=== 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: 47690d86814a0e25b6e15635e68dedc68a09033e ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, # Cray-FFTW 3.3.8.10, COSMA 2.6.6, ELPA 2023.05.001, # HDF5 1.14.2, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.2.2, LIBVORI 220621, LIBXSMM 1.17, # PLUMED 2.9.0, SPGLIB 1.16.2, LIBGRPP 20231215 # # 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: 25.01.2024 # # \ 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-hdf5 --with-libvdwxc --with-pexsi --with-plumed --with-sirius=no; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2023.05.001 USE_HDF5 := 1.14.2 USE_LIBGRPP := 20231215 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.9.0 #USE_QUIP := 0.9.10 #USE_DEEPMD := 2.2.7 #USE_SIRIUS := 7.5.2 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS #LIBVDWXC_VER := 0.4.0 #SPFFT_VER := 1.0.6 #SPLA_VER := 1.5.5 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(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_DEEPMD),) USE_DEEPMD := $(strip $(USE_DEEPMD)) DEEPMD_INC := $(INSTALL_PATH)/libdeepmd_c-$(USE_DEEPMD)/include DEEPMD_LIB := $(INSTALL_PATH)/libdeepmd_c-$(USE_DEEPMD)/lib CFLAGS += -I$(DEEPMD_INC) DFLAGS += -D__DEEPMD LIBS += $(DEEPMD_LIB)/libdeepmd.so LIBS += $(DEEPMD_LIB)/libdeepmd_c.so LIBS += $(DEEPMD_LIB)/libdeepmd_cc.so LIBS += $(DEEPMD_LIB)/libdeepmd_dyn_cudart.so LIBS += $(DEEPMD_LIB)/libdeepmd_op.so LIBS += $(DEEPMD_LIB)/libdeepmd_op_cuda.so LIBS += $(DEEPMD_LIB)/libtensorflow_cc.so.2 LIBS += $(DEEPMD_LIB)/libtensorflow_framework.so.2 endif ifneq ($(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_LIBGRPP),) USE_LIBGRPP := $(strip $(USE_LIBGRPP)) LIBGRPP_INC := $(INSTALL_PATH)/libgrpp-main-$(USE_LIBGRPP)/include LIBGRPP_LIB := $(INSTALL_PATH)/libgrpp-main-$(USE_LIBGRPP)/lib CFLAGS += -I$(LIBGRPP_INC) DFLAGS += -D__LIBGRPP LIBS += $(LIBGRPP_LIB)/liblibgrpp.a endif ifneq ($(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/47690d86814a0e25b6e15635e68dedc68a09033e_performance_tests/01 job id: 51866460 --- 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/47690d86814a0e25b6e15635e68dedc68a09033e_performance_tests/02 job id: 51866462 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/47690d86814a0e25b6e15635e68dedc68a09033e_performance_tests/03 job id: 51866464 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/47690d86814a0e25b6e15635e68dedc68a09033e_performance_tests/04 job id: 51866465 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/47690d86814a0e25b6e15635e68dedc68a09033e_performance_tests/05 job id: 51866466 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/47690d86814a0e25b6e15635e68dedc68a09033e_performance_tests/06 job id: 51866467 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/47690d86814a0e25b6e15635e68dedc68a09033e_performance_tests/07 job id: 51866470 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/47690d86814a0e25b6e15635e68dedc68a09033e_performance_tests/08 job id: 51866472 --- 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/47690d86814a0e25b6e15635e68dedc68a09033e_performance_tests/09 job id: 51866473 --- 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/47690d86814a0e25b6e15635e68dedc68a09033e_performance_tests/10 job id: 51866474 --- 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/47690d86814a0e25b6e15635e68dedc68a09033e_performance_tests/11 job id: 51866475 --- 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/47690d86814a0e25b6e15635e68dedc68a09033e_performance_tests/12 job id: 51866477 --- 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/47690d86814a0e25b6e15635e68dedc68a09033e_performance_tests/13 job id: 51866478 --- 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/47690d86814a0e25b6e15635e68dedc68a09033e_performance_tests/14 job id: 51866479 --- 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/47690d86814a0e25b6e15635e68dedc68a09033e_performance_tests/15 job id: 51866480 --- 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/47690d86814a0e25b6e15635e68dedc68a09033e_performance_tests/16 job id: 51866481 --- 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/47690d86814a0e25b6e15635e68dedc68a09033e_performance_tests/17 job id: 51866482 --- 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/47690d86814a0e25b6e15635e68dedc68a09033e_performance_tests/18 job id: 51866483 --- 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/47690d86814a0e25b6e15635e68dedc68a09033e_performance_tests/19 job id: 51866484 --- 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/47690d86814a0e25b6e15635e68dedc68a09033e_performance_tests/20 job id: 51866486 --- 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] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/47690d86814a0e25b6e15635e68dedc68a09033e_performance_tests/21 job id: 51866487 --- 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] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/47690d86814a0e25b6e15635e68dedc68a09033e_performance_tests/22 job id: 51866488 --- 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] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/47690d86814a0e25b6e15635e68dedc68a09033e_performance_tests/23 job id: 51866489 --- 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] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/47690d86814a0e25b6e15635e68dedc68a09033e_performance_tests/24 job id: 51866490 --- 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] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/47690d86814a0e25b6e15635e68dedc68a09033e_performance_tests/25 job id: 51866491 --- 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] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/47690d86814a0e25b6e15635e68dedc68a09033e_performance_tests/26 job id: 51866493 --- 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] = 20 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/47690d86814a0e25b6e15635e68dedc68a09033e_performance_tests/27 job id: 51866494 --- 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/47690d86814a0e25b6e15635e68dedc68a09033e_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 172669. 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 0.044 0.045 144.700 146.134 farming_run 1 2.0 143.913 145.340 144.638 146.072 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.550778E+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 0.006 0.027 122.402 123.616 qs_energies 1 2.0 0.005 0.006 122.187 123.399 mp2_main 1 3.0 0.000 0.000 119.436 120.620 mp2_gpw_main 1 4.0 0.063 0.359 117.919 119.089 mp2_ri_gpw_compute_in 1 5.0 0.192 0.196 98.215 99.211 mp2_ri_gpw_compute_in_loop 1 6.0 0.003 0.004 57.440 58.032 mp2_eri_3c_integrate_gpw 272 7.0 0.151 0.160 43.835 48.516 get_2c_integrals 1 6.0 0.008 0.009 39.922 40.471 integrate_v_rspace 273 8.0 0.439 0.454 25.573 29.768 fft_wrap_pw1pw2 5465 10.4 0.067 0.071 28.210 28.937 fft_wrap_pw1pw2_100 2178 11.4 1.334 1.442 25.639 26.371 grid_integrate_task_list 273 9.0 20.798 25.929 20.798 25.929 compute_2c_integrals 1 7.0 0.002 0.002 21.599 21.813 compute_2c_integrals_loop_lm 1 8.0 0.001 0.003 20.978 21.370 mp2_eri_2c_integrate_gpw 1 9.0 2.402 2.431 20.976 21.369 rpa_ri_compute_en 1 5.0 0.019 0.021 19.258 19.491 cp_fm_cholesky_decompose 12 8.2 18.455 19.058 18.455 19.058 cholesky_decomp 1 7.0 0.000 0.000 17.161 17.772 fft3d_s 5443 12.4 16.861 17.305 16.882 17.327 ao_to_mo_and_store_B_mult_1 272 7.0 10.810 15.493 10.810 15.493 calculate_wavefunction 272 8.0 5.466 5.592 12.945 13.833 calc_potential_gpw 544 9.5 0.004 0.004 10.781 11.797 rpa_num_int 1 6.0 0.014 0.224 11.164 11.275 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.924 11.036 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 9.852 10.653 potential_pw2rs 545 10.0 0.106 0.108 8.624 10.468 calc_mat_Q 8 8.0 0.000 0.000 9.602 9.767 contract_S_to_Q 8 9.0 0.000 0.000 9.027 9.182 parallel_gemm_fm 14 9.1 0.000 0.000 8.600 8.691 parallel_gemm_fm_cosma 14 10.1 8.600 8.691 8.600 8.691 collocate_single_gaussian 272 10.0 0.040 0.041 7.910 8.651 create_integ_mat 1 6.0 0.039 0.044 7.914 7.993 array2fm 1 7.0 0.000 0.000 6.706 7.205 pw_poisson_solve 545 10.5 0.010 0.010 4.670 4.953 pw_gather_s 2722 12.2 4.097 4.945 4.097 4.945 pw_scatter_s 2720 12.7 4.422 4.777 4.422 4.777 pw_poisson_set 548 11.5 0.017 0.018 3.330 3.636 array2fm_buffer_send 1 8.0 2.973 3.144 2.973 3.144 pw_copy 4911 11.6 2.626 2.749 2.626 2.749 scf_env_do_scf 1 3.0 0.000 0.000 2.465 2.489 cp_fm_syevd 4 7.5 0.000 0.000 2.455 2.482 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=117.586390, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2821.