_diag_elpa_base 48 14.0 5.487 5.515 5.500 5.528 mp_sum_l 6594 12.7 4.370 5.215 4.370 5.215 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 4.880 4.884 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.505 4.563 make_m2s 4110 13.4 0.062 0.066 4.268 4.408 make_images 4110 14.4 0.177 0.191 4.171 4.312 wfi_extrapolate 11 7.9 0.001 0.001 4.206 4.206 calculate_first_density_matrix 1 7.0 0.000 0.001 3.960 3.965 calculate_dm_sparse 110 9.5 0.001 0.001 3.821 3.934 acc_transpose_blocks 49320 15.4 0.213 0.221 3.819 3.904 density_rs2pw 110 9.6 0.004 0.005 3.710 3.896 multiply_cannon_metrocomm1 49320 15.4 0.063 0.065 2.645 3.650 prepare_preconditioner 11 7.9 0.000 0.000 3.621 3.643 make_preconditioner 11 8.9 0.000 0.000 3.620 3.643 pw_transfer 1331 11.6 0.055 0.067 3.412 3.503 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.388 3.437 grid_integrate_task_list 110 12.3 3.250 3.417 3.250 3.417 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 3.324 3.414 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.331 3.400 potential_pw2rs 110 12.3 0.005 0.006 2.837 2.863 fft_wrap_pw1pw2_140 451 13.1 0.390 0.431 2.762 2.857 mp_alltoall_d11v 2046 13.8 2.278 2.734 2.278 2.734 fft3d_ps 1111 14.6 0.797 0.890 2.619 2.686 jit_kernel_multiply 13 15.9 2.582 2.612 2.582 2.612 acc_transpose_blocks_kernels 49320 16.4 0.422 0.440 2.301 2.342 grid_collocate_task_list 110 9.6 2.160 2.290 2.160 2.290 transfer_rs2pw 451 10.6 0.005 0.006 2.099 2.256 mp_waitany 14300 13.8 1.865 2.142 1.865 2.142 make_images_data 4110 15.4 0.044 0.047 1.910 2.066 mp_sum_d 3889 11.9 1.399 1.985 1.399 1.985 cp_fm_cholesky_invert 11 10.9 1.980 1.984 1.980 1.984 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.932 1.955 hybrid_alltoall_any 4261 16.3 0.085 0.486 1.657 1.949 transfer_pw2rs 451 13.1 0.006 0.007 1.893 1.910 jit_kernel_transpose 5 15.6 1.879 1.902 1.879 1.902 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=86.690000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=475.545455, yerr=2.535696 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/04dd4284328246f8f7259e2ea433868052050a67_performance_tests/10/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 390.715586E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 5019072 0.0% 0.0% 100.0% average stack size 0.0 0.0 196.1 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 592.048128E+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 1935295. 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 0.035 0.121 71.076 71.091 qs_mol_dyn_low 1 2.0 0.005 0.017 70.599 70.639 qs_forces 11 3.9 0.004 0.011 70.502 70.502 qs_energies 11 4.9 0.003 0.016 67.167 67.171 scf_env_do_scf 11 5.9 0.000 0.001 57.155 57.158 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 49.521 49.522 dbcsr_multiply_generic 2055 12.4 0.116 0.121 39.455 39.655 velocity_verlet 10 3.0 0.002 0.003 36.270 36.272 multiply_cannon 2055 13.4 0.224 0.245 32.780 33.973 qs_scf_new_mos 99 7.5 0.001 0.001 33.406 33.549 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.406 33.548 multiply_cannon_loop 2055 14.4 1.169 1.191 31.490 32.265 ot_scf_mini 99 9.5 0.003 0.003 31.735 31.872 ot_mini 99 10.5 0.001 0.001 18.421 18.556 multiply_cannon_multrec 24660 15.4 6.974 8.696 14.055 15.705 rebuild_ks_matrix 110 8.3 0.000 0.000 13.670 13.773 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.016 13.670 13.773 qs_ot_get_derivative 99 11.5 0.001 0.001 12.654 12.795 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.033 12.124 mp_waitall_1 176588 16.5 7.609 10.110 7.609 10.110 multiply_cannon_metrocomm3 24660 15.4 0.072 0.076 5.105 7.812 init_scf_run 11 5.9 0.000 0.002 7.601 7.601 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.601 7.601 init_scf_loop 11 6.9 0.000 0.000 7.597 7.597 multiply_cannon_sync_h2d 24660 15.4 6.382 7.255 6.382 7.255 dbcsr_mm_accdrv_process 52282 16.1 5.503 6.447 6.910 7.225 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.444 7.058 apply_single 110 13.6 0.000 0.001 6.444 7.058 qs_ot_get_p 110 10.4 0.002 0.003 6.629 6.778 sum_up_and_integrate 110 10.3 0.002 0.003 6.476 6.490 integrate_v_rspace 110 11.3 0.003 0.006 6.448 6.463 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.038 6.051 calculate_rho_elec 110 8.6 0.039 0.047 6.038 6.050 ot_diis_step 99 11.5 0.010 0.011 5.719 5.719 prepare_preconditioner 11 7.9 0.000 0.000 5.566 5.591 make_preconditioner 11 8.9 0.000 0.000 5.566 5.591 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.752 5.435 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.142 5.307 make_m2s 4110 13.4 0.057 0.059 4.310 4.785 qs_ot_p2m_diag 48 11.0 0.029 0.045 4.694 4.715 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 4.697 4.699 make_images 4110 14.4 0.409 0.472 4.200 4.673 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.335 4.373 cp_dbcsr_syevd 48 12.0 0.004 0.004 4.224 4.225 calculate_first_density_matrix 1 7.0 0.001 0.002 3.950 3.952 pw_transfer 1331 11.6 0.066 0.074 3.592 3.734 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.636 3.648 cp_fm_diag_elpa_base 48 14.0 3.586 3.609 3.633 3.645 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.485 3.631 wfi_extrapolate 11 7.9 0.001 0.001 3.554 3.554 acc_transpose_blocks 24660 15.4 0.116 0.120 3.412 3.439 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.313 3.391 density_rs2pw 110 9.6 0.004 0.005 3.248 3.380 grid_integrate_task_list 110 12.3 3.164 3.337 3.164 3.337 fft_wrap_pw1pw2_140 451 13.1 0.459 0.474 2.954 3.105 calculate_dm_sparse 110 9.5 0.001 0.001 2.975 3.006 make_images_data 4110 15.4 0.049 0.052 2.346 2.850 hybrid_alltoall_any 4261 16.3 0.107 0.458 2.034 2.816 fft3d_ps 1111 14.6 1.112 1.336 2.582 2.737 cp_fm_cholesky_invert 11 10.9 2.590 2.597 2.590 2.597 mp_sum_l 6594 12.7 1.840 2.478 1.840 2.478 acc_transpose_blocks_kernels 24660 16.4 0.319 0.329 2.340 2.355 potential_pw2rs 110 12.3 0.008 0.009 2.336 2.350 grid_collocate_task_list 110 9.6 2.172 2.319 2.172 2.319 jit_kernel_transpose 5 15.6 2.021 2.029 2.021 2.029 mp_alltoall_d11v 2046 13.8 1.793 2.017 1.793 2.017 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.970 1.990 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 1.867 1.868 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.794 1.802 transfer_rs2pw 451 10.6 0.007 0.007 1.535 1.690 mp_allgather_i34 2055 14.4 0.600 1.639 0.600 1.639 jit_kernel_multiply 8 16.4 1.046 1.600 1.046 1.600 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.541 1.552 multiply_cannon_metrocomm4 22605 15.4 0.079 0.084 0.773 1.541 build_core_hamiltonian_matrix_ 11 4.9 0.002 0.012 1.385 1.484 dbcsr_complete_redistribute 325 12.2 0.255 0.315 1.172 1.443 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=71.091000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=560.272727, yerr=6.620779 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/04dd4284328246f8f7259e2ea433868052050a67_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 669.368320E+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 0.010 0.024 62.407 62.409 qs_mol_dyn_low 1 2.0 0.003 0.003 62.019 62.029 qs_forces 11 3.9 0.003 0.006 61.927 61.930 qs_energies 11 4.9 0.002 0.004 58.703 58.709 scf_env_do_scf 11 5.9 0.000 0.001 50.409 50.409 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 41.454 41.455 velocity_verlet 10 3.0 0.002 0.003 33.675 33.677 dbcsr_multiply_generic 2055 12.4 0.108 0.114 29.884 30.231 qs_scf_new_mos 99 7.5 0.001 0.001 26.527 26.647 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.526 26.646 ot_scf_mini 99 9.5 0.002 0.003 25.235 25.372 multiply_cannon 2055 13.4 0.213 0.226 23.281 24.504 multiply_cannon_loop 2055 14.4 0.818 0.858 22.135 23.231 ot_mini 99 10.5 0.001 0.001 14.432 14.582 rebuild_ks_matrix 110 8.3 0.000 0.000 12.338 12.511 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.015 12.338 12.511 multiply_cannon_multrec 16440 15.4 3.550 4.592 10.148 11.106 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.866 11.023 mp_waitall_1 139946 16.5 7.562 10.591 7.562 10.591 qs_ot_get_derivative 99 11.5 0.001 0.001 9.953 10.096 init_scf_loop 11 6.9 0.000 0.000 8.917 8.919 multiply_cannon_metrocomm3 16440 15.4 0.046 0.048 4.740 7.640 prepare_preconditioner 11 7.9 0.000 0.000 7.120 7.153 make_preconditioner 11 8.9 0.000 0.000 7.120 7.153 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.407 6.785 dbcsr_mm_accdrv_process 34862 16.1 5.672 6.238 6.445 6.660 sum_up_and_integrate 110 10.3 0.001 0.002 6.441 6.455 integrate_v_rspace 110 11.3 0.003 0.003 6.415 6.428 qs_ot_get_p 110 10.4 0.002 0.003 5.853 6.011 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.910 5.922 calculate_rho_elec 110 8.6 0.059 0.059 5.910 5.921 init_scf_run 11 5.9 0.000 0.001 5.843 5.843 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.843 5.843 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.921 5.418 apply_single 110 13.6 0.000 0.000 4.921 5.418 make_m2s 4110 13.4 0.050 0.052 4.340 4.668 make_images 4110 14.4 0.398 0.522 4.224 4.551 ot_diis_step 99 11.5 0.011 0.011 4.444 4.444 qs_ot_p2m_diag 48 11.0 0.042 0.044 4.081 4.089 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.456 4.044 multiply_cannon_sync_h2d 16440 15.4 3.218 3.739 3.218 3.739 pw_transfer 1331 11.6 0.066 0.072 3.690 3.701 cp_dbcsr_syevd 48 12.0 0.003 0.004 3.681 3.682 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.583 3.597 grid_integrate_task_list 110 12.3 3.195 3.435 3.195 3.435 density_rs2pw 110 9.6 0.004 0.005 3.123 3.311 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.121 3.129 cp_fm_diag_elpa_base 48 14.0 3.050 3.081 3.119 3.127 wfi_extrapolate 11 7.9 0.001 0.001 3.093 3.093 fft_wrap_pw1pw2_140 451 13.1 0.578 0.585 3.046 3.062 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.972 2.973 cp_fm_cholesky_invert 11 10.9 2.898 2.904 2.898 2.904 make_images_data 4110 15.4 0.046 0.050 2.449 2.848 hybrid_alltoall_any 4261 16.3 0.109 0.383 2.167 2.815 calculate_dm_sparse 110 9.5 0.001 0.001 2.687 2.719 calculate_first_density_matrix 1 7.0 0.000 0.001 2.653 2.654 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.576 2.642 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.544 2.619 multiply_cannon_metrocomm4 14385 15.4 0.049 0.053 0.902 2.572 fft3d_ps 1111 14.6 1.091 1.101 2.526 2.535 mp_sum_l 6594 12.7 1.640 2.445 1.640 2.445 mp_irecv_dv 48980 15.7 0.827 2.439 0.827 2.439 grid_collocate_task_list 110 9.6 2.219 2.394 2.219 2.394 potential_pw2rs 110 12.3 0.011 0.012 2.286 2.293 mp_alltoall_d11v 2046 13.8 1.880 2.190 1.880 2.190 dbcsr_complete_redistribute 325 12.2 0.354 0.375 1.545 2.010 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.999 2.002 cp_fm_upper_to_full 70 14.2 1.422 1.811 1.422 1.811 acc_transpose_blocks 16440 15.4 0.078 0.080 1.641 1.692 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.667 1.684 cp_fm_cholesky_decompose 22 10.9 1.642 1.664 1.642 1.664 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.590 1.605 transfer_rs2pw 451 10.6 0.005 0.006 1.386 1.586 mp_allgather_i34 2055 14.4 0.459 1.551 0.459 1.551 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.072 1.524 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.363 1.466 mp_waitany 17072 13.8 1.196 1.433 1.196 1.433 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.386 1.395 transfer_pw2rs 451 13.1 0.005 0.005 1.291 1.299 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=62.409000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=634.818182, yerr=8.684555 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/04dd4284328246f8f7259e2ea433868052050a67_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 731.783168E+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 0.009 0.030 66.329 66.330 qs_mol_dyn_low 1 2.0 0.003 0.003 66.089 66.100 qs_forces 11 3.9 0.003 0.003 66.007 66.009 qs_energies 11 4.9 0.001 0.003 62.600 62.605 scf_env_do_scf 11 5.9 0.000 0.001 54.279 54.281 scf_env_do_scf_inner_loop 99 6.5 0.003 0.006 42.639 42.641 velocity_verlet 10 3.0 0.002 0.003 37.462 37.465 dbcsr_multiply_generic 2055 12.4 0.115 0.121 31.025 31.243 qs_scf_new_mos 99 7.5 0.001 0.001 27.978 28.083 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.978 28.082 ot_scf_mini 99 9.5 0.003 0.003 26.292 26.402 multiply_cannon 2055 13.4 0.238 0.259 23.728 24.963 multiply_cannon_loop 2055 14.4 1.404 1.483 22.350 22.958 ot_mini 99 10.5 0.001 0.001 15.185 15.320 multiply_cannon_multrec 24660 15.4 4.119 6.711 13.067 14.227 rebuild_ks_matrix 110 8.3 0.000 0.000 12.033 12.157 qs_ks_build_kohn_sham_matrix 110 9.3 0.016 0.019 12.033 12.157 init_scf_loop 11 6.9 0.000 0.000 11.600 11.601 qs_ot_get_derivative 99 11.5 0.001 0.001 11.019 11.129 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.611 10.724 prepare_preconditioner 11 7.9 0.000 0.000 9.845 9.859 make_preconditioner 11 8.9 0.000 0.000 9.845 9.859 dbcsr_mm_accdrv_process 52304 16.0 7.644 9.099 8.790 9.746 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.119 9.527 mp_waitall_1 121746 16.5 4.588 6.298 4.588 6.298 sum_up_and_integrate 110 10.3 0.001 0.002 6.282 6.294 integrate_v_rspace 110 11.3 0.003 0.003 6.256 6.268 qs_ot_get_p 110 10.4 0.009 0.013 5.946 6.106 make_m2s 4110 13.4 0.060 0.062 5.655 5.953 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.861 5.871 calculate_rho_elec 110 8.6 0.078 0.081 5.860 5.870 make_images 4110 14.4 0.580 0.701 5.513 5.805 init_scf_run 11 5.9 0.000 0.001 5.564 5.564 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.564 5.564 cp_fm_upper_to_full 70 14.2 3.281 4.732 3.281 4.732 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.048 4.146 apply_single 110 13.6 0.000 0.000 4.048 4.146 ot_diis_step 99 11.5 0.011 0.012 4.126 4.127 qs_ot_p2m_diag 48 11.0 0.055 0.064 4.039 4.058 dbcsr_complete_redistribute 325 12.2 0.423 0.479 2.595 3.714 pw_transfer 1331 11.6 0.066 0.074 3.670 3.705 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.549 3.603 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.563 3.601 cp_dbcsr_syevd 48 12.0 0.003 0.004 3.473 3.481 grid_integrate_task_list 110 12.3 3.271 3.461 3.271 3.461 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.377 3.438 multiply_cannon_metrocomm3 24660 15.4 0.038 0.040 1.482 3.251 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.083 3.188 make_images_data 4110 15.4 0.049 0.053 2.861 3.185 hybrid_alltoall_any 4261 16.3 0.123 0.460 2.397 3.174 density_rs2pw 110 9.6 0.004 0.004 2.951 3.113 fft_wrap_pw1pw2_140 451 13.1 0.610 0.633 3.051 3.090 wfi_extrapolate 11 7.9 0.001 0.001 3.027 3.027 calculate_dm_sparse 110 9.5 0.001 0.001 2.985 3.022 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.927 2.936 cp_fm_diag_elpa_base 48 14.0 2.778 2.836 2.925 2.934 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.854 2.856 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.717 2.808 cp_fm_cholesky_invert 11 10.9 2.741 2.749 2.741 2.749 mp_alltoall_i22 605 13.7 1.597 2.726 1.597 2.726 acc_transpose_blocks 24660 15.4 0.114 0.118 2.418 2.581 multiply_cannon_sync_h2d 24660 15.4 2.358 2.564 2.358 2.564 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.473 2.519 grid_collocate_task_list 110 9.6 2.265 2.479 2.265 2.479 fft3d_ps 1111 14.6 1.083 1.119 2.445 2.461 calculate_first_density_matrix 1 7.0 0.000 0.001 2.444 2.445 qs_energies_init_hamiltonians 11 5.9 0.001 0.011 2.260 2.262 mp_alltoall_d11v 2046 13.8 1.855 2.117 1.855 2.117 potential_pw2rs 110 12.3 0.012 0.013 2.078 2.081 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.803 1.834 cp_fm_cholesky_decompose 22 10.9 1.708 1.754 1.708 1.754 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.608 1.704 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.663 1.676 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.542 1.562 mp_sum_l 6594 12.7 0.956 1.559 0.956 1.559 acc_transpose_blocks_sync 73980 16.4 1.386 1.549 1.386 1.549 multiply_cannon_metrocomm4 20550 15.4 0.062 0.065 0.856 1.471 jit_kernel_multiply 8 15.9 0.800 1.454 0.800 1.454 transfer_rs2pw 451 10.6 0.005 0.006 1.201 1.423 mp_allgather_i34 2055 14.4 0.460 1.412 0.460 1.412 mp_irecv_dv 62702 16.1 0.751 1.392 0.751 1.392 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=66.330000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=693.727273, yerr=8.214078 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/04dd4284328246f8f7259e2ea433868052050a67_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 859.058176E+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 0.027 0.067 55.976 55.987 qs_mol_dyn_low 1 2.0 0.003 0.003 55.696 55.718 qs_forces 11 3.9 0.003 0.003 55.626 55.627 qs_energies 11 4.9 0.001 0.002 51.977 51.982 scf_env_do_scf 11 5.9 0.000 0.001 43.599 43.599 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 35.747 35.748 velocity_verlet 10 3.0 0.010 0.011 31.547 31.550 dbcsr_multiply_generic 2055 12.4 0.106 0.110 23.628 23.747 qs_scf_new_mos 99 7.5 0.001 0.001 21.102 21.140 qs_scf_loop_do_ot 99 8.5 0.001 0.001 21.102 21.139 ot_scf_mini 99 9.5 0.002 0.002 19.845 19.877 multiply_cannon 2055 13.4 0.236 0.243 17.907 19.356 multiply_cannon_loop 2055 14.4 0.604 0.625 16.622 16.962 rebuild_ks_matrix 110 8.3 0.000 0.000 11.648 11.663 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 11.647 11.662 ot_mini 99 10.5 0.001 0.001 11.017 11.041 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.361 10.372 multiply_cannon_multrec 8220 15.4 3.432 4.716 7.919 8.825 init_scf_loop 11 6.9 0.000 0.000 7.804 7.805 mp_waitall_1 103326 16.6 5.823 7.570 5.823 7.570 qs_ot_get_derivative 99 11.5 0.001 0.001 7.328 7.358 sum_up_and_integrate 110 10.3 0.002 0.002 6.246 6.257 integrate_v_rspace 110 11.3 0.003 0.003 6.219 6.231 prepare_preconditioner 11 7.9 0.000 0.000 6.145 6.148 make_preconditioner 11 8.9 0.000 0.000 6.145 6.148 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.061 6.077 calculate_rho_elec 110 8.6 0.114 0.114 6.061 6.076 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.708 5.791 dbcsr_mm_accdrv_process 17442 15.9 3.153 4.880 4.348 5.591 init_scf_run 11 5.9 0.000 0.001 5.224 5.224 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.223 5.224 qs_ot_get_p 110 10.4 0.005 0.007 4.856 4.892 make_m2s 4110 13.4 0.039 0.040 4.349 4.640 make_images 4110 14.4 0.653 0.715 4.219 4.507 multiply_cannon_metrocomm3 8220 15.4 0.019 0.020 2.870 4.230 pw_transfer 1331 11.6 0.066 0.071 3.897 3.904 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.789 3.799 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.731 3.771 apply_single 110 13.6 0.000 0.000 3.731 3.771 ot_diis_step 99 11.5 0.012 0.012 3.665 3.666 grid_integrate_task_list 110 12.3 3.386 3.475 3.386 3.475 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.432 3.439 fft_wrap_pw1pw2_140 451 13.1 0.777 0.786 3.264 3.280 density_rs2pw 110 9.6 0.004 0.004 3.010 3.171 cp_dbcsr_syevd 48 12.0 0.004 0.004 3.073 3.074 cp_fm_cholesky_invert 11 10.9 2.949 2.953 2.949 2.953 hybrid_alltoall_any 4261 16.3 0.200 0.850 2.325 2.870 wfi_extrapolate 11 7.9 0.001 0.001 2.780 2.780 make_images_data 4110 15.4 0.041 0.046 2.341 2.767 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.668 2.669 calculate_dm_sparse 110 9.5 0.001 0.001 2.553 2.589 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.536 2.538 grid_collocate_task_list 110 9.6 2.364 2.536 2.364 2.536 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.514 