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/47690d86814a0e25b6e15635e68dedc68a09033e_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 200775. 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 0.053 0.117 533.729 638.300 farming_run 1 2.0 532.548 636.885 533.584 638.125 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.277456E+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 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 0.012 0.033 326.977 391.037 qs_energies 1 2.0 0.008 0.031 326.764 390.782 mp2_main 1 3.0 0.002 0.015 210.860 252.180 mp2_gpw_main 1 4.0 0.062 0.154 209.989 251.127 mp2_ri_gpw_compute_en 1 5.0 0.039 0.048 129.094 155.632 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.789 2.202 127.352 152.307 scf_env_do_scf 1 3.0 0.000 0.000 115.611 138.261 qs_ks_update_qs_env 5 5.0 0.000 0.000 114.733 137.217 rebuild_ks_matrix 4 6.0 0.000 0.000 114.731 137.215 qs_ks_build_kohn_sham_matrix 4 7.0 0.054 0.070 114.731 137.215 hfx_ks_matrix 4 8.0 0.001 0.001 114.378 136.790 integrate_four_center 4 9.0 0.154 0.475 114.378 136.789 mp2_ri_gpw_compute_en_expansio 172 7.0 0.468 0.602 111.467 133.918 local_gemm 172 8.0 111.000 133.316 111.000 133.316 integrate_four_center_main 4 10.0 0.115 0.494 102.229 121.496 integrate_four_center_bin 264 11.0 102.114 121.489 102.114 121.489 init_scf_loop 1 4.0 0.000 0.000 97.335 116.405 mp2_ri_gpw_compute_in 1 5.0 0.072 0.095 80.690 95.245 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.007 57.065 66.980 mp2_eri_3c_integrate_gpw 91 7.0 0.148 0.192 44.627 49.948 fft_wrap_pw1pw2 1868 10.4 0.030 0.035 30.704 37.272 fft_wrap_pw1pw2_100 730 11.4 0.670 0.822 28.385 34.509 integrate_v_rspace 95 8.0 0.399 0.564 28.171 33.240 ao_to_mo_and_store_B_mult_1 91 7.0 10.750 28.925 10.750 28.925 grid_integrate_task_list 95 9.0 23.443 28.711 23.443 28.711 get_2c_integrals 1 6.0 0.012 0.104 23.426 28.006 compute_2c_integrals 1 7.0 0.011 0.074 22.392 26.793 compute_2c_integrals_loop_lm 1 8.0 0.002 0.009 21.918 26.387 mp2_eri_2c_integrate_gpw 1 9.0 1.722 2.069 21.916 26.385 fft3d_s 1823 12.4 19.316 22.872 19.330 22.889 scf_env_do_scf_inner_loop 4 4.0 0.000 0.001 18.274 21.854 calc_potential_gpw 182 9.5 0.002 0.002 12.458 14.673 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.013 10.827 13.678 potential_pw2rs 186 10.0 0.033 0.040 9.129 12.515 calculate_wavefunction 91 8.0 2.037 2.432 10.096 11.624 collocate_single_gaussian 91 10.0 0.018 0.040 8.374 9.494 mp2_ri_gpw_compute_en_comm 22 7.0 0.506 0.597 7.964 8.991 integrate_four_center_load 4 10.0 0.000 0.000 6.797 8.140 hfx_load_balance 1 11.0 0.000 0.000 6.796 8.140 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=208.890040, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1530.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/47690d86814a0e25b6e15635e68dedc68a09033e_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 447.074304E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62379. MP_Allreduce 10329 270. MP_Sync 530 MP_Alltoall 2083 MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.027 55.417 58.824 qs_mol_dyn_low 1 2.0 0.013 0.088 55.119 58.508 qs_forces 11 3.9 0.005 0.011 54.897 58.275 qs_energies 11 4.9 0.001 0.002 53.385 56.668 scf_env_do_scf 11 5.9 0.000 0.001 46.596 49.460 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 44.527 47.265 qs_scf_new_mos 108 7.5 0.000 0.001 34.403 36.844 qs_scf_loop_do_ot 108 8.5 0.000 0.001 34.402 36.844 dbcsr_multiply_generic 2286 12.5 0.100 0.105 34.307 36.759 ot_scf_mini 108 9.5 0.002 0.002 32.688 34.870 multiply_cannon 2286 13.5 0.182 0.195 26.798 29.192 multiply_cannon_loop 2286 14.5 1.846 2.051 26.129 28.519 velocity_verlet 10 3.0 0.012 0.066 26.485 28.121 ot_mini 108 10.5 0.001 0.001 19.755 21.220 qs_ot_get_derivative 108 11.5 0.001 0.001 16.691 17.883 mp_waitall_1 245248 16.5 8.625 15.809 8.625 15.809 multiply_cannon_metrocomm3 54864 15.5 0.076 0.082 6.114 13.571 multiply_cannon_multrec 54864 15.5 3.712 5.925 7.843 11.772 qs_ot_get_p 119 10.4 0.001 0.002 8.188 8.963 rebuild_ks_matrix 119 8.3 0.000 0.000 7.986 8.629 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 7.986 8.628 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.030 7.601 mp_sum_l 7287 12.8 5.497 7.447 5.497 7.447 multiply_cannon_sync_h2d 54864 15.5 5.257 6.660 5.257 6.660 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.697 6.299 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.627 6.052 dbcsr_mm_accdrv_process 76910 16.1 1.873 3.133 4.044 6.045 init_scf_run 11 5.9 0.000 0.001 5.352 5.681 scf_env_initial_rho_setup 11 6.9 0.005 0.014 5.352 5.681 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.307 5.661 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.496 4.772 sum_up_and_integrate 119 10.3 0.001 0.002 4.488 4.770 integrate_v_rspace 119 11.3 0.002 0.003 4.477 4.759 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.295 4.559 cp_fm_redistribute_end 50 14.0 2.198 4.531 2.203 4.535 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.121 4.502 calculate_rho_elec 119 8.7 0.011 0.018 4.120 4.502 cp_fm_diag_elpa_base 50 14.0 2.064 4.324 2.068 4.332 calculate_dm_sparse 119 9.5 0.000 0.000 3.140 3.491 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.935 3.353 apply_single 119 13.6 0.000 0.000 2.935 3.353 calculate_first_density_matrix 1 7.0 0.000 0.000 3.122 3.325 acc_transpose_blocks 54864 15.5 0.231 0.255 2.273 2.966 multiply_cannon_metrocomm1 54864 15.5 0.058 0.064 1.851 2.946 ot_diis_step 108 11.5 0.006 0.007 2.758 2.928 jit_kernel_multiply 13 15.8 2.109 2.885 2.109 2.885 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.485 2.701 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.361 2.506 wfi_extrapolate 11 7.9 0.001 0.001 2.152 2.285 density_rs2pw 119 9.7 0.004 0.004 2.097 2.275 grid_integrate_task_list 119 12.3 2.062 2.249 2.062 2.249 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.071 2.235 init_scf_loop 11 6.9 0.000 0.000 2.052 2.178 mp_sum_d 4139 12.0 1.427 2.162 1.427 2.162 potential_pw2rs 119 12.3 0.004 0.004 1.808 1.930 make_m2s 4572 13.5 0.054 0.058 1.683 1.819 make_images 4572 14.5 0.138 0.149 1.599 1.731 fft_wrap_pw1pw2 1201 11.6 0.011 0.013 1.532 1.703 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.588 1.689 transfer_rs2pw 487 10.6 0.006 0.007 1.447 1.613 mp_alltoall_d11v 2130 13.8 1.334 1.605 1.334 1.605 acc_transpose_blocks_sync 164592 16.5 1.225 1.551 1.225 1.551 grid_collocate_task_list 119 9.7 1.375 1.533 1.375 1.533 mp_waitany 12084 13.8 1.315 1.529 1.315 1.529 fft3d_ps 1201 13.6 0.373 0.504 1.303 1.457 transfer_pw2rs 487 13.2 0.005 0.007 1.354 1.446 fft_wrap_pw1pw2_140 487 12.2 0.049 0.057 1.182 1.331 dbcsr_dot_sd 1205 11.9 0.048 0.061 0.803 1.308 parallel_gemm_fm 81 9.0 0.000 0.000 1.221 1.291 parallel_gemm_fm_cosma 81 10.0 1.221 1.291 1.221 1.291 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=58.824000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=426.272727, yerr=1.052349 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/47690d86814a0e25b6e15635e68dedc68a09033e_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 484.171776E+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 649770. 