2.521 cp_fm_diag_elpa_base 48 14.0 2.457 2.485 2.512 2.519 fft3d_ps 1111 14.6 1.135 1.170 2.454 2.468 multiply_cannon_sync_h2d 8220 15.4 2.362 2.453 2.362 2.453 calculate_first_density_matrix 1 7.0 0.001 0.002 2.348 2.348 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.144 2.151 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.079 2.111 potential_pw2rs 110 12.3 0.015 0.015 2.024 2.027 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.780 1.995 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.975 1.987 mp_alltoall_d11v 2046 13.8 1.784 1.867 1.784 1.867 cp_fm_cholesky_decompose 22 10.9 1.753 1.769 1.753 1.769 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.616 1.620 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.503 1.617 mp_allgather_i34 2055 14.4 0.458 1.575 0.458 1.575 dbcsr_complete_redistribute 325 12.2 0.565 0.592 1.484 1.569 jit_kernel_multiply 7 15.7 0.882 1.550 0.882 1.550 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.502 1.512 acc_transpose_blocks 8220 15.4 0.039 0.040 1.434 1.467 transfer_rs2pw 451 10.6 0.005 0.005 1.180 1.334 qs_create_task_list 11 7.9 0.000 0.001 1.220 1.318 generate_qs_task_list 11 8.9 0.378 0.446 1.219 1.318 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.203 1.233 mp_waitany 9240 13.8 1.069 1.228 1.069 1.228 multiply_cannon_metrocomm1 8220 15.4 0.022 0.023 0.746 1.133 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=55.987000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=805.818182, yerr=11.823076 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/04dd4284328246f8f7259e2ea433868052050a67_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 1.403310E+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 0.033 0.085 88.870 88.888 qs_mol_dyn_low 1 2.0 0.003 0.003 88.380 88.519 qs_forces 11 3.9 0.013 0.024 88.284 88.285 qs_energies 11 4.9 0.008 0.036 84.138 84.151 scf_env_do_scf 11 5.9 0.001 0.001 73.892 73.893 velocity_verlet 10 3.0 0.010 0.011 56.603 56.611 scf_env_do_scf_inner_loop 99 6.5 0.004 0.007 45.096 45.098 dbcsr_multiply_generic 2055 12.4 0.121 0.126 30.096 30.223 init_scf_loop 11 6.9 0.000 0.000 28.723 28.726 qs_scf_new_mos 99 7.5 0.001 0.001 27.631 27.678 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.630 27.677 prepare_preconditioner 11 7.9 0.000 0.000 26.705 26.712 make_preconditioner 11 8.9 0.000 0.000 26.704 26.712 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.945 26.168 ot_scf_mini 99 9.5 0.002 0.002 25.806 25.851 multiply_cannon 2055 13.4 0.329 0.350 22.843 23.519 multiply_cannon_loop 2055 14.4 0.826 0.837 21.019 21.343 cp_fm_upper_to_full 70 14.2 12.759 18.229 12.759 18.229 ot_mini 99 10.5 0.001 0.001 14.505 14.551 rebuild_ks_matrix 110 8.3 0.000 0.001 13.719 13.765 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 13.719 13.764 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.442 12.484 dbcsr_complete_redistribute 325 12.2 1.016 1.045 7.366 10.566 multiply_cannon_multrec 8220 15.4 4.074 4.241 9.783 9.887 qs_ot_get_derivative 99 11.5 0.001 0.001 9.830 9.874 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.335 9.537 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.744 8.919 mp_waitall_1 84994 16.7 7.899 8.796 7.899 8.796 mp_alltoall_i22 605 13.7 5.375 8.565 5.375 8.565 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.189 7.221 calculate_rho_elec 110 8.6 0.224 0.225 7.189 7.221 sum_up_and_integrate 110 10.3 0.002 0.002 6.828 6.842 integrate_v_rspace 110 11.3 0.004 0.004 6.800 6.814 qs_ot_get_p 110 10.4 0.003 0.003 5.886 5.948 make_m2s 4110 13.4 0.044 0.044 5.434 5.941 init_scf_run 11 5.9 0.001 0.004 5.850 5.850 scf_env_initial_rho_setup 11 6.9 0.004 0.004 5.849 5.850 dbcsr_mm_accdrv_process 11614 15.7 3.993 4.823 5.560 5.773 make_images 4110 14.4 0.889 0.948 5.244 5.750 cp_fm_cholesky_invert 11 10.9 5.563 5.568 5.563 5.568 multiply_cannon_metrocomm3 8220 15.4 0.019 0.020 5.170 5.508 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.815 5.291 apply_single 110 13.6 0.000 0.000 4.815 5.291 pw_transfer 1331 11.6 0.074 0.075 5.065 5.072 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 4.948 4.956 ot_diis_step 99 11.5 0.015 0.016 4.651 4.651 fft_wrap_pw1pw2_140 451 13.1 1.279 1.283 4.306 4.316 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.236 4.244 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.736 3.747 grid_integrate_task_list 110 12.3 3.685 3.747 3.685 3.747 cp_dbcsr_syevd 48 12.0 0.004 0.004 3.688 3.688 density_rs2pw 110 9.6 0.004 0.004 3.630 3.641 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.105 3.563 hybrid_alltoall_any 4261 16.3 0.263 0.562 2.903 3.559 make_images_data 4110 15.4 0.046 0.049 2.842 3.508 wfi_extrapolate 11 7.9 0.001 0.001 3.391 3.391 calculate_dm_sparse 110 9.5 0.001 0.001 3.203 3.238 multiply_cannon_sync_h2d 8220 15.4 3.122 3.132 3.122 3.132 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.077 3.077 cp_fm_diag_elpa_base 48 14.0 2.522 2.719 3.075 3.075 fft3d_ps 1111 14.6 1.303 1.315 3.002 3.008 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.985 2.988 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.734 2.762 grid_collocate_task_list 110 9.6 2.675 2.687 2.675 2.687 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.446 2.468 calculate_first_density_matrix 1 7.0 0.001 0.001 2.338 2.339 potential_pw2rs 110 12.3 0.021 0.021 2.331 2.335 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.251 2.321 mp_alltoall_d11v 2046 13.8 2.122 2.186 2.122 2.186 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.100 2.179 cp_fm_cholesky_decompose 22 10.9 2.129 2.151 2.129 2.151 qs_create_task_list 11 7.9 0.001 0.003 1.896 1.940 generate_qs_task_list 11 8.9 0.741 0.793 1.895 1.939 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.882 1.886 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.794 1.839 acc_transpose_blocks 8220 15.4 0.041 0.042 1.756 1.790 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=88.888000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1250.909091, yerr=73.624915 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/04dd4284328246f8f7259e2ea433868052050a67_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 632.545280E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175955383376 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4010 57647. MP_Allreduce 11104 795. MP_Sync 87 MP_Alltoall 2226 2754334. 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 0.039 0.071 205.252 205.254 qs_mol_dyn_low 1 2.0 0.003 0.003 203.913 204.766 qs_forces 11 3.9 0.005 0.005 203.799 203.800 qs_energies 11 4.9 0.002 0.002 198.064 198.081 scf_env_do_scf 11 5.9 0.001 0.001 181.359 181.363 scf_env_do_scf_inner_loop 117 6.6 0.003 0.014 160.735 160.738 velocity_verlet 10 3.0 0.002 0.014 124.241 124.243 dbcsr_multiply_generic 2507 12.6 0.181 0.185 123.120 123.764 qs_scf_new_mos 117 7.6 0.001 0.001 120.949 121.187 qs_scf_loop_do_ot 117 8.6 0.001 0.001 120.949 121.186 ot_scf_mini 117 9.6 0.003 0.003 114.387 114.648 multiply_cannon 2507 13.6 0.242 0.252 101.114 103.192 multiply_cannon_loop 2507 14.6 2.314 2.400 98.901 100.745 ot_mini 117 10.6 0.001 0.001 64.242 64.467 multiply_cannon_multrec 60168 15.6 32.465 34.370 41.972 43.845 qs_ot_get_derivative 117 11.6 0.001 0.001 39.168 39.406 rebuild_ks_matrix 128 8.3 0.001 0.001 33.909 34.234 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 33.909 34.234 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.473 30.742 mp_waitall_1 267128 16.5 27.499 30.525 27.499 30.525 multiply_cannon_sync_h2d 60168 15.6 27.322 29.454 27.322 29.454 qs_ot_get_p 128 10.4 0.001 0.003 28.239 28.508 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.619 25.361 apply_single 128 13.6 0.001 0.001 24.619 25.361 ot_diis_step 117 11.6 0.008 0.008 24.845 24.847 qs_ot_p2m_diag 83 11.4 0.079 0.091 21.541 21.617 init_scf_loop 11 6.9 0.000 0.000 20.541 20.543 cp_dbcsr_syevd 83 12.4 0.005 0.006 19.150 19.151 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 16.925 17.093 multiply_cannon_metrocomm3 60168 15.6 0.121 0.126 15.086 16.778 cp_fm_diag_elpa 83 13.4 0.000 0.000 16.121 16.152 cp_fm_diag_elpa_base 83 14.4 16.038 16.084 16.116 16.145 prepare_preconditioner 11 7.9 0.000 0.000 15.877 15.917 make_preconditioner 11 8.9 0.000 0.000 15.877 15.917 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.112 15.272 make_m2s 5014 13.6 0.106 0.112 14.055 14.425 make_images 5014 14.6 0.398 0.416 13.871 14.245 sum_up_and_integrate 128 10.3 0.002 0.005 14.159 14.173 integrate_v_rspace 128 11.3 0.003 0.004 14.100 14.118 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.288 13.367 calculate_rho_elec 128 8.7 0.045 0.064 13.288 13.367 init_scf_run 11 5.9 0.000 0.002 12.493 12.493 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.493 12.493 dbcsr_mm_accdrv_process 124484 16.2 4.425 4.762 9.048 9.760 wfi_extrapolate 11 7.9 0.001 0.001 9.221 9.222 cp_fm_cholesky_invert 11 10.9 9.028 9.035 9.028 9.035 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.637 8.807 calculate_dm_sparse 128 9.5 0.001 0.001 8.429 8.537 mp_sum_l 7950 12.9 6.690 7.980 6.690 7.980 pw_transfer 1547 11.6 0.075 0.093 7.749 7.926 make_images_data 5014 15.6 0.067 0.074 6.892 7.918 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.767 7.885 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 7.547 7.722 density_rs2pw 128 9.7 0.006 0.007 6.960 7.572 hybrid_alltoall_any 5200 16.5 0.296 2.256 6.032 7.522 multiply_cannon_metrocomm1 60168 15.6 0.094 0.099 5.720 7.457 grid_integrate_task_list 128 12.3 7.038 7.379 7.038 7.379 fft_wrap_pw1pw2_140 523 13.2 1.126 1.169 6.563 6.734 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.691 6.701 fft3d_ps 1291 14.7 2.191 2.841 5.552 5.885 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.790 5.854 mp_alltoall_d11v 2415 14.1 4.369 5.416 4.369 5.416 grid_collocate_task_list 128 9.7 4.846 5.269 4.846 5.269 cp_fm_cholesky_decompose 22 10.9 4.679 4.694 4.679 4.694 potential_pw2rs 128 12.3 0.009 0.010 4.641 4.659 mp_sum_d 4474 12.1 3.536 4.194 3.536 4.194 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=205.254000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=598.272727, yerr=6.579457 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/04dd4284328246f8f7259e2ea433868052050a67_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 839.327744E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100943E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834845264 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 58214. MP_Allreduce 11083 960. MP_Sync 87 MP_Alltoall 1969 6213081. MP_SendRecv 12032 47072. MP_ISendRecv 12032 47072. MP_Wait 25916 MP_ISend 11748 212467. MP_IRecv 11748 212467. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.030 189.944 189.945 qs_mol_dyn_low 1 2.0 0.003 0.003 189.579 189.595 qs_forces 11 3.9 0.004 0.005 189.073 189.075 qs_energies 11 4.9 0.002 0.002 182.307 182.317 scf_env_do_scf 11 5.9 0.001 0.001 165.892 165.903 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 132.927 132.929 velocity_verlet 10 3.0 0.002 0.002 119.917 119.928 dbcsr_multiply_generic 2507 12.6 0.190 0.196 97.551 98.747 qs_scf_new_mos 117 7.6 0.001 0.001 94.229 94.738 qs_scf_loop_do_ot 117 8.6 0.001 0.001 94.228 94.737 ot_scf_mini 117 9.6 0.004 0.004 89.411 90.078 multiply_cannon 2507 13.6 0.503 0.561 77.349 82.115 multiply_cannon_loop 2507 14.6 1.581 1.649 73.917 76.754 ot_mini 117 10.6 0.001 0.001 50.291 50.911 mp_waitall_1 214728 16.6 24.155 38.991 24.155 38.991 multiply_cannon_multrec 30084 15.6 20.951 25.795 31.547 36.704 rebuild_ks_matrix 128 8.3 0.001 0.001 32.422 33.006 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.019 32.421 33.005 init_scf_loop 11 6.9 0.000 0.000 32.876 32.877 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.141 29.662 qs_ot_get_derivative 117 11.6 0.001 0.002 28.353 29.016 multiply_cannon_metrocomm3 30084 15.6 0.097 0.104 15.650 28.813 prepare_preconditioner 11 7.9 0.000 0.000 28.534 28.608 make_preconditioner 11 8.9 0.000 0.000 28.534 28.608 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.184 27.730 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.038 23.204 apply_single 128 13.6 0.001 0.001 22.038 23.204 qs_ot_get_p 128 10.4 0.002 0.003 21.118 21.918 ot_diis_step 117 11.6 0.014 0.015 21.759 21.761 multiply_cannon_sync_h2d 30084 15.6 17.976 20.522 17.976 20.522 cp_fm_cholesky_invert 11 10.9 16.613 16.625 16.613 16.625 qs_ot_p2m_diag 83 11.4 0.189 0.217 16.169 16.204 make_m2s 5014 13.6 0.090 0.095 14.252 15.988 make_images 5014 14.6 1.172 1.374 14.040 15.776 cp_dbcsr_syevd 83 12.4 0.006 0.008 14.963 14.964 sum_up_and_integrate 128 10.3 0.002 0.004 14.077 14.106 integrate_v_rspace 128 11.3 0.003 0.003 14.017 14.048 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.438 13.471 calculate_rho_elec 128 8.7 0.087 0.104 13.437 13.470 cp_fm_diag_elpa 83 13.4 0.000 0.000 11.785 11.817 cp_fm_diag_elpa_base 83 14.4 11.524 11.626 11.778 11.807 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 11.261 11.728 init_scf_run 11 5.9 0.000 0.002 11.609 11.610 scf_env_initial_rho_setup 11 6.9 0.001 0.002 11.608 11.610 multiply_cannon_metrocomm4 27577 15.6 0.106 0.120 3.898 10.841 dbcsr_mm_accdrv_process 62242 16.2 5.453 6.394 10.052 10.610 make_images_data 5014 15.6 0.068 0.076 8.416 10.469 mp_irecv_dv 69486 16.3 3.692 10.437 3.692 10.437 hybrid_alltoall_any 5200 16.5 0.351 1.495 7.206 10.175 pw_transfer 1547 11.6 0.085 0.095 8.635 8.690 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.411 8.470 wfi_extrapolate 11 7.9 0.001 0.001 8.426 8.426 grid_integrate_task_list 128 12.3 7.149 7.580 7.149 7.580 density_rs2pw 128 9.7 0.006 0.007 7.117 7.522 fft_wrap_pw1pw2_140 523 13.2 1.212 1.238 7.427 7.498 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.281 7.090 cp_fm_cholesky_decompose 22 10.9 6.965 7.043 6.965 7.043 calculate_dm_sparse 128 9.5 0.001 0.001 6.477 6.633 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.158 6.167 mp_sum_l 7950 12.9 4.147 6.057 4.147 6.057 fft3d_ps 1291 14.7 2.846 3.023 5.973 6.023 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.375 5.441 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.264 5.411 grid_collocate_task_list 128 9.7 5.004 5.368 5.004 5.368 mp_allgather_i34 2507 14.6 1.961 5.135 1.961 5.135 mp_alltoall_d11v 2415 14.1 4.249 4.934 4.249 4.934 potential_pw2rs 128 12.3 0.016 0.019 4.548 4.560 dbcsr_complete_redistribute 395 12.7 0.772 0.864 3.220 4.093 mp_sum_d 4470 12.1 2.617 3.966 2.617 3.966 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=189.945000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=800.181818, yerr=1.192262 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/04dd4284328246f8f7259e2ea433868052050a67_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430457200640 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986252263424 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992001093632 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613077360640 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239162695680 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239162695680 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.232994E+12 0.0% 0.0% 100.0% flops max/rank 2.951022E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806382528 0.0% 0.0% 100.0% number of processed stacks 4017216 0.0% 0.0% 100.0% average stack size 0.0 0.0 1694.3 marketing flops 145.650931E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 963.178496E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1052064 MPI messages size (bytes): total size 2.737208E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.601750E+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 284544 37295751168 131072 < size <= 4194304 665856 1004519030784 4194304 < size <= 16777216 66080 937890532432 16777216 < size 28896 757491302400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 57629. MP_Allreduce 11184 996. MP_Sync 88 MP_Alltoall 1724 9394234. MP_SendRecv 7998 75008. MP_ISendRecv 7998 75008. MP_Wait 21986 MP_ISend 11836 275177. MP_IRecv 11836 275177. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.030 0.052 186.187 186.192 qs_mol_dyn_low 1 2.0 0.003 0.003 185.689 185.749 qs_forces 11 3.9 0.004 0.005 185.582 185.585 qs_energies 11 4.9 0.009 0.013 178.919 178.929 scf_env_do_scf 11 5.9 0.001 0.001 162.740 162.741 scf_env_do_scf_inner_loop 118 6.6 0.003 0.009 124.688 124.689 velocity_verlet 10 3.0 0.001 0.002 120.124 120.127 dbcsr_multiply_generic 2529 12.6 0.183 0.190 85.069 86.372 qs_scf_new_mos 118 7.6 0.001 0.001 85.788 86.132 qs_scf_loop_do_ot 118 8.6 0.001 0.001 85.788 86.131 ot_scf_mini 118 9.6 0.003 0.004 81.327 81.729 multiply_cannon 2529 13.6 0.513 0.547 63.911 67.992 multiply_cannon_loop 2529 14.6 1.143 1.212 60.639 63.691 ot_mini 118 10.6 0.001 0.001 44.722 45.140 init_scf_loop 11 6.9 0.000 0.000 37.942 37.943 mp_waitall_1 172006 16.6 26.855 37.036 26.855 37.036 prepare_preconditioner 11 7.9 0.000 0.000 33.806 33.863 make_preconditioner 11 8.9 0.000 0.000 33.806 33.863 make_full_inverse_cholesky 11 9.9 0.000 0.000 31.353 32.796 rebuild_ks_matrix 129 8.3 0.001 0.001 31.259 31.795 qs_ks_build_kohn_sham_matrix 129 9.3 0.016 0.018 31.259 31.795 qs_ks_update_qs_env 129 7.6 0.001 0.001 28.194 28.695 multiply_cannon_metrocomm3 20232 15.6 0.063 0.066 16.563 26.926 multiply_cannon_multrec 20232 15.6 13.024 16.087 22.409 25.497 qs_ot_get_derivative 118 11.6 0.002 0.002 24.552 24.959 qs_ot_get_p 129 10.4 0.002 0.003 21.524 22.061 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 20.171 21.338 apply_single 129 13.6 0.001 0.001 20.171 21.338 ot_diis_step 118 11.6 0.018 0.019 20.035 20.035 qs_ot_p2m_diag 84 11.4 0.269 0.275 16.898 16.917 make_m2s 5058 13.6 0.081 0.086 15.530 16.655 make_images 5058 14.6 1.162 1.268 15.294 16.419 cp_fm_cholesky_invert 11 10.9 16.165 16.176 16.165 16.176 cp_dbcsr_syevd 84 12.4 0.006 0.007 15.630 15.639 multiply_cannon_sync_h2d 20232 15.6 13.524 15.274 13.524 15.274 sum_up_and_integrate 129 10.3 0.002 0.003 14.660 14.684 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.568 14.626 calculate_rho_elec 129 8.7 0.131 0.146 14.567 14.626 integrate_v_rspace 129 11.3 0.003 0.004 14.599 14.625 cp_fm_diag_elpa 84 13.4 0.000 0.000 12.188 12.211 cp_fm_diag_elpa_base 84 14.4 11.727 11.931 12.184 12.206 init_scf_run 11 5.9 0.000 0.002 11.016 11.017 scf_env_initial_rho_setup 11 6.9 0.001 0.003 11.016 11.016 make_images_data 5058 15.6 0.063 0.073 9.331 10.978 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.899 10.206 hybrid_alltoall_any 5245 16.5 0.455 2.070 8.055 9.916 multiply_cannon_metrocomm4 17703 15.6 0.068 0.078 3.550 9.679 pw_transfer 1559 11.6 0.086 0.104 9.390 9.501 mp_irecv_dv 50659 16.2 3.421 9.418 3.421 9.418 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 9.164 9.283 dbcsr_mm_accdrv_process 41846 16.2 5.668 5.964 8.854 8.996 cp_fm_cholesky_decompose 22 10.9 8.233 8.303 8.233 8.303 fft_wrap_pw1pw2_140 527 13.2 1.313 1.353 7.968 8.094 density_rs2pw 129 9.7 0.006 0.006 7.696 8.094 wfi_extrapolate 11 7.9 0.001 0.001 7.825 7.825 grid_integrate_task_list 129 12.3 7.338 7.740 7.338 7.740 cp_fm_upper_to_full 106 14.8 5.884 7.478 5.884 7.478 dbcsr_complete_redistribute 397 12.7 1.169 1.196 