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 0.015 0.028 40.918 41.876 qs_mol_dyn_low 1 2.0 0.003 0.003 40.593 41.543 qs_forces 11 3.9 0.003 0.004 40.510 41.459 qs_energies 11 4.9 0.012 0.031 38.799 39.707 scf_env_do_scf 11 5.9 0.000 0.001 32.823 33.591 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 30.259 30.967 dbcsr_multiply_generic 2286 12.5 0.110 0.119 22.742 23.589 qs_scf_new_mos 108 7.5 0.001 0.001 21.335 21.870 qs_scf_loop_do_ot 108 8.5 0.001 0.001 21.335 21.870 ot_scf_mini 108 9.5 0.002 0.003 20.394 20.863 velocity_verlet 10 3.0 0.003 0.011 18.966 19.412 multiply_cannon 2286 13.5 0.209 0.217 17.342 19.046 multiply_cannon_loop 2286 14.5 1.182 1.245 16.114 17.678 ot_mini 108 10.5 0.001 0.001 12.493 12.910 mp_waitall_1 200699 16.5 5.934 11.095 5.934 11.095 qs_ot_get_derivative 108 11.5 0.001 0.001 10.108 10.424 multiply_cannon_metrocomm3 27432 15.5 0.074 0.077 4.245 9.630 multiply_cannon_multrec 27432 15.5 1.811 4.099 6.309 9.225 rebuild_ks_matrix 119 8.3 0.000 0.000 7.008 7.316 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.008 7.316 dbcsr_mm_accdrv_process 47894 16.0 3.601 5.954 4.419 6.714 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.168 6.445 qs_ot_get_p 119 10.4 0.001 0.001 4.943 5.240 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.881 4.746 init_scf_run 11 5.9 0.000 0.002 4.591 4.698 scf_env_initial_rho_setup 11 6.9 0.005 0.010 4.590 4.698 mp_sum_l 7287 12.8 2.114 4.118 2.114 4.118 sum_up_and_integrate 119 10.3 0.001 0.001 3.849 3.944 integrate_v_rspace 119 11.3 0.002 0.003 3.835 3.929 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.770 3.858 calculate_rho_elec 119 8.7 0.021 0.025 3.769 3.857 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.065 3.738 apply_single 119 13.6 0.000 0.000 3.065 3.737 qs_ot_p2m_diag 50 11.0 0.008 0.012 3.197 3.288 make_m2s 4572 13.5 0.052 0.054 2.822 3.134 calculate_first_density_matrix 1 7.0 0.000 0.000 2.990 3.066 make_images 4572 14.5 0.207 0.249 2.732 3.046 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.785 2.850 init_scf_loop 11 6.9 0.000 0.000 2.537 2.596 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.385 2.506 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.420 2.477 cp_fm_redistribute_end 50 14.0 1.226 2.435 1.231 2.441 ot_diis_step 108 11.5 0.011 0.011 2.337 2.392 cp_fm_diag_elpa_base 50 14.0 1.119 2.281 1.149 2.321 calculate_dm_sparse 119 9.5 0.000 0.001 2.178 2.260 multiply_cannon_sync_h2d 27432 15.5 1.702 2.238 1.702 2.238 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.183 2.234 density_rs2pw 119 9.7 0.004 0.004 2.059 2.133 acc_transpose_blocks 27432 15.5 0.117 0.124 1.641 1.994 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.901 1.984 grid_integrate_task_list 119 12.3 1.846 1.919 1.846 1.919 fft_wrap_pw1pw2 1201 11.6 0.012 0.014 1.760 1.833 make_images_data 4572 15.5 0.048 0.054 1.286 1.689 jit_kernel_multiply 8 16.4 0.759 1.665 0.759 1.665 prepare_preconditioner 11 7.9 0.000 0.000 1.606 1.654 make_preconditioner 11 8.9 0.000 0.000 1.605 1.654 potential_pw2rs 119 12.3 0.006 0.006 1.575 1.613 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.507 1.580 wfi_extrapolate 11 7.9 0.001 0.001 1.527 1.562 fft3d_ps 1201 13.6 0.513 0.566 1.479 1.541 hybrid_alltoall_any 4725 16.4 0.055 0.117 1.150 1.504 mp_alltoall_d11v 2130 13.8 1.286 1.436 1.286 1.436 fft_wrap_pw1pw2_140 487 12.2 0.047 0.049 1.339 1.405 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.354 1.390 transfer_rs2pw 487 10.6 0.005 0.006 1.293 1.388 grid_collocate_task_list 119 9.7 1.294 1.361 1.294 1.361 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.228 1.276 transfer_pw2rs 487 13.2 0.004 0.005 1.055 1.080 qs_energies_init_hamiltonians 11 5.9 0.041 0.070 1.044 1.073 mp_sum_d 4139 12.0 0.560 1.033 0.560 1.033 mp_allgather_i34 2286 14.5 0.641 1.029 0.641 1.029 acc_transpose_blocks_sync 82296 16.5 0.816 0.944 0.816 0.944 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.898 0.930 acc_transpose_blocks_kernels 27432 16.5 0.190 0.281 0.681 0.912 make_images_sizes 4572 15.5 0.005 0.006 0.647 0.906 mp_alltoall_i44 4572 16.5 0.641 0.901 0.641 0.901 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=41.876000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=461.090909, yerr=1.564059 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/47690d86814a0e25b6e15635e68dedc68a09033e_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 512.970752E+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 0.030 0.076 34.489 34.981 qs_mol_dyn_low 1 2.0 0.003 0.003 33.606 34.085 qs_forces 11 3.9 0.005 0.006 33.547 34.026 qs_energies 11 4.9 0.088 0.090 31.996 32.454 scf_env_do_scf 11 5.9 0.000 0.001 26.017 26.389 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 23.438 23.773 dbcsr_multiply_generic 2286 12.5 0.097 0.100 17.785 18.021 qs_scf_new_mos 108 7.5 0.001 0.001 15.797 16.028 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.797 16.028 velocity_verlet 10 3.0 0.007 0.008 15.264 15.483 ot_scf_mini 108 9.5 0.002 0.002 15.034 15.254 multiply_cannon 2286 13.5 0.193 0.198 14.505 15.169 multiply_cannon_loop 2286 14.5 0.862 0.909 13.657 14.462 ot_mini 108 10.5 0.001 0.001 9.361 9.500 qs_ot_get_derivative 108 11.5 0.001 0.001 7.861 7.981 multiply_cannon_multrec 18288 15.5 1.924 2.919 7.452 7.756 dbcsr_mm_accdrv_process 38222 16.0 5.312 6.315 5.435 6.381 rebuild_ks_matrix 119 8.3 0.000 0.000 6.037 6.119 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.037 6.119 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.325 5.398 init_scf_run 11 5.9 0.001 0.012 4.599 4.665 scf_env_initial_rho_setup 11 6.9 0.003 0.012 4.598 4.664 mp_waitall_1 158411 16.6 2.821 3.868 2.821 3.868 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.100 3.688 sum_up_and_integrate 119 10.3 0.001 0.001 3.508 3.558 integrate_v_rspace 119 11.3 0.003 0.003 3.495 3.546 qs_ot_get_p 119 10.4 0.002 0.002 3.409 3.487 calculate_first_density_matrix 1 7.0 0.000 0.000 3.370 3.418 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.206 3.251 calculate_rho_elec 119 8.7 0.031 0.032 3.206 3.250 multiply_cannon_metrocomm3 18288 15.5 0.049 0.051 1.658 2.651 init_scf_loop 11 6.9 0.000 0.000 2.561 2.597 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.999 2.409 apply_single 119 13.6 0.000 0.000 1.999 2.409 qs_ot_p2m_diag 50 11.0 0.012 0.012 2.275 2.304 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.160 2.192 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.003 2.032 make_m2s 4572 13.5 0.044 0.045 1.855 2.030 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.974 2.002 calculate_dm_sparse 119 9.5 0.000 0.001 1.911 1.943 make_images 4572 14.5 0.194 0.206 1.768 1.943 grid_integrate_task_list 119 12.3 1.829 1.909 1.829 1.909 density_rs2pw 119 9.7 0.003 0.004 1.764 1.860 acc_transpose_blocks 18288 15.5 0.081 0.084 1.696 1.809 prepare_preconditioner 11 7.9 0.000 0.000 1.751 1.778 make_preconditioner 11 8.9 0.000 0.000 1.751 1.778 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.739 1.773 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.722 1.753 cp_fm_diag_elpa_base 50 14.0 1.699 1.734 1.721 1.751 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.609 1.701 fft_wrap_pw1pw2 1201 11.6 0.012 0.014 1.639 1.667 ot_diis_step 108 11.5 0.011 0.012 1.486 1.507 mp_sum_l 7287 12.8 1.066 1.480 1.066 1.480 fft3d_ps 1201 13.6 0.539 0.566 1.322 1.348 grid_collocate_task_list 119 9.7 1.242 1.340 1.242 1.340 fft_wrap_pw1pw2_140 487 12.2 0.061 0.063 1.314 1.338 potential_pw2rs 119 12.3 0.007 0.008 1.297 1.317 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.244 1.264 multiply_cannon_sync_h2d 18288 15.5 1.061 1.240 1.061 1.240 wfi_extrapolate 11 7.9 0.001 0.001 1.175 1.191 transfer_rs2pw 487 10.6 0.005 0.005 1.009 1.113 jit_kernel_multiply 4 16.0 0.064 1.017 0.064 1.017 qs_energies_init_hamiltonians 11 5.9 0.003 0.004 0.981 0.996 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.957 0.986 parallel_gemm_fm 81 9.0 0.000 0.000 0.963 0.983 parallel_gemm_fm_cosma 81 10.0 0.963 0.982 0.963 0.982 make_images_data 4572 15.5 0.048 0.052 0.787 0.959 acc_transpose_blocks_kernels 18288 16.5 0.219 0.227 0.851 0.902 make_basis_sm 11 9.8 0.000 0.000 0.867 0.879 hybrid_alltoall_any 4725 16.4 0.060 0.118 0.673 0.874 transfer_pw2rs 487 13.2 0.004 0.004 0.833 0.847 acc_transpose_blocks_sync 54864 16.5 0.744 0.835 0.744 0.835 mp_alltoall_d11v 2130 13.8 0.665 0.801 0.665 0.801 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.782 0.793 multiply_cannon_metrocomm1 18288 15.5 0.031 0.032 0.398 0.769 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.677 0.743 cp_fm_cholesky_invert 11 10.9 0.717 0.729 0.717 0.729 mp_alltoall_z22v 1201 15.6 0.634 0.708 0.634 0.708 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=34.981000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=489.454545, yerr=0.782030 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/47690d86814a0e25b6e15635e68dedc68a09033e_performance_tests/06/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 114.044384E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3805952 0.0% 0.0% 100.0% average stack size 0.0 0.0 38.6 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 559.419392E+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 0.060 0.115 39.627 42.845 qs_mol_dyn_low 1 2.0 0.003 0.004 39.003 42.169 qs_forces 11 3.9 0.029 0.044 38.915 42.074 qs_energies 11 4.9 0.001 0.001 37.120 40.127 scf_env_do_scf 11 5.9 0.000 0.001 30.307 32.767 scf_env_do_scf_inner_loop 108 6.5 0.003 0.008 26.760 28.932 dbcsr_multiply_generic 2286 12.5 0.127 0.136 21.728 23.578 velocity_verlet 10 3.0 0.006 0.018 18.966 20.509 qs_scf_new_mos 108 7.5 0.001 0.001 18.580 20.136 qs_scf_loop_do_ot 108 8.5 0.001 0.001 18.580 20.135 multiply_cannon 2286 13.5 0.226 0.247 17.869 19.430 ot_scf_mini 108 9.5 0.002 0.003 17.514 18.927 multiply_cannon_loop 2286 14.5 1.587 1.790 16.891 18.403 ot_mini 108 10.5 0.001 0.001 10.856 11.735 multiply_cannon_multrec 27432 15.5 2.582 3.298 9.827 10.550 qs_ot_get_derivative 108 11.5 0.001 0.001 8.921 9.637 dbcsr_mm_accdrv_process 47916 15.9 6.597 8.904 7.141 8.994 rebuild_ks_matrix 119 8.3 0.000 0.000 6.500 7.043 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 6.499 7.042 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.757 6.239 init_scf_run 11 5.9 0.001 0.017 5.368 5.803 scf_env_initial_rho_setup 11 6.9 0.012 0.023 5.367 5.803 calculate_first_density_matrix 1 7.0 0.000 0.000 3.929 4.241 qs_ot_get_p 119 10.4 0.002 0.007 3.722 4.031 qs_ot_get_derivative_taylor 59 13.0 0.001 0.002 3.274 3.869 init_scf_loop 11 6.9 0.000 0.000 3.527 3.814 sum_up_and_integrate 119 10.3 0.001 0.001 3.530 3.813 integrate_v_rspace 119 11.3 0.003 0.005 3.518 3.801 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.333 3.615 calculate_rho_elec 119 8.7 0.042 0.052 3.333 3.615 mp_waitall_1 137007 16.6 2.360 3.155 2.360 3.155 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.837 3.067 acc_transpose_blocks 27432 15.5 0.124 0.135 2.599 2.945 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.197 2.897 apply_single 119 13.6 0.000 0.000 2.196 2.897 prepare_preconditioner 11 7.9 0.000 0.000 2.634 2.848 make_preconditioner 11 8.9 0.000 0.000 2.634 2.848 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.230 2.734 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.503 2.713 calculate_dm_sparse 119 9.5 0.000 0.001 2.437 2.689 make_m2s 4572 13.5 0.056 0.061 2.366 2.644 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.412 2.605 make_images 4572 14.5 0.284 0.370 2.253 2.524 qs_ot_p2m_diag 50 11.0 0.016 0.024 2.235 2.417 grid_integrate_task_list 119 12.3 1.900 2.100 1.900 2.100 jit_kernel_multiply 8 16.1 0.475 2.090 0.475 2.090 ot_diis_step 108 11.5 0.013 0.014 1.892 2.046 cp_dbcsr_syevd 50 12.0 0.003 0.004 1.824 1.971 density_rs2pw 119 9.7 0.004 0.004 1.735 1.940 fft_wrap_pw1pw2 1201 11.6 0.012 0.015 1.760 1.934 mp_sum_l 7287 12.8 1.040 1.923 1.040 1.923 acc_transpose_blocks_sync 82296 16.5 1.489 1.741 1.489 1.741 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.582 1.718 cp_fm_diag_elpa_base 50 14.0 1.547 1.688 1.580 1.715 multiply_cannon_metrocomm3 27432 15.5 0.043 0.048 1.051 1.658 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.504 1.630 fft_wrap_pw1pw2_140 487 12.2 0.074 0.080 1.460 1.612 fft3d_ps 1201 13.6 0.601 0.705 1.407 1.552 wfi_extrapolate 11 7.9 0.001 0.001 1.369 1.480 grid_collocate_task_list 119 9.7 1.302 1.460 1.302 1.460 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.335 1.442 potential_pw2rs 119 12.3 0.008 0.010 1.248 1.349 cp_fm_upper_to_full 72 14.2 0.872 1.307 0.872 1.307 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.166 1.271 dbcsr_complete_redistribute 329 12.2 0.147 0.265 0.944 1.236 transfer_rs2pw 487 10.6 0.005 0.006 0.937 1.110 make_images_data 4572 15.5 0.050 0.054 0.859 1.052 acc_transpose_blocks_kernels 27432 16.5 0.283 0.309 0.956 1.041 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.740 1.019 multiply_cannon_metrocomm1 27432 15.5 0.037 0.040 0.501 1.014 hybrid_alltoall_any 4725 16.4 0.069 0.158 0.724 0.959 mp_alltoall_d11v 2130 13.8 0.732 0.922 0.732 0.922 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.835 0.917 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.834 0.903 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=42.845000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=531.454545, yerr=4.207431 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/47690d86814a0e25b6e15635e68dedc68a09033e_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 625.287168E+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 0.152 0.176 38.502 40.260 qs_mol_dyn_low 1 2.0 0.003 0.003 37.751 39.475 qs_forces 11 3.9 0.014 0.023 35.190 36.797 qs_energies 11 4.9 0.002 0.003 33.368 34.894 scf_env_do_scf 11 5.9 0.000 0.001 26.699 27.918 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 23.399 24.468 velocity_verlet 10 3.0 0.003 0.003 19.547 20.604 dbcsr_multiply_generic 2286 12.5 0.099 0.101 15.721 16.408 qs_scf_new_mos 108 7.5 0.001 0.001 14.777 15.474 qs_scf_loop_do_ot 108 8.5 0.001 0.001 14.776 15.473 ot_scf_mini 108 9.5 0.002 0.002 13.977 14.624 multiply_cannon 2286 13.5 0.228 0.239 12.217 12.786 multiply_cannon_loop 2286 14.5 0.653 0.676 11.092 11.351 