5.015 6.837 fft3d_ps 1301 14.7 2.786 3.018 6.518 6.606 calculate_dm_sparse 129 9.5 0.001 0.001 6.101 6.227 grid_collocate_task_list 129 9.7 5.197 5.644 5.197 5.644 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.614 5.618 mp_alltoall_d11v 2429 14.1 4.941 5.535 4.941 5.535 copy_fm_to_dbcsr 210 11.7 0.002 0.002 3.694 5.520 mp_sum_l 8016 12.9 3.782 5.518 3.782 5.518 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.828 5.516 mp_allgather_i34 2529 14.6 1.763 5.189 1.763 5.189 potential_pw2rs 129 12.3 0.020 0.022 4.913 4.933 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.606 4.756 transfer_fm_to_dbcsr 11 9.9 0.019 0.024 2.434 4.248 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.149 4.189 mp_alltoall_i22 720 14.1 2.089 4.065 2.089 4.065 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.792 3.793 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=186.192000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=908.545455, yerr=10.236864 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/04dd4284328246f8f7259e2ea433868052050a67_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 1.147785E+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 0.015 0.042 192.123 192.133 qs_mol_dyn_low 1 2.0 0.003 0.003 191.739 191.755 qs_forces 11 3.9 0.004 0.004 191.626 191.632 qs_energies 11 4.9 0.002 0.002 184.502 184.511 scf_env_do_scf 11 5.9 0.001 0.001 167.623 167.631 velocity_verlet 10 3.0 0.002 0.003 127.546 127.549 scf_env_do_scf_inner_loop 117 6.6 0.004 0.009 120.176 120.177 qs_scf_new_mos 117 7.6 0.001 0.001 83.305 83.685 qs_scf_loop_do_ot 117 8.6 0.001 0.001 83.305 83.684 dbcsr_multiply_generic 2507 12.6 0.189 0.197 81.248 81.838 ot_scf_mini 117 9.6 0.003 0.003 78.707 79.043 multiply_cannon 2507 13.6 0.546 0.578 55.710 58.313 multiply_cannon_loop 2507 14.6 1.830 1.935 52.057 53.609 init_scf_loop 11 6.9 0.000 0.000 47.315 47.316 ot_mini 117 10.6 0.001 0.001 43.402 43.737 prepare_preconditioner 11 7.9 0.000 0.000 43.155 43.188 make_preconditioner 11 8.9 0.000 0.000 43.155 43.188 make_full_inverse_cholesky 11 9.9 0.012 0.022 36.912 41.758 multiply_cannon_multrec 30084 15.6 13.771 18.885 26.694 31.511 rebuild_ks_matrix 128 8.3 0.001 0.001 29.714 29.984 qs_ks_build_kohn_sham_matrix 128 9.3 0.019 0.028 29.714 29.984 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.853 27.083 mp_waitall_1 147882 16.7 17.572 26.980 17.572 26.980 qs_ot_get_derivative 117 11.6 0.002 0.002 23.799 24.140 make_m2s 5014 13.6 0.097 0.101 21.112 21.966 qs_ot_get_p 128 10.4 0.006 0.018 21.326 21.687 make_images 5014 14.6 1.966 2.247 20.801 21.654 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 19.037 19.571 apply_single 128 13.6 0.001 0.001 19.037 19.570 ot_diis_step 117 11.6 0.018 0.018 19.481 19.483 qs_ot_p2m_diag 83 11.4 0.343 0.390 16.973 17.028 cp_fm_cholesky_invert 11 10.9 16.951 16.960 16.951 16.960 cp_fm_upper_to_full 105 14.8 11.307 16.578 11.307 16.578 cp_dbcsr_syevd 83 12.4 0.006 0.006 15.531 15.534 multiply_cannon_metrocomm3 30084 15.6 0.050 0.052 6.127 14.335 sum_up_and_integrate 128 10.3 0.002 0.003 14.234 14.261 integrate_v_rspace 128 11.3 0.003 0.004 14.174 14.203 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.986 14.029 calculate_rho_elec 128 8.7 0.173 0.190 13.985 14.029 make_images_data 5014 15.6 0.067 0.074 11.447 13.196 dbcsr_mm_accdrv_process 62264 16.2 8.416 9.308 12.491 13.016 dbcsr_complete_redistribute 395 12.7 1.515 1.622 9.120 12.889 cp_fm_diag_elpa 83 13.4 0.000 0.000 12.165 12.179 cp_fm_diag_elpa_base 83 14.4 11.180 11.512 12.158 12.170 hybrid_alltoall_any 5200 16.5 0.530 2.208 10.008 12.158 copy_fm_to_dbcsr 209 11.7 0.001 0.002 7.655 11.362 multiply_cannon_sync_h2d 30084 15.6 10.474 11.069 10.474 11.069 init_scf_run 11 5.9 0.000 0.002 10.866 10.867 scf_env_initial_rho_setup 11 6.9 0.001 0.004 10.866 10.867 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.990 10.246 transfer_fm_to_dbcsr 11 9.9 0.002 0.008 6.220 9.867 pw_transfer 1547 11.6 0.086 0.099 9.216 9.283 mp_alltoall_i22 716 14.1 5.490 9.267 5.490 9.267 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.990 9.063 cp_fm_cholesky_decompose 22 10.9 7.919 8.030 7.919 8.030 fft_wrap_pw1pw2_140 523 13.2 1.446 1.484 7.934 8.024 grid_integrate_task_list 128 12.3 7.531 7.844 7.531 7.844 wfi_extrapolate 11 7.9 0.001 0.001 7.750 7.750 density_rs2pw 128 9.7 0.006 0.006 7.031 7.453 multiply_cannon_metrocomm4 25070 15.6 0.085 0.094 2.770 6.573 calculate_dm_sparse 128 9.5 0.001 0.001 6.273 6.376 mp_irecv_dv 76098 16.2 2.615 6.306 2.615 6.306 fft3d_ps 1291 14.7 2.852 2.921 6.169 6.220 grid_collocate_task_list 128 9.7 5.279 5.595 5.279 5.595 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.516 5.568 mp_alltoall_d11v 2415 14.1 5.110 5.522 5.110 5.522 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.590 4.687 potential_pw2rs 128 12.3 0.023 0.023 4.457 4.469 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 4.460 4.461 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.373 4.439 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.284 4.361 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=192.133000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1082.090909, yerr=20.438175 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/04dd4284328246f8f7259e2ea433868052050a67_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430454546432 0.0% 0.0% 100.0% flops 32 x 32 x 32 1975684956160 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992006770688 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613065416704 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239182565376 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239182565376 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.245913E+12 0.0% 0.0% 100.0% flops max/rank 5.910926E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806580192 0.0% 0.0% 100.0% number of processed stacks 1979136 0.0% 0.0% 100.0% average stack size 0.0 0.0 3439.2 marketing flops 145.663816E+12 ------------------------------------------------------------------------------- # multiplications 2535 max memory usage/rank 1.550709E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 243360 MPI messages size (bytes): total size 1.342108E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.514910E+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 115584 60599304192 4194304 < size <= 16777216 105840 554906419200 16777216 < size 20352 726592540656 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 9028 51. MP_Alltoall 9736 793528. MP_ISend 40516 2095952. MP_IRecv 40516 2095058. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4041 57652. MP_Allreduce 11179 1163. MP_Sync 88 MP_Alltoall 1724 18848021. MP_SendRecv 3870 122880. MP_ISendRecv 3870 122880. MP_Wait 16244 MP_ISend 10760 423501. MP_IRecv 10760 423501. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.027 0.079 174.648 174.674 qs_mol_dyn_low 1 2.0 0.003 0.003 174.199 174.223 qs_forces 11 3.9 0.005 0.006 174.071 174.074 qs_energies 11 4.9 0.002 0.004 166.658 166.666 scf_env_do_scf 11 5.9 0.001 0.001 149.157 149.171 velocity_verlet 10 3.0 0.023 0.028 115.332 115.336 scf_env_do_scf_inner_loop 118 6.6 0.005 0.010 112.103 112.105 qs_scf_new_mos 118 7.6 0.001 0.001 74.619 74.757 qs_scf_loop_do_ot 118 8.6 0.001 0.001 74.618 74.756 dbcsr_multiply_generic 2535 12.6 0.184 0.195 73.248 73.775 ot_scf_mini 118 9.6 0.003 0.003 70.131 70.270 multiply_cannon 2535 13.6 0.561 0.594 53.886 57.353 multiply_cannon_loop 2535 14.6 0.813 0.854 50.902 51.994 ot_mini 118 10.6 0.001 0.001 38.254 38.377 init_scf_loop 11 6.9 0.000 0.000 36.899 36.901 prepare_preconditioner 11 7.9 0.000 0.000 33.011 33.036 make_preconditioner 11 8.9 0.000 0.000 33.011 33.036 mp_waitall_1 127164 16.7 25.459 31.702 25.459 31.702 make_full_inverse_cholesky 11 9.9 0.015 0.027 30.838 31.135 rebuild_ks_matrix 129 8.3 0.001 0.001 29.499 29.652 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.019 29.498 29.652 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.916 27.062 qs_ot_get_derivative 118 11.6 0.002 0.002 21.110 21.256 multiply_cannon_multrec 10140 15.6 10.469 14.166 18.118 21.162 multiply_cannon_metrocomm3 10140 15.6 0.026 0.028 12.348 19.534 qs_ot_get_p 129 10.4 0.003 0.004 19.006 19.141 cp_fm_cholesky_invert 11 10.9 19.080 19.086 19.080 19.086 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 17.123 17.403 apply_single 129 13.6 0.001 0.001 17.123 17.402 ot_diis_step 118 11.6 0.020 0.021 17.058 17.059 make_m2s 5070 13.6 0.067 0.070 15.372 16.228 make_images 5070 14.6 2.197 2.625 15.062 15.914 qs_ot_p2m_diag 84 11.4 0.502 0.507 14.999 15.017 sum_up_and_integrate 129 10.3 0.002 0.002 14.358 14.408 integrate_v_rspace 129 11.3 0.004 0.004 14.297 14.349 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.286 14.315 calculate_rho_elec 129 8.7 0.258 0.269 14.285 14.315 cp_dbcsr_syevd 84 12.4 0.006 0.007 13.793 13.795 multiply_cannon_sync_h2d 10140 15.6 10.900 11.224 10.900 11.224 cp_fm_diag_elpa 84 13.4 0.000 0.000 10.805 10.816 cp_fm_diag_elpa_base 84 14.4 10.538 10.641 10.795 10.805 init_scf_run 11 5.9 0.000 0.002 10.582 10.582 scf_env_initial_rho_setup 11 6.9 0.001 0.002 10.582 10.582 make_images_data 5070 15.6 0.057 0.067 8.902 10.277 hybrid_alltoall_any 5257 16.5 0.857 3.837 8.693 10.143 pw_transfer 1559 11.6 0.086 0.096 9.566 9.586 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 9.341 9.369 qs_ot_get_derivative_diag 78 12.4 0.003 0.003 8.394 8.497 multiply_cannon_metrocomm1 10140 15.6 0.030 0.030 4.910 8.401 cp_fm_cholesky_decompose 22 10.9 8.150 8.276 8.150 8.276 fft_wrap_pw1pw2_140 527 13.2 1.822 1.857 8.221 8.250 grid_integrate_task_list 129 12.3 7.818 8.118 7.818 8.118 dbcsr_mm_accdrv_process 20958 16.1 3.373 4.564 7.296 7.997 wfi_extrapolate 11 7.9 0.001 0.001 7.483 7.484 density_rs2pw 129 9.7 0.006 0.006 6.907 7.440 calculate_dm_sparse 129 9.5 0.001 0.001 6.196 6.287 fft3d_ps 1301 14.7 2.798 2.877 6.036 6.068 grid_collocate_task_list 129 9.7 5.601 6.010 5.601 6.010 dbcsr_complete_redistribute 397 12.7 2.104 2.194 5.406 5.797 mp_alltoall_d11v 2429 14.1 4.892 5.367 4.892 5.367 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.308 5.309 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.212 5.222 mp_allgather_i34 2535 14.6 1.171 4.798 1.171 4.798 potential_pw2rs 129 12.3 0.026 0.027 4.413 4.426 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.080 4.136 multiply_cannon_metrocomm4 7605 15.6 0.027 0.029 1.816 3.949 mp_irecv_dv 29150 15.9 1.776 3.871 1.776 3.871 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.577 3.862 copy_fm_to_dbcsr 210 11.7 0.001 0.002 3.497 3.831 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.707 3.766 copy_dbcsr_to_fm 187 11.8 0.004 0.004 3.561 3.678 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.633 3.677 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.495 3.518 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=174.674000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1472.000000, yerr=17.529196 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/04dd4284328246f8f7259e2ea433868052050a67_performance_tests/20/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 11.785520E+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 1979552 0.0% 0.0% 100.0% average stack size 0.0 0.0 3438.3 marketing flops 145.646636E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 3.205345E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101080 MPI messages size (bytes): total size 1.144901E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.326685E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 45568 35366371328 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726591885696 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4024 58825. MP_Allreduce 11137 1501. MP_Sync 87 MP_Alltoall 1724 36993750. MP_SendRecv 1806 218624. MP_ISendRecv 1806 218624. MP_Wait 9876 MP_ISend 6456 1080169. MP_IRecv 6456 1080169. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.025 0.055 349.025 349.028 qs_mol_dyn_low 1 2.0 0.003 0.003 348.412 348.439 qs_forces 11 3.9 0.004 0.005 348.300 348.304 qs_energies 11 4.9 0.002 0.003 338.595 338.605 scf_env_do_scf 11 5.9 0.001 0.001 312.809 312.816 velocity_verlet 10 3.0 0.012 0.014 249.244 249.254 scf_env_do_scf_inner_loop 118 6.6 0.012 0.018 169.818 169.821 init_scf_loop 11 6.9 0.000 0.000 142.665 142.670 prepare_preconditioner 11 7.9 0.000 0.000 137.046 137.061 make_preconditioner 11 8.9 0.000 0.000 137.046 137.061 make_full_inverse_cholesky 11 9.9 0.037 0.039 112.201 133.959 qs_scf_new_mos 118 7.6 0.001 0.001 113.839 113.920 qs_scf_loop_do_ot 118 8.6 0.001 0.001 113.838 113.919 ot_scf_mini 118 9.6 0.004 0.004 108.492 108.537 dbcsr_multiply_generic 2527 12.6 0.216 0.227 92.498 92.865 cp_fm_upper_to_full 105 14.8 54.042 78.099 54.042 78.099 multiply_cannon 2527 13.6 0.696 0.739 61.519 64.334 multiply_cannon_loop 2527 14.6 1.067 1.090 56.546 57.765 ot_mini 118 10.6 0.001 0.001 51.694 51.746 dbcsr_complete_redistribute 395 12.7 3.989 4.032 32.388 46.090 cp_fm_cholesky_invert 11 10.9 43.400 43.411 43.400 43.411 copy_fm_to_dbcsr 209 11.7 0.001 0.001 28.065 41.768 rebuild_ks_matrix 129 8.3 0.001 0.001 41.551 41.581 qs_ks_build_kohn_sham_matrix 129 9.3 0.019 0.021 41.551 41.580 mp_waitall_1 104500 16.8 35.448 41.202 35.448 41.202 qs_ot_get_p 129 10.4 0.018 0.020 40.700 40.771 qs_ks_update_qs_env 129 7.6 0.001 0.001 38.374 38.403 transfer_fm_to_dbcsr 11 9.9 0.030 0.031 24.803 38.402 mp_alltoall_i22 718 14.1 22.449 36.137 22.449 36.137 qs_ot_p2m_diag 83 11.4 0.880 0.885 35.218 35.235 cp_dbcsr_syevd 83 12.4 0.008 0.011 32.805 32.811 qs_ot_get_derivative 118 11.6 0.002 0.002 31.177 31.215 cp_fm_diag_elpa 83 13.4 0.000 0.000 28.313 28.314 cp_fm_diag_elpa_base 83 14.4 23.371 25.482 28.304 28.305 make_m2s 5054 13.6 0.078 0.081 26.039 28.092 make_images 5054 14.6 3.105 3.293 25.552 27.607 qs_rho_update_rho_low 129 7.7 0.001 0.001 23.327 23.350 calculate_rho_elec 129 8.7 0.483 0.484 23.326 23.349 sum_up_and_integrate 129 10.3 0.002 0.002 20.922 21.003 integrate_v_rspace 129 11.3 0.004 0.004 20.858 20.940 multiply_cannon_metrocomm3 10108 15.6 0.025 0.026 18.935 20.531 ot_diis_step 118 11.6 0.022 0.023 20.474 20.476 make_images_data 5054 15.6 0.064 0.071 15.895 19.986 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.739 19.855 apply_single 129 13.6 0.001 0.001 19.739 19.854 pw_transfer 1559 11.6 0.096 0.097 18.664 18.683 hybrid_alltoall_any 5240 16.5 1.320 3.067 14.869 18.543 multiply_cannon_multrec 10108 15.6 10.478 12.442 18.407 18.502 fft_wrap_pw1pw2 1301 12.7 0.011 0.012 18.424 18.442 fft_wrap_pw1pw2_140 527 13.2 3.069 3.143 15.706 15.720 cp_fm_cholesky_decompose 22 10.9 14.731 14.754 14.731 14.754 init_scf_run 11 5.9 0.000 0.001 14.722 14.723 scf_env_initial_rho_setup 11 6.9 0.001 0.001 14.721 14.722 multiply_cannon_sync_h2d 10108 15.6 14.394 14.441 14.394 14.441 fft3d_ps 1301 14.7 2.883 2.895 13.635 13.684 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 12.792 12.819 density_rs2pw 129 9.7 0.005 0.006 12.372 12.426 mp_alltoall_d11v 2423 14.1 12.083 12.207 12.083 12.207 wfi_extrapolate 11 7.9 0.001 0.001 11.285 11.286 mp_alltoall_z22v 1301 16.7 10.097 10.153 10.097 10.153 dbcsr_mm_accdrv_process 20926 16.1 4.446 6.452 7.675 9.723 grid_integrate_task_list 129 12.3 8.646 8.855 8.646 8.855 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 8.409 8.411 potential_pw2rs 129 12.3 0.041 0.041 7.988 7.998 calculate_dm_sparse 129 9.5 0.001 0.001 7.372 7.467 copy_dbcsr_to_fm 186 11.8 0.004 0.004 7.237 7.289 make_images_sizes 5054 15.6 0.006 0.007 3.092 7.142 mp_alltoall_i44 5054 16.6 3.086 7.136 3.086 7.136 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 7.088 7.124 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=349.028000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2904.090909, yerr=152.995220 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/04dd4284328246f8f7259e2ea433868052050a67_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 1.259811E+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 29975755. 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 0.017 0.044 85.297 85.299 qs_energies 1 2.0 0.000 0.000 84.804 84.817 ls_scf 1 3.0 0.000 0.000 83.908 83.921 dbcsr_multiply_generic 111 6.7 0.015 0.026 72.841 73.013 multiply_cannon 111 7.7 0.017 0.020 55.922 56.929 multiply_cannon_loop 111 8.7 0.227 0.241 52.516 53.747 ls_scf_main 1 4.0 0.000 0.000 52.470 52.471 density_matrix_trs4 2 5.0 0.002 0.003 46.914 47.009 ls_scf_init_scf 1 4.0 0.000 0.000 28.398 28.400 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.262 27.317 mp_waitall_1 11031 10.9 22.451 26.199 22.451 26.199 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.003 25.114 25.131 multiply_cannon_multrec 2664 9.7 8.177 8.968 15.580 17.317 multiply_cannon_sync_h2d 2664 9.7 13.663 15.305 13.663 15.305 make_m2s 222 7.7 0.009 0.011 13.218 13.694 make_images 222 8.7 0.098 0.108 13.196 13.675 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.678 11.939 make_images_data 222 9.7 0.004 0.005 7.746 8.380 dbcsr_mm_accdrv_process 4760 10.4 0.597 0.730 7.021 7.971 hybrid_alltoall_any 227 10.6 0.215 1.835 6.632 7.866 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.360 7.512 dbcsr_mm_accdrv_process_sort 4760 11.4 6.225 7.119 6.225 7.119 calculate_norms 4752 9.8 5.523 6.193 5.523 6.193 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.041 5.204 mp_sum_l 887 5.1 3.171 4.574 3.171 4.574 make_images_sizes 222 9.7 0.000 0.000 0.812 3.707 mp_alltoall_i44 222 10.7 0.812 3.707 0.812 3.707 multiply_cannon_metrocomm4 2442 9.7 0.012 0.014 2.056 3.563 mp_irecv_dv 6231 10.9 2.040 3.547 2.040 3.547 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.355 3.288 arnoldi_extremal 4 6.8 0.000 0.000 3.247 3.274 arnoldi_normal_ev 4 7.8 0.001 0.005 3.247 3.274 build_subspace 16 8.4 0.009 0.012 3.144 3.146 ls_scf_post 1 4.0 0.000 0.000 3.039 3.053 ls_scf_store_result 1 5.0 0.000 0.000 2.856 2.898 dbcsr_special_finalize 555 9.7 0.005 0.006 2.308 2.762 dbcsr_merge_single_wm 555 10.7 0.459 0.590 2.300 2.753 make_images_pack 222 9.7 2.205 2.633 2.207 2.635 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.331 2.575 dbcsr_sort_data 658 11.4 2.098 2.521 2.098 2.521 dbcsr_matrix_vector_mult_local 304 10.0 2.064 2.459 2.066 2.461 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.312 2.384 buffer_matrices_ensure_size 222 8.7 1.757 2.152 1.757 2.152 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.772 1.774 rebuild_ks_matrix 3 7.3 0.000 0.000 1.763 1.764 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.763 1.764 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.299000, 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/04dd4284328246f8f7259e2ea433868052050a67_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 2.172830E+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 0.024 0.035 89.767 89.768 qs_energies 1 2.0 0.000 0.000 89.314 89.318 ls_scf 1 3.0 0.000 0.000 87.975 87.980 dbcsr_multiply_generic 111 6.7 0.015 0.016 74.086 74.357 multiply_cannon 111 7.7 0.026 0.037 52.742 57.114 ls_scf_main 1 4.0 0.000 0.000 54.362 54.369 multiply_cannon_loop 111 8.7 0.137 0.149 49.923 52.849 density_matrix_trs4 2 5.0 0.002 0.003 48.726 48.927 ls_scf_init_scf 