ot_mini 108 10.5 0.001 0.001 7.652 8.024 multiply_cannon_multrec 9144 15.5 1.713 1.882 6.538 6.843 rebuild_ks_matrix 119 8.3 0.000 0.000 6.446 6.712 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.445 6.712 qs_ot_get_derivative 108 11.5 0.001 0.001 6.198 6.490 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.750 5.989 init_scf_run 11 5.9 0.000 0.002 5.030 5.259 scf_env_initial_rho_setup 11 6.9 0.002 0.005 5.029 5.259 dbcsr_mm_accdrv_process 12550 15.8 3.854 4.971 4.715 5.014 qs_ot_get_p 119 10.4 0.002 0.003 4.211 4.421 sum_up_and_integrate 119 10.3 0.001 0.001 3.903 4.081 integrate_v_rspace 119 11.3 0.003 0.003 3.893 4.071 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.844 4.016 calculate_rho_elec 119 8.7 0.060 0.065 3.844 4.016 calculate_first_density_matrix 1 7.0 0.000 0.000 3.647 3.812 mp_waitall_1 115863 16.7 2.770 3.443 2.770 3.443 init_scf_loop 11 6.9 0.000 0.000 3.268 3.417 qs_ot_p2m_diag 50 11.0 0.022 0.023 2.813 2.939 mp_sum_dm 438 4.9 2.428 2.707 2.428 2.707 update_particle_set 20 4.0 0.000 0.000 2.385 2.658 make_m2s 4572 13.5 0.034 0.035 2.361 2.646 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.516 2.632 make_images 4572 14.5 0.272 0.301 2.269 2.554 md_write_output 11 3.9 0.003 0.034 0.159 2.518 md_output 10 3.0 0.000 0.000 0.160 2.516 prepare_preconditioner 11 7.9 0.000 0.000 2.382 2.495 make_preconditioner 11 8.9 0.000 0.000 2.382 2.495 write_trajectory 44 4.9 0.001 0.007 0.156 2.483 write_particle_coordinates 11 5.9 0.155 2.477 0.155 2.477 calculate_dm_sparse 119 9.5 0.000 0.000 2.330 2.450 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.252 2.363 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.177 2.283 density_rs2pw 119 9.7 0.003 0.003 2.134 2.267 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.162 2.262 fft_wrap_pw1pw2 1201 11.6 0.012 0.014 2.115 2.229 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.090 2.181 cp_fm_diag_elpa_base 50 14.0 2.058 2.154 2.087 2.177 grid_integrate_task_list 119 12.3 1.916 1.973 1.916 1.973 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.894 1.971 fft3d_ps 1201 13.6 0.648 0.689 1.723 1.816 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.694 1.776 fft_wrap_pw1pw2_140 487 12.2 0.095 0.100 1.673 1.767 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.648 1.724 acc_transpose_blocks 9144 15.5 0.043 0.044 1.582 1.652 potential_pw2rs 119 12.3 0.010 0.011 1.576 1.639 ot_diis_step 108 11.5 0.013 0.013 1.431 1.497 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.336 1.439 apply_single 119 13.6 0.000 0.000 1.336 1.439 grid_collocate_task_list 119 9.7 1.337 1.403 1.337 1.403 cp_fm_cholesky_invert 11 10.9 1.321 1.381 1.321 1.381 wfi_extrapolate 11 7.9 0.001 0.001 1.320 1.381 make_images_data 4572 15.5 0.043 0.047 1.054 1.378 qs_energies_init_hamiltonians 11 5.9 0.024 0.044 1.311 1.374 jit_kernel_multiply 6 15.4 0.821 1.363 0.821 1.363 hybrid_alltoall_any 4725 16.4 0.067 0.178 1.025 1.359 mp_alltoall_d11v 2130 13.8 1.107 1.248 1.107 1.248 multiply_cannon_metrocomm3 9144 15.5 0.021 0.021 0.622 1.230 transfer_rs2pw 487 10.6 0.004 0.005 1.129 1.199 multiply_cannon_metrocomm1 9144 15.5 0.024 0.025 0.851 1.058 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.990 1.028 mp_alltoall_z22v 1201 15.6 0.949 1.009 0.949 1.009 parallel_gemm_fm 81 9.0 0.000 0.000 0.941 0.983 parallel_gemm_fm_cosma 81 10.0 0.941 0.983 0.941 0.983 transfer_pw2rs 487 13.2 0.003 0.004 0.947 0.982 mp_allgather_i34 2286 14.5 0.428 0.967 0.428 0.967 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.885 0.949 mp_sum_l 7287 12.8 0.686 0.933 0.686 0.933 make_basis_sm 11 9.8 0.000 0.000 0.863 0.902 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.828 0.862 dbcsr_complete_redistribute 329 12.2 0.174 0.220 0.766 0.824 acc_transpose_blocks_kernels 9144 16.5 0.120 0.124 0.790 0.824 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=40.260000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=592.545455, yerr=7.475580 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/47690d86814a0e25b6e15635e68dedc68a09033e_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 784.875520E+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 0.089 0.121 47.496 47.989 qs_mol_dyn_low 1 2.0 0.003 0.003 46.637 47.140 qs_forces 11 3.9 0.005 0.006 46.573 47.056 qs_energies 11 4.9 0.010 0.011 44.505 44.972 scf_env_do_scf 11 5.9 0.001 0.001 37.382 37.770 scf_env_do_scf_inner_loop 108 6.5 0.004 0.008 28.969 29.266 velocity_verlet 10 3.0 0.002 0.004 25.321 25.577 dbcsr_multiply_generic 2286 12.5 0.106 0.108 20.795 21.158 qs_scf_new_mos 108 7.5 0.001 0.001 19.244 19.523 qs_scf_loop_do_ot 108 8.5 0.001 0.001 19.243 19.522 ot_scf_mini 108 9.5 0.002 0.002 18.020 18.291 multiply_cannon 2286 13.5 0.301 0.307 16.190 17.299 multiply_cannon_loop 2286 14.5 0.868 0.887 14.771 15.562 ot_mini 108 10.5 0.001 0.001 10.865 11.086 multiply_cannon_multrec 9144 15.5 3.641 5.310 9.471 9.661 qs_ot_get_derivative 108 11.5 0.001 0.001 8.689 8.868 init_scf_loop 11 6.9 0.000 0.000 8.378 8.468 rebuild_ks_matrix 119 8.3 0.000 0.000 7.389 7.620 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.389 7.619 prepare_preconditioner 11 7.9 0.000 0.000 7.305 7.391 make_preconditioner 11 8.9 0.000 0.000 7.305 7.391 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.823 7.256 dbcsr_mm_accdrv_process 12550 15.8 5.163 6.961 5.697 6.994 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.655 6.862 init_scf_run 11 5.9 0.000 0.002 4.900 4.951 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.900 4.949 cp_fm_upper_to_full 72 14.2 3.212 4.675 3.212 4.675 qs_ot_get_p 119 10.4 0.007 0.015 4.103 4.281 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.168 4.210 calculate_rho_elec 119 8.7 0.119 0.124 4.168 4.210 sum_up_and_integrate 119 10.3 0.001 0.001 3.932 3.978 integrate_v_rspace 119 11.3 0.004 0.004 3.921 3.968 mp_waitall_1 94719 16.7 2.751 3.738 2.751 3.738 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.951 3.391 calculate_first_density_matrix 1 7.0 0.000 0.000 3.326 3.360 dbcsr_complete_redistribute 329 12.2 0.283 0.294 2.099 2.944 make_m2s 4572 13.5 0.038 0.038 2.719 2.923 make_images 4572 14.5 0.354 0.386 2.598 2.801 fft_wrap_pw1pw2 1201 11.6 0.014 0.014 2.589 2.616 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.746 2.589 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.349 2.559 apply_single 119 13.6 0.000 0.000 2.349 2.559 calculate_dm_sparse 119 9.5 0.000 0.000 2.478 2.517 multiply_cannon_metrocomm3 9144 15.5 0.021 0.022 1.520 2.499 qs_ot_p2m_diag 50 11.0 0.042 0.043 2.471 2.496 mp_alltoall_i22 627 13.8 1.475 2.363 1.475 2.363 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.328 2.352 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.476 2.308 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.232 2.303 density_rs2pw 119 9.7 0.003 0.003 2.136 2.204 grid_integrate_task_list 119 12.3 2.140 2.203 2.140 2.203 fft_wrap_pw1pw2_140 487 12.2 0.181 0.183 2.140 2.167 ot_diis_step 108 11.5 0.014 0.014 2.140 2.163 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.040 2.112 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.088 2.109 mp_sum_l 7287 12.8 1.348 2.065 1.348 2.065 fft3d_ps 1201 13.6 0.867 0.888 2.035 2.058 acc_transpose_blocks 9144 15.5 0.045 0.045 1.926 1.960 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.874 1.894 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.742 1.760 cp_fm_cholesky_invert 11 10.9 1.741 1.758 1.741 1.758 cp_fm_diag_elpa_base 50 14.0 1.590 1.659 1.739 