1 4.0 0.000 0.000 29.924 29.925 mp_waitall_1 9105 10.9 20.567 29.211 20.567 29.211 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.721 28.814 multiply_cannon_multrec 1332 9.7 13.350 16.960 22.650 27.408 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.320 26.329 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.478 20.296 make_m2s 222 7.7 0.006 0.007 14.903 15.497 make_images 222 8.7 1.363 1.697 14.873 15.467 dbcsr_mm_accdrv_process 4041 10.4 0.324 0.511 8.898 10.444 dbcsr_mm_accdrv_process_sort 4041 11.4 8.413 9.912 8.413 9.912 make_images_data 222 9.7 0.004 0.004 8.621 9.531 hybrid_alltoall_any 227 10.6 0.542 2.539 8.059 9.479 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.190 8.018 mp_irecv_dv 3311 11.0 3.169 7.963 3.169 7.963 mp_sum_l 887 5.1 4.951 7.587 4.951 7.587 calculate_norms 2376 9.8 6.067 6.747 6.067 6.747 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.736 6.112 multiply_cannon_sync_h2d 1332 9.7 4.912 5.999 4.912 5.999 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.885 5.117 arnoldi_extremal 4 6.8 0.000 0.000 4.672 4.686 arnoldi_normal_ev 4 7.8 0.001 0.005 4.672 4.686 build_subspace 16 8.4 0.014 0.021 4.395 4.398 ls_scf_post 1 4.0 0.000 0.000 3.689 3.693 ls_scf_store_result 1 5.0 0.000 0.000 3.412 3.517 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.147 3.376 dbcsr_matrix_vector_mult_local 304 10.0 2.741 3.220 2.743 3.221 mp_allgather_i34 111 8.7 0.813 3.030 0.813 3.030 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.598 2.706 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.113 2.454 dbcsr_data_new 4174 10.1 2.116 2.411 2.116 2.411 make_images_pack 222 9.7 1.818 2.120 1.820 2.122 dbcsr_sort_data 436 11.2 1.850 2.059 1.850 2.059 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.852 1.854 rebuild_ks_matrix 3 7.3 0.000 0.000 1.839 1.841 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.839 1.841 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=89.768000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1779.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/04dd4284328246f8f7259e2ea433868052050a67_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 2.919899E+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 0.054 0.114 93.662 93.666 qs_energies 1 2.0 0.002 0.013 92.790 92.814 ls_scf 1 3.0 0.001 0.011 91.306 91.323 dbcsr_multiply_generic 111 6.7 0.016 0.022 75.541 75.781 ls_scf_main 1 4.0 0.001 0.012 56.703 56.726 multiply_cannon 111 7.7 0.035 0.074 52.195 56.053 multiply_cannon_loop 111 8.7 0.117 0.130 49.416 53.837 density_matrix_trs4 2 5.0 0.003 0.018 50.853 50.991 mp_waitall_1 7281 11.0 23.740 33.982 23.740 33.982 ls_scf_init_scf 1 4.0 0.002 0.014 30.927 30.942 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.325 29.433 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 26.909 26.921 multiply_cannon_multrec 888 9.7 12.694 15.436 21.291 24.595 multiply_cannon_metrocomm3 888 9.7 0.004 0.005 11.153 23.423 make_m2s 222 7.7 0.006 0.007 16.405 17.147 make_images 222 8.7 1.581 1.853 16.366 17.106 make_images_data 222 9.7 0.004 0.004 9.588 10.656 hybrid_alltoall_any 227 10.6 0.642 2.947 9.097 10.278 dbcsr_mm_accdrv_process 3754 10.4 0.312 0.505 8.120 9.385 dbcsr_mm_accdrv_process_sort 3754 11.4 7.677 8.880 7.677 8.880 mp_sum_l 887 5.1 5.248 8.824 5.248 8.824 multiply_cannon_metrocomm4 777 9.7 0.004 0.006 2.444 7.201 mp_irecv_dv 2335 11.1 2.428 7.143 2.428 7.143 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.118 7.126 multiply_cannon_sync_h2d 888 9.7 6.057 7.072 6.057 7.072 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.644 6.599 arnoldi_extremal 4 6.8 0.000 0.000 5.114 5.130 arnoldi_normal_ev 4 7.8 0.001 0.006 5.114 5.130 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.697 4.927 build_subspace 16 8.4 0.014 0.020 4.786 4.792 calculate_norms 1584 9.8 4.348 4.735 4.348 4.735 mp_allgather_i34 111 8.7 0.892 3.825 0.892 3.825 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.450 3.791 ls_scf_post 1 4.0 0.002 0.013 3.675 3.691 dbcsr_matrix_vector_mult_local 304 10.0 3.029 3.617 3.031 3.619 ls_scf_store_result 1 5.0 0.000 0.000 3.401 3.500 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.827 2.917 dbcsr_data_new 4116 9.9 2.113 2.456 2.113 2.456 make_images_sizes 222 9.7 0.000 0.000 0.981 2.360 mp_alltoall_i44 222 10.7 0.980 2.360 0.980 2.360 dbcsr_sort_data 325 11.1 1.887 2.155 1.887 2.155 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.066 2.078 rebuild_ks_matrix 3 7.3 0.000 0.000 2.048 2.059 qs_ks_build_kohn_sham_matrix 3 8.3 0.019 0.031 2.048 2.059 make_images_pack 222 9.7 1.623 1.873 1.626 1.876 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=93.666000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2286.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/04dd4284328246f8f7259e2ea433868052050a67_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 3.336450E+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 0.029 0.044 116.796 116.797 qs_energies 1 2.0 0.000 0.000 116.167 116.173 ls_scf 1 3.0 0.000 0.000 114.506 114.511 dbcsr_multiply_generic 111 6.7 0.017 0.018 97.959 98.203 multiply_cannon 111 7.7 0.040 0.085 63.805 75.251 multiply_cannon_loop 111 8.7 0.153 0.170 58.685 68.336 ls_scf_main 1 4.0 0.000 0.000 58.546 58.547 ls_scf_init_scf 1 4.0 0.000 0.000 52.767 52.770 density_matrix_trs4 2 5.0 0.002 0.003 52.397 52.541 ls_scf_init_matrix_S 1 5.0 0.000 0.000 51.556 51.625 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 49.049 49.060 mp_waitall_1 6369 11.0 24.502 39.371 24.502 39.371 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 10.775 25.900 multiply_cannon_multrec 1332 9.7 14.138 17.052 22.065 24.656 make_m2s 222 7.7 0.007 0.007 21.009 22.493 make_images 222 8.7 3.135 3.598 20.958 22.446 compute_matrix_preconditioner 1 6.0 0.001 0.001 20.863 20.871 mp_sum_l 887 5.1 11.571 17.030 11.571 17.030 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 10.539 15.364 make_images_data 222 9.7 0.004 0.004 11.764 13.307 hybrid_alltoall_any 227 10.6 0.796 3.772 11.061 13.084 acc_transpose_blocks 1332 9.7 0.008 0.008 10.521 10.795 acc_transpose_blocks_kernels 1332 10.7 0.018 0.019 10.410 10.685 jit_kernel_transpose 1 13.0 10.393 10.667 10.393 10.667 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.789 10.169 dbcsr_mm_accdrv_process 3641 10.4 0.303 0.540 7.573 9.084 dbcsr_mm_accdrv_process_sort 3641 11.4 7.122 8.587 7.122 8.587 multiply_cannon_sync_h2d 1332 9.7 5.421 6.210 5.421 6.210 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.097 6.086 mp_irecv_dv 3229 10.9 2.072 6.011 2.072 6.011 arnoldi_extremal 4 6.8 0.000 0.000 5.343 5.359 arnoldi_normal_ev 4 7.8 0.001 0.004 5.343 5.359 build_subspace 16 8.4 0.014 0.021 4.987 4.997 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.572 4.700 mp_allgather_i34 111 8.7 2.241 4.683 2.241 4.683 calculate_norms 2376 9.8 4.180 4.477 4.180 4.477 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.606 3.932 dbcsr_matrix_vector_mult_local 304 10.0 3.195 3.702 3.197 3.704 dbcsr_sort_data 658 11.4 3.038 3.474 3.038 3.474 ls_scf_post 1 4.0 0.000 0.000 3.192 3.198 dbcsr_special_finalize 555 9.7 0.006 0.007 2.787 3.180 dbcsr_merge_single_wm 555 10.7 0.539 0.659 2.778 3.171 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.013 3.064 ls_scf_store_result 1 5.0 0.000 0.000 2.942 2.994 dbcsr_data_release 10477 10.7 1.575 2.452 1.575 2.452 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=116.797000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2752.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/04dd4284328246f8f7259e2ea433868052050a67_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 4.755423E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265536. MP_Allreduce 3129 15263. MP_Sync 4 MP_Alltoall 47 46208988. MP_SendRecv 45 115200. MP_ISendRecv 45 115200. MP_Wait 528 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.040 0.071 95.045 95.047 qs_energies 1 2.0 0.000 0.000 94.211 94.215 ls_scf 1 3.0 0.000 0.000 92.263 92.268 dbcsr_multiply_generic 111 6.7 0.017 0.019 73.627 73.892 multiply_cannon 111 7.7 0.078 0.132 54.944 57.861 ls_scf_main 1 4.0 0.000 0.000 57.219 57.219 multiply_cannon_loop 111 8.7 0.088 0.094 52.338 54.257 density_matrix_trs4 2 5.0 0.002 0.003 50.246 50.321 ls_scf_init_scf 1 4.0 0.000 0.000 31.644 31.649 ls_scf_init_matrix_S 1 5.0 0.000 0.000 30.388 30.421 mp_waitall_1 5436 11.0 24.939 29.105 24.939 29.105 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.217 28.229 multiply_cannon_multrec 444 9.7 13.696 16.214 20.796 22.179 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.059 15.722 make_m2s 222 7.7 0.005 0.005 13.814 14.789 make_images 222 8.7 2.037 2.476 13.747 14.719 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.160 14.336 hybrid_alltoall_any 227 10.6 0.799 3.803 8.316 10.098 make_images_data 222 9.7 0.003 0.004 8.535 9.856 multiply_cannon_sync_h2d 444 9.7 6.730 8.235 6.730 8.235 dbcsr_mm_accdrv_process 3003 10.4 0.357 0.428 6.791 7.910 dbcsr_mm_accdrv_process_sort 3003 11.4 6.421 7.510 6.421 7.510 arnoldi_extremal 4 6.8 0.000 0.000 5.969 5.981 arnoldi_normal_ev 4 7.8 0.002 0.006 5.969 5.981 build_subspace 16 8.4 0.015 0.020 5.553 5.563 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.421 4.662 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.208 4.413 mp_sum_l 887 5.1 2.886 4.391 2.886 4.391 dbcsr_matrix_vector_mult_local 304 10.0 3.726 4.202 3.728 4.204 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.552 3.820 mp_allgather_i34 111 8.7 1.178 3.812 1.178 3.812 mp_irecv_dv 1241 11.2 1.540 3.771 1.540 3.771 calculate_norms 792 9.8 3.601 3.730 3.601 3.730 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.531 3.602 make_images_sizes 222 9.7 0.000 0.000 0.884 3.449 mp_alltoall_i44 222 10.7 0.883 3.448 0.883 3.448 ls_scf_post 1 4.0 0.000 0.000 3.399 3.404 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.047 3.386 ls_scf_store_result 1 5.0 0.000 0.000 3.183 3.223 compute_matrix_preconditioner 1 6.0 0.001 0.002 2.742 2.748 dbcsr_data_new 4608 9.7 1.821 2.363 1.821 2.363 acc_transpose_blocks 444 9.7 0.003 0.003 2.300 2.306 dbcsr_finalize 304 7.8 0.062 0.077 2.194 2.299 acc_transpose_blocks_kernels 444 10.7 0.006 0.007 2.234 2.239 jit_kernel_transpose 1 13.0 2.228 2.232 2.228 2.232 dbcsr_merge_all 275 8.9 0.479 0.527 2.052 2.137 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.030 2.030 rebuild_ks_matrix 3 7.3 0.000 0.000 1.996 1.997 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.996 1.997 qs_energies_init_hamiltonians 1 3.0 0.001 0.004 1.933 1.933 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=95.047000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3764.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/04dd4284328246f8f7259e2ea433868052050a67_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 8.980578E+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 0.076 0.092 110.452 110.453 qs_energies 1 2.0 0.000 0.000 109.083 109.087 ls_scf 1 3.0 0.000 0.000 106.156 106.159 dbcsr_multiply_generic 111 6.7 0.024 0.028 79.719 79.862 ls_scf_main 1 4.0 0.000 0.000 63.852 63.853 multiply_cannon 111 7.7 0.155 0.249 53.410 55.722 density_matrix_trs4 2 5.0 0.002 0.003 54.874 54.928 multiply_cannon_loop 111 8.7 0.100 0.111 50.313 50.933 ls_scf_init_scf 1 4.0 0.000 0.000 38.503 38.504 ls_scf_init_matrix_S 1 5.0 0.000 0.000 37.003 37.027 matrix_sqrt_Newton_Schulz 2 6.5 0.002 0.006 34.302 34.309 mp_waitall_1 4527 11.1 22.425 26.639 22.425 26.639 make_m2s 222 7.7 0.005 0.005 22.681 23.858 make_images 222 8.7 3.573 3.868 22.573 23.749 multiply_cannon_multrec 444 9.7 17.879 18.540 22.543 23.205 hybrid_alltoall_any 227 10.6 1.655 3.632 12.772 15.693 make_images_data 222 9.7 0.003 0.004 13.044 15.305 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.760 15.292 multiply_cannon_sync_h2d 444 9.7 8.792 8.824 8.792 8.824 arnoldi_extremal 4 6.8 0.000 0.000 7.559 7.568 arnoldi_normal_ev 4 7.8 0.003 0.010 7.559 7.568 build_subspace 16 8.4 0.026 0.036 6.982 6.997 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.536 5.721 compute_matrix_preconditioner 1 6.0 0.002 0.002 5.584 5.591 dbcsr_matrix_vector_mult_local 304 10.0 5.114 5.447 5.117 5.450 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.098 5.191 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.873 5.132 acc_transpose_blocks 444 9.7 0.003 0.003 4.474 5.090 acc_transpose_blocks_kernels 444 10.7 0.006 0.006 4.390 5.006 jit_kernel_transpose 1 13.0 4.385 5.000 4.385 5.000 dbcsr_mm_accdrv_process 1814 10.4 0.269 0.361 4.467 4.601 dbcsr_mm_accdrv_process_sort 1814 11.4 4.127 4.264 4.127 4.264 mp_allgather_i34 111 8.7 1.219 3.874 1.219 3.874 ls_scf_post 1 4.0 0.000 0.000 3.800 3.805 make_images_sizes 222 9.7 0.000 0.000 1.504 3.685 mp_alltoall_i44 222 10.7 1.504 3.684 1.504 3.684 ls_scf_store_result 1 5.0 0.000 0.000 3.517 3.554 calculate_norms 792 9.8 3.231 3.268 3.231 3.268 dbcsr_finalize 304 7.8 0.082 0.089 3.070 3.146 dbcsr_merge_all 275 8.9 0.883 0.917 2.856 2.926 qs_energies_init_hamiltonians 1 3.0 0.001 0.002 2.897 2.897 dbcsr_complete_redistribute 5 7.6 1.426 1.470 2.735 2.849 matrix_ls_to_qs 2 6.0 0.000 0.000 2.406 2.516 dbcsr_sort_data 325 11.1 2.440 2.485 2.440 2.485 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.395 2.397 rebuild_ks_matrix 3 7.3 0.000 0.000 2.325 2.327 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.002 2.325 2.327 dbcsr_data_new 6591 9.6 1.865 2.323 1.865 2.323 dbcsr_new_transposed 4 7.5 0.247 0.263 2.306 2.316 dbcsr_frobenius_norm 74 6.6 2.054 2.129 2.191 2.222 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=110.453000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6989.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/04dd4284328246f8f7259e2ea433868052050a67_performance_tests/27/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 7009386627072 0.0% 0.0% 100.0% flops 9 x 9 x 32 7335108845568 0.0% 0.0% 100.0% flops 9 x 22 x 32 9866241589248 0.0% 0.0% 100.0% flops 22 x 9 x 32 9884108906496 0.0% 0.0% 100.0% flops 22 x 22 x 32 13354440523776 0.0% 0.0% 100.0% flops 32 x 32 x 9 20607185977344 0.0% 0.0% 100.0% flops 32 x 32 x 22 25186560638976 0.0% 0.0% 100.0% flops 9 x 32 x 32 28458319085568 0.0% 0.0% 100.0% flops 22 x 32 x 32 34782389993472 0.0% 0.0% 100.0% flops 9 x 32 x 9 42881542373376 0.0% 0.0% 100.0% flops 22 x 32 x 9 55680402235392 0.0% 0.0% 100.0% flops 9 x 32 x 22 55680402235392 0.0% 0.0% 100.0% flops 22 x 32 x 22 72328573419520 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 383.054662E+12 0.0% 0.0% 100.0% flops max/rank 733.641090E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 26899403712 0.0% 0.0% 100.0% number of processed stacks 118860288 0.0% 0.0% 100.0% average stack size 0.0 0.0 226.3 marketing flops 780.439111E+12 ------------------------------------------------------------------------------- # multiplications 1445 max memory usage/rank 593.113088E+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 4802885. 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 0.032 0.082 243.870 243.875 qs_mol_dyn_low 1 2.0 0.003 0.014 242.795 242.838 qs_forces 5 3.8 0.022 0.052 242.689 242.693 qs_energies 5 4.8 0.003 0.032 239.572 239.612 scf_env_do_scf 5 5.8 0.000 0.009 222.101 222.104 scf_env_do_scf_inner_loop 105 6.6 0.003 0.026 193.227 193.241 qs_scf_new_mos 105 7.6 0.000 0.001 150.470 150.647 qs_scf_loop_do_ot 105 8.6 0.001 0.001 150.470 150.647 ot_scf_mini 105 9.6 0.003 0.004 140.422 140.571 dbcsr_multiply_generic 1445 12.2 0.127 0.133 137.444 137.964 multiply_cannon 1445 13.2 0.275 0.286 117.389 120.106 multiply_cannon_loop 1445 14.2 2.861 3.018 115.499 117.067 velocity_verlet 4 3.0 0.009 0.025 110.627 110.629 ot_mini 105 10.6 0.001 0.001 61.081 61.216 qs_ot_get_p 112 10.4 0.001 0.003 49.061 49.374 multiply_cannon_multrec 69360 15.2 29.488 34.431 40.071 45.105 mp_waitall_1 488190 16.1 37.053 43.895 37.053 43.895 qs_ot_get_derivative 55 11.6 0.001 0.001 39.271 39.405 qs_ot_p2m_diag 40 11.0 0.020 0.031 37.507 37.648 multiply_cannon_metrocomm3 69360 15.2 0.202 0.212 26.867 34.292 cp_dbcsr_syevd 40 12.0 0.002 0.002 34.151 34.153 multiply_cannon_sync_h2d 69360 15.2 28.798 32.851 28.798 32.851 rebuild_ks_matrix 110 8.4 0.000 0.000 30.384 30.561 qs_ks_build_kohn_sham_matrix 110 9.4 0.012 0.026 30.384 30.561 cp_fm_syevd 40 13.0 0.000 0.000 28.844 28.996 init_scf_loop 7 6.6 0.001 0.042 28.834 28.836 qs_ks_update_qs_env 112 7.6 0.001 0.001 27.947 28.114 cp_fm_redistribute_end 40 14.0 12.007 23.959 12.013 23.962 cp_fm_syevd_base 40 14.0 11.936 23.889 11.936 23.889 prepare_preconditioner 7 7.6 0.000 0.001 23.597 23.636 make_preconditioner 7 8.6 0.000 0.000 23.597 23.636 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 23.136 23.403 apply_single 62 13.6 0.000 0.000 23.136 23.403 ot_new_cg_direction 55 11.6 0.001 0.001 21.109 21.110 qs_rho_update_rho_low 110 7.6 0.001 0.001 18.403 18.765 calculate_rho_elec 110 8.6 0.030 0.032 18.402 18.764 make_full_inverse_cholesky 7 9.6 0.000 0.000 15.878 15.941 qs_ot_get_orbitals 105 10.6 0.001 0.001 15.049 15.224 init_scf_run 5 5.8 0.000 0.001 14.620 14.621 scf_env_initial_rho_setup 5 6.8 0.002 0.003 14.620 14.621 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 14.063 14.155 mp_sum_l 4764 12.2 12.470 13.318 12.470 13.318 pw_transfer 1645 12.4 0.079 0.101 12.910 13.164 fft_wrap_pw1pw2 1425 13.5 0.012 0.015 12.771 13.030 density_rs2pw 110 9.6 0.006 0.007 12.467 12.961 calculate_dm_sparse 110 9.5 0.000 0.001 12.656 12.868 dbcsr_mm_accdrv_process 154766 15.8 6.290 6.473 10.453 11.676 fft_wrap_pw1pw2_240 915 15.0 1.130 1.229 11.312 11.556 qs_vxc_create 110 10.4 0.002 0.003 11.044 11.099 cp_fm_cholesky_invert 7 10.6 11.030 11.041 11.030 11.041 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 10.557 10.642 check_diag 80 13.5 8.596 8.888 9.699 9.852 fft3d_pb 915 16.0 2.374 2.660 9.547 9.800 calculate_first_density_matrix 1 7.0 0.001 0.005 9.645 9.664 transfer_rs2pw 445 10.6 0.007 0.008 8.808 9.328 acc_transpose_blocks 69360 15.2 0.355 0.371 8.789 9.315 sum_up_and_integrate 60 10.3 0.001 0.002 9.195 9.210 integrate_v_rspace 60 11.3 0.002 0.032 9.178 9.194 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.018 8.693 8.709 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 8.246 8.312 multiply_cannon_metrocomm1 69360 15.2 0.091 0.096 4.981 8.068 xc_rho_set_and_dset_create 110 12.4 0.076 0.098 7.720 7.973 make_m2s 2890 13.2 0.079 0.088 7.114 7.693 make_images 2890 14.2 0.242 0.262 7.007 7.587 xc_vxc_pw_create 60 11.3 0.039 0.051 7.425 7.479 make_full_single_inverse 7 9.6 0.001 0.002 7.397 7.437 mp_alltoall_z22v 2340 17.7 6.645 6.919 6.645 6.919 xc_pw_derive 510 13.4 0.005 0.007 6.769 6.854 acc_transpose_blocks_kernels 69360 16.2 0.851 0.896 6.024 6.449 jit_kernel_transpose 5 15.0 5.174 5.569 5.174 5.569 potential_pw2rs 60 12.3 0.003 0.003 5.292 5.326 mp_waitany 7680 13.5 4.653 5.300 4.653 5.300 mp_alltoall_d11v 1300 13.8 3.789 4.985 3.789 4.985 multiply_cannon_metrocomm4 67915 15.2 0.185 0.202 2.035 4.940 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=243.875000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=562.200000, yerr=3.487119 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 04dd4284328246f8f7259e2ea433868052050a67 Summary: empty Status: OK