1.757 grid_collocate_task_list 119 9.7 1.564 1.602 1.564 1.602 wfi_extrapolate 11 7.9 0.001 0.001 1.510 1.525 hybrid_alltoall_any 4725 16.4 0.092 0.155 1.253 1.515 make_images_data 4572 15.5 0.046 0.049 1.222 1.470 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.408 1.450 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.422 1.444 potential_pw2rs 119 12.3 0.014 0.014 1.351 1.367 mp_alltoall_d11v 2130 13.8 1.275 1.319 1.275 1.319 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.195 1.211 acc_transpose_blocks_sync 27432 16.5 1.109 1.140 1.109 1.140 mp_alltoall_z22v 1201 15.6 1.033 1.073 1.033 1.073 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.005 1.046 jit_kernel_multiply 5 15.8 0.505 0.994 0.505 0.994 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.952 0.977 qs_create_task_list 11 7.9 0.006 0.013 0.959 0.975 generate_qs_task_list 11 8.9 0.371 0.391 0.953 0.970 make_basis_sm 11 9.8 0.000 0.000 0.951 0.961 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=47.989000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=739.363636, yerr=11.795783 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/47690d86814a0e25b6e15635e68dedc68a09033e_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 499.589120E+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 910155. 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 0.087 0.225 93.245 94.714 qs_mol_dyn_low 1 2.0 0.067 0.158 92.229 93.735 qs_forces 11 3.9 0.003 0.004 91.937 93.351 qs_energies 11 4.9 0.015 0.114 88.909 90.274 scf_env_do_scf 11 5.9 0.001 0.002 77.367 78.637 scf_env_do_scf_inner_loop 99 6.5 0.008 0.044 70.361 71.512 dbcsr_multiply_generic 2055 12.4 0.116 0.122 55.238 56.464 qs_scf_new_mos 99 7.5 0.000 0.001 51.636 52.484 qs_scf_loop_do_ot 99 8.5 0.001 0.001 51.635 52.484 ot_scf_mini 99 9.5 0.002 0.003 49.071 49.849 velocity_verlet 10 3.0 0.055 0.140 47.364 48.135 multiply_cannon 2055 13.4 0.179 0.187 45.217 47.004 multiply_cannon_loop 2055 14.4 1.793 1.843 43.943 45.732 ot_mini 99 10.5 0.001 0.001 28.550 29.041 qs_ot_get_derivative 99 11.5 0.001 0.001 21.533 21.912 multiply_cannon_multrec 49320 15.4 11.232 12.017 17.936 18.900 rebuild_ks_matrix 110 8.3 0.000 0.000 15.556 16.032 qs_ks_build_kohn_sham_matrix 110 9.3 0.015 0.046 15.556 16.032 mp_waitall_1 220248 16.4 13.123 14.295 13.123 14.295 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.689 14.117 qs_ot_get_p 110 10.4 0.002 0.013 11.644 11.873 multiply_cannon_sync_h2d 49320 15.4 9.393 10.093 9.393 10.093 init_scf_run 11 5.9 0.000 0.002 8.872 9.017 scf_env_initial_rho_setup 11 6.9 0.004 0.019 8.872 9.017 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 8.205 8.867 multiply_cannon_metrocomm3 49320 15.4 0.085 0.088 6.962 8.441 qs_ot_p2m_diag 48 11.0 0.012 0.018 7.950 8.117 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.343 7.861 apply_single 110 13.6 0.000 0.000 7.342 7.861 sum_up_and_integrate 110 10.3 0.002 0.003 7.652 7.789 integrate_v_rspace 110 11.3 0.003 0.003 7.627 7.765 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.267 7.437 calculate_rho_elec 110 8.6 0.021 0.034 7.267 7.436 dbcsr_mm_accdrv_process 87628 16.1 3.105 3.189 6.577 7.394 init_scf_loop 11 6.9 0.004 0.034 6.964 7.084 cp_dbcsr_syevd 48 12.0 0.002 0.003 6.917 7.030 ot_diis_step 99 11.5 0.006 0.006 6.756 6.867 cp_fm_diag_elpa 48 13.0 0.000 0.000 6.163 6.277 cp_fm_diag_elpa_base 48 14.0 6.137 6.253 6.159 6.274 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.033 6.163 mp_sum_l 6594 12.7 4.936 5.787 4.936 5.787 calculate_dm_sparse 110 9.5 0.001 0.001 4.820 4.980 make_m2s 4110 13.4 0.062 0.066 4.778 4.963 multiply_cannon_metrocomm1 49320 15.4 0.065 0.067 3.816 4.936 make_images 4110 14.4 0.182 0.195 4.680 4.867 prepare_preconditioner 11 7.9 0.000 0.000 4.497 4.576 make_preconditioner 11 8.9 0.010 0.075 4.497 4.576 wfi_extrapolate 11 7.9 0.002 0.013 4.379 4.451 calculate_first_density_matrix 1 7.0 0.002 0.019 4.365 4.444 make_full_inverse_cholesky 11 9.9 0.000 0.000 4.196 4.313 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 4.216 4.288 density_rs2pw 110 9.6 0.004 0.004 3.976 4.180 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.681 3.824 jit_kernel_multiply 13 15.9 3.196 3.823 3.196 3.823 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.529 3.631 fft_wrap_pw1pw2 1111 11.6 0.014 0.016 3.415 3.551 grid_integrate_task_list 110 12.3 3.270 3.453 3.270 3.453 mp_alltoall_d11v 2046 13.8 2.827 3.327 2.827 3.327 potential_pw2rs 110 12.3 0.006 0.006 3.022 3.082 fft3d_ps 1111 13.6 0.740 0.831 2.957 3.076 fft_wrap_pw1pw2_140 451 12.1 0.091 0.098 2.746 2.872 transfer_rs2pw 451 10.6 0.006 0.006 2.493 2.866 acc_transpose_blocks 49320 15.4 0.215 0.225 2.721 2.778 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 2.556 2.606 mp_waitany 14300 13.8 2.210 2.586 2.210 2.586 cp_fm_cholesky_invert 11 10.9 2.437 2.480 2.437 2.480 mp_sum_d 3893 11.9 1.744 2.402 1.744 2.402 grid_collocate_task_list 110 9.6 2.167 2.282 2.167 2.282 make_images_data 4110 15.4 0.045 0.049 2.047 2.240 transfer_pw2rs 451 13.1 0.006 0.006 2.081 2.137 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.032 2.083 hybrid_alltoall_any 4261 16.3 0.086 0.482 1.789 2.067 mp_alltoall_z22v 1111 15.6 1.849 2.033 1.849 2.033 qs_energies_init_hamiltonians 11 5.9 0.012 0.032 1.866 1.957 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=94.714000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=473.000000, yerr=3.219260 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/47690d86814a0e25b6e15635e68dedc68a09033e_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 593.162240E+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 2833836. 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 0.046 0.070 83.193 101.391 qs_mol_dyn_low 1 2.0 0.003 0.004 82.628 100.702 qs_forces 11 3.9 0.004 0.005 82.548 100.605 qs_energies 11 4.9 0.001 0.002 78.977 96.254 scf_env_do_scf 11 5.9 0.000 0.001 68.711 83.743 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 58.374 71.142 velocity_verlet 10 3.0 0.006 0.011 42.342 51.605 dbcsr_multiply_generic 2055 12.4 0.135 0.168 42.261 51.509 qs_scf_new_mos 99 7.5 0.001 0.001 40.144 48.984 qs_scf_loop_do_ot 99 8.5 0.001 0.001 40.144 48.984 ot_scf_mini 99 9.5 0.003 0.003 38.267 46.721 multiply_cannon 2055 13.4 0.230 0.301 33.270 40.219 multiply_cannon_loop 2055 14.4 1.178 1.451 31.683 38.282 ot_mini 99 10.5 0.001 0.001 21.475 26.252 multiply_cannon_multrec 24660 15.4 6.958 10.187 14.554 19.253 qs_ot_get_derivative 99 11.5 0.001 0.001 15.393 18.835 rebuild_ks_matrix 110 8.3 0.000 0.001 15.006 18.279 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.015 15.005 18.278 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.229 16.116 mp_waitall_1 176588 16.5 9.393 14.044 9.393 14.044 init_scf_loop 11 6.9 0.000 0.000 10.287 12.536 qs_ot_get_p 110 10.4 0.001 0.002 9.588 11.810 multiply_cannon_metrocomm3 24660 15.4 0.075 0.090 5.934 10.926 prepare_preconditioner 11 7.9 0.000 0.000 8.077 9.849 make_preconditioner 11 8.9 0.000 0.000 8.077 9.849 init_scf_run 11 5.9 0.004 0.160 7.734 9.426 scf_env_initial_rho_setup 11 6.9 0.075 0.316 7.730 9.426 make_full_inverse_cholesky 11 9.9 0.000 0.000 7.560 9.404 dbcsr_mm_accdrv_process 52282 16.1 5.861 7.203 7.429 9.177 sum_up_and_integrate 110 10.3 0.001 0.003 7.508 9.146 integrate_v_rspace 110 11.3 0.003 0.003 7.480 9.115 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.102 8.651 calculate_rho_elec 110 8.6 0.040 0.047 7.102 8.650 qs_ot_p2m_diag 48 11.0 0.029 0.053 7.058 8.620 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.617 8.408 apply_single 110 13.6 0.000 0.001 6.617 8.408 cp_dbcsr_syevd 48 12.0 0.003 0.004 6.466 7.882 multiply_cannon_sync_h2d 24660 15.4 6.327 7.689 6.327 7.689 ot_diis_step 99 11.5 0.011 0.013 5.909 7.202 qs_ot_get_derivative_taylor 52 13.0 0.001 0.002 5.574 7.088 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.580 6.811 cp_fm_diag_elpa_base 48 14.0 5.519 6.736 5.577 6.809 make_m2s 4110 13.4 0.058 0.073 5.210 6.350 make_images 4110 14.4 0.412 0.535 5.092 6.202 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 4.263 5.238 cp_fm_cholesky_invert 11 10.9 4.270 5.224 4.270 5.224 density_rs2pw 110 9.6 0.004 0.005 4.260 5.204 wfi_extrapolate 11 7.9 0.001 0.001 3.946 4.809 fft_wrap_pw1pw2 1111 11.6 0.015 0.020 3.864 4.721 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.760 4.579 calculate_first_density_matrix 1 7.0 0.000 0.000 3.589 4.469 mp_sum_l 6594 12.7 3.127 4.361 3.127 4.361 calculate_dm_sparse 110 9.5 0.001 0.001 3.441 4.222 grid_integrate_task_list 110 12.3 3.197 4.174 3.197 4.174 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.301 3.999 fft3d_ps 1111 13.6 1.096 1.286 3.238 3.958 potential_pw2rs 110 12.3 0.008 0.010 3.210 3.881 fft_wrap_pw1pw2_140 451 12.1 0.112 0.142 2.957 3.616 make_images_data 4110 15.4 0.050 0.061 2.568 3.485 hybrid_alltoall_any 4261 16.3 0.108 0.453 2.210 3.021 qs_ot_get_orbitals 99 10.5 0.001 0.001 2.425 2.949 transfer_rs2pw 451 10.6 0.007 0.009 2.330 2.823 mp_alltoall_d11v 2046 13.8 2.207 2.754 2.207 2.754 grid_collocate_task_list 110 9.6 2.215 2.743 2.215 2.743 acc_transpose_blocks 24660 15.4 0.120 0.143 2.210 2.674 transfer_pw2rs 451 13.1 0.005 0.007 2.063 2.521 cp_fm_cholesky_decompose 22 10.9 2.045 2.484 2.045 2.484 multiply_cannon_metrocomm4 22605 15.4 0.081 0.100 0.830 2.432 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.947 2.366 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.940 2.366 mp_alltoall_z22v 1111 15.6 1.812 2.364 1.812 2.364 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.821 2.232 dbcsr_complete_redistribute 325 12.2 0.255 0.351 1.545 2.202 mp_irecv_dv 57340 16.2 0.695 2.193 0.695 2.193 mp_sum_d 3893 11.9 1.436 2.155 1.436 2.155 jit_kernel_multiply 8 16.4 1.187 2.071 1.187 2.071 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=101.391000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=560.363636, yerr=7.486627 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/47690d86814a0e25b6e15635e68dedc68a09033e_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 664.416256E+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 0.082 0.199 65.394 71.468 qs_mol_dyn_low 1 2.0 0.012 0.072 64.328 70.299 qs_forces 11 3.9 0.004 0.007 64.082 70.043 qs_energies 11 4.9 0.018 0.133 60.849 66.507 scf_env_do_scf 11 5.9 0.001 0.001 50.643 55.344 scf_env_do_scf_inner_loop 99 6.5 0.013 0.055 41.636 45.501 velocity_verlet 10 3.0 0.018 0.061 33.550 36.672 dbcsr_multiply_generic 2055 12.4 0.121 0.135 30.597 33.222 qs_scf_new_mos 99 7.5 0.001 0.001 26.842 29.418 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.841 29.418 ot_scf_mini 99 9.5 0.003 0.003 25.560 28.034 multiply_cannon 2055 13.4 0.213 0.235 23.814 25.818 multiply_cannon_loop 2055 14.4 0.819 0.901 22.556 24.576 ot_mini 99 10.5 0.001 0.001 14.706 16.175 rebuild_ks_matrix 110 8.3 0.000 0.001 12.465 13.500 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.018 12.464 13.499 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.997 11.910 multiply_cannon_multrec 16440 15.4 3.593 4.500 10.305 11.782 qs_ot_get_derivative 99 11.5 0.001 0.001 10.110 11.153 mp_waitall_1 139946 16.5 7.788 10.699 7.788 10.699 init_scf_loop 11 6.9 0.001 0.007 8.969 9.803 multiply_cannon_metrocomm3 16440 15.4 0.046 0.049 4.808 7.975 init_scf_run 11 5.9 0.000 0.002 7.092 7.751 scf_env_initial_rho_setup 11 6.9 0.004 0.027 7.092 7.751 prepare_preconditioner 11 7.9 0.000 0.000 7.070 7.736 make_preconditioner 11 8.9 0.001 0.004 7.070 7.736 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.348 7.336 dbcsr_mm_accdrv_process 34862 16.1 5.749 6.404 6.558 7.115 sum_up_and_integrate 110 10.3 0.001 0.002 6.304 6.890 integrate_v_rspace 110 11.3 0.003 0.003 6.278 6.861 qs_ot_get_p 110 10.4 0.015 0.026 5.928 6.612 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.668 6.199 calculate_rho_elec 110 8.6 0.059 0.064 5.668 6.199 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.047 5.518 apply_single 110 13.6 0.000 0.000 5.046 5.517 make_m2s 4110 13.4 0.050 0.054 4.511 5.003 ot_diis_step 99 11.5 0.011 0.012 4.558 4.981 make_images 4110 14.4 0.399 0.518 4.393 4.878 qs_ot_p2m_diag 48 11.0 0.042 0.047 4.133 4.515 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.449 4.380 calculate_first_density_matrix 1 7.0 0.003 0.021 3.925 4.294 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.743 4.090 multiply_cannon_sync_h2d 16440 15.4 3.229 3.872 3.229 3.872 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.413 3.730 grid_integrate_task_list 110 12.3 3.234 3.649 3.234 3.649 fft_wrap_pw1pw2 1111 11.6 0.014 0.017 3.177 3.470 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.156 3.445 cp_fm_diag_elpa_base 48 14.0 3.085 3.361 3.154 3.443 wfi_extrapolate 11 7.9 0.001 0.001 3.062 3.347 density_rs2pw 110 9.6 0.004 0.007 2.850 3.235 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.962 3.189 cp_fm_cholesky_invert 11 10.9 2.812 3.076 2.812 3.076 make_images_data 4110 15.4 0.046 0.051 2.546 3.032 fft_wrap_pw1pw2_140 451 12.1 0.127 0.144 2.693 2.949 calculate_dm_sparse 110 9.5 0.001 0.001 2.695 2.926 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.569 2.861 hybrid_alltoall_any 4261 16.3 0.110 0.380 2.235 2.860 qs_energies_init_hamiltonians 11 5.9 0.022 0.031 2.473 2.713 fft3d_ps 1111 13.6 1.095 1.215 2.486 2.708 multiply_cannon_metrocomm4 14385 15.4 0.049 0.054 0.935 2.549 mp_sum_l 6594 12.7 1.797 2.499 1.797 2.499 grid_collocate_task_list 110 9.6 2.242 2.472 2.242 2.472 mp_irecv_dv 48980 15.7 0.857 2.405 0.857 2.405 potential_pw2rs 110 12.3 0.011 0.012 2.093 2.289 dbcsr_complete_redistribute 325 12.2 0.343 0.435 1.631 2.234 mp_alltoall_d11v 2046 13.8 1.903 2.169 1.903 2.169 transfer_rs2pw 451 10.6 0.005 0.006 1.637 1.980 cp_fm_upper_to_full 70 14.2 1.425 1.962 1.425 1.962 acc_transpose_blocks 16440 15.4 0.079 0.083 1.673 1.865 mp_waitany 17072 13.8 1.461 1.826 1.461 1.826 cp_fm_cholesky_decompose 22 10.9 1.666 1.818 1.666 1.818 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.672 1.813 mp_allgather_i34 2055 14.4 0.559 1.791 0.559 1.791 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.606 1.752 copy_fm_to_dbcsr 174 11.2 0.001 0.004 1.146 1.719 qs_env_update_s_mstruct 11 6.9 0.003 0.008 1.498 1.634 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.392 1.520 parallel_gemm_fm 81 9.0 0.000 0.000 1.368 1.499 parallel_gemm_fm_cosma 81 10.0 1.367 1.499 1.367 1.499 make_basis_sm 11 9.8 0.000 0.000 1.369 1.496 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.337 1.463 transfer_rs2pw_140 121 11.5 0.174 0.190 1.154 1.459 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=71.468000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=629.363636, yerr=8.025785 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/47690d86814a0e25b6e15635e68dedc68a09033e_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 734.339072E+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 0.037 0.067 71.837 72.843 qs_mol_dyn_low 1 2.0 0.003 0.003 71.499 72.499 qs_forces 11 3.9 0.006 0.009 71.110 72.106 qs_energies 11 4.9 0.001 0.003 67.634 68.587 scf_env_do_scf 11 5.9 0.000 0.001 58.750 59.575 scf_env_do_scf_inner_loop 99 6.5 0.007 0.013 46.151 46.797 velocity_verlet 10 3.0 0.009 0.013 39.167 39.730 dbcsr_multiply_generic 2055 12.4 0.143 0.195 32.272 32.818 qs_scf_new_mos 99 7.5 0.001 0.001 31.019 31.403 qs_scf_loop_do_ot 99 8.5 0.001 0.001 31.019 31.403 ot_scf_mini 99 9.5 0.002 0.003 29.262 29.643 multiply_cannon 2055 13.4 0.238 0.256 24.350 25.359 multiply_cannon_loop 2055 14.4 1.394 1.451 22.740 23.707 ot_mini 99 10.5 0.001 0.001 15.968 16.215 multiply_cannon_multrec 24660 15.4 4.096 6.794 13.286 14.508 init_scf_loop 11 6.9 0.000 0.000 12.556 12.731 rebuild_ks_matrix 110 8.3 0.000 0.000 12.427 12.704 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.426 12.703 qs_ot_get_derivative 99 11.5 0.001 0.001 11.650 11.847 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.980 11.221 prepare_preconditioner 11 7.9 0.000 0.000 10.720 10.863 make_preconditioner 11 8.9 0.000 0.000 10.720 10.863 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.886 10.490 dbcsr_mm_accdrv_process 52304 16.0 7.884 9.316 9.035 10.072 qs_ot_get_p 110 10.4 0.007 0.020 8.030 8.230 mp_waitall_1 121746 16.5 4.861 6.980 4.861 6.980 sum_up_and_integrate 110 10.3 0.001 0.002 6.461 6.558 integrate_v_rspace 110 11.3 0.003 0.003 6.435 6.530 make_m2s 4110 13.4 0.060 0.062 5.981 6.430 make_images 4110 14.4 0.572 0.701 5.837 6.282 init_scf_run 11 5.9 0.000 0.002 6.084 6.169 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.083 6.169 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.997 6.066 calculate_rho_elec 110 8.6 0.077 0.082 5.996 6.065 qs_ot_p2m_diag 48 11.0 0.054 0.063 5.831 5.914 cp_dbcsr_syevd 48 12.0 0.003 0.003 5.219 5.291 cp_fm_upper_to_full 70 14.2 3.487 5.089 3.487 5.089 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.506 4.572 cp_fm_diag_elpa_base 48 14.0 4.320 4.398 4.502 4.569 ot_diis_step 99 11.5 0.011 0.011 4.268 4.328 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.165 4.322 apply_single 110 13.6 0.000 0.000 4.165 4.322 dbcsr_complete_redistribute 325 12.2 0.450 0.526 2.868 4.050 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.864 3.941 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.671 3.752 fft_wrap_pw1pw2 1111 11.6 0.014 0.017 3.525 3.583 multiply_cannon_metrocomm3 24660 15.4 0.039 0.041 1.570 3.514 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.298 3.471 grid_integrate_task_list 110 12.3 3.301 3.448 3.301 3.448 make_images_data 4110 15.4 0.049 0.053 2.908 3.425 hybrid_alltoall_any 4261 16.3 0.123 0.463 2.487 3.347 cp_fm_cholesky_invert 11 10.9 3.180 3.226 3.180 3.226 calculate_dm_sparse 110 9.5 0.001 0.001 3.151 3.224 density_rs2pw 110 9.6 0.004 0.004 3.058 3.217 wfi_extrapolate 11 7.9 0.001 0.001 3.118 3.162 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.060 3.103 fft_wrap_pw1pw2_140 451 12.1 0.146 0.148 2.943 2.998 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.825 2.991 mp_alltoall_i22 605 13.7 1.721 2.926 1.721 2.926 calculate_first_density_matrix 1 7.0 0.000 0.000 2.863 2.905 fft3d_ps 1111 13.6 1.133 1.179 2.793 2.839 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.599 2.665 multiply_cannon_sync_h2d 24660 15.4 2.370 2.616 2.370 2.616 acc_transpose_blocks 24660 15.4 0.111 0.115 2.446 2.542 grid_collocate_task_list 110 9.6 2.275 2.364 2.275 2.364 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.254 2.288 potential_pw2rs 110 12.3 0.012 0.012 2.204 2.242 mp_alltoall_d11v 2046 13.8 1.974 2.093 1.974 2.093 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.904 1.957 cp_fm_cholesky_decompose 22 10.9 1.855 1.901 1.855 1.901 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.699 1.737 mp_sum_l 6594 12.7 1.227 1.716 1.227 1.716 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.621 1.709 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.655 1.700 mp_allgather_i34 2055 14.4 0.689 1.663 0.689 1.663 mp_alltoall_z22v 1111 15.6 1.445 1.529 1.445 1.529 transfer_rs2pw 451 10.6 0.005 0.006 1.350 1.521 jit_kernel_multiply 7 15.8 0.805 1.500 0.805 1.500 multiply_cannon_metrocomm4 20550 15.4 0.060 0.064 0.858 1.476 acc_transpose_blocks_sync 73980 16.4 1.384 1.468 1.384 1.468 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=72.843000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=696.454545, yerr=8.595598 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/47690d86814a0e25b6e15635e68dedc68a09033e_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 860.577792E+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 0.082 0.102 62.497 69.281 qs_mol_dyn_low 1 2.0 0.003 0.003 62.004 68.733 qs_forces 11 3.9 0.008 0.020 61.927 68.650 qs_energies 11 4.9 0.002 0.003 58.060 64.361 scf_env_do_scf 11 5.9 0.001 0.001 48.082 53.302 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 39.344 43.616 velocity_verlet 10 3.0 0.011 0.014 33.630 37.284 dbcsr_multiply_generic 2055 12.4 0.114 0.125 25.643 28.235 qs_scf_new_mos 99 7.5 0.001 0.001 23.759 26.307 qs_scf_loop_do_ot 99 8.5 0.001 0.001 23.758 26.306 ot_scf_mini 99 9.5 0.002 0.003 22.400 24.834 multiply_cannon 2055 13.4 0.250 0.284 19.212 21.196 multiply_cannon_loop 2055 14.4 0.636 0.715 17.765 19.437 rebuild_ks_matrix 110 8.3 0.000 0.001 12.438 13.768 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.015 12.438 13.768 ot_mini 99 10.5 0.001 0.001 12.101 13.435 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.091 12.273 multiply_cannon_multrec 8220 15.4 3.392 5.101 8.281 9.990 init_scf_loop 11 6.9 0.000 0.000 8.687 9.630 qs_ot_get_derivative 99 11.5 0.001 0.001 8.174 9.067 mp_waitall_1 103326 16.6 6.599 8.472 6.599 8.472 prepare_preconditioner 11 7.9 0.000 0.000 6.840 7.585 make_preconditioner 11 8.9 0.000 0.000 6.840 7.585 init_scf_run 11 5.9 0.000 0.004 6.621 7.339 scf_env_initial_rho_setup 11 6.9 0.002 0.004 6.620 7.339 sum_up_and_integrate 110 10.3 0.001 0.002 6.602 7.310 integrate_v_rspace 110 11.3 0.003 0.004 6.574 7.279 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.385 7.155 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.261 6.945 calculate_rho_elec 110 8.6 0.121 0.135 6.260 6.945 qs_ot_get_p 110 10.4 0.018 0.023 6.109 6.779 make_m2s 4110 13.4 0.040 0.043 4.732 5.545 dbcsr_mm_accdrv_process 17442 15.9 3.481 4.418 4.741 5.485 make_images 4110 14.4 0.677 0.824 4.594 5.394 qs_ot_p2m_diag 48 11.0 0.085 0.093 4.412 4.883 multiply_cannon_metrocomm3 8220 15.4 0.020 0.022 3.142 4.745 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.022 4.459 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.006 4.397 apply_single 110 13.6 0.000 0.000 4.006 4.397 ot_diis_step 99 11.5 0.013 0.014 3.897 4.320 fft_wrap_pw1pw2 1111 11.6 0.015 0.017 3.764 4.172 grid_integrate_task_list 110 12.3 3.546 4.107 3.546 4.107 calculate_first_density_matrix 1 7.0 0.000 0.000 3.600 3.988 cp_fm_cholesky_invert 11 10.9 3.357 3.718 3.357 3.718 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.346 3.701 cp_fm_diag_elpa_base 48 14.0 3.275 3.607 3.343 3.697 fft_wrap_pw1pw2_140 451 12.1 0.201 0.221 3.198 3.545 density_rs2pw 110 9.6 0.004 0.004 3.046 3.367 hybrid_alltoall_any 4261 16.3 0.216 0.846 2.507 3.296 wfi_extrapolate 11 7.9 0.001 0.001 2.911 3.226 fft3d_ps 1111 13.6 1.340 1.484 2.881 3.194 make_images_data 4110 15.4 0.043 0.048 2.543 3.182 qs_energies_init_hamiltonians 11 5.9 0.013 0.024 2.830 3.154 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.805 3.108 calculate_dm_sparse 110 9.5 0.001 0.001 2.771 3.066 grid_collocate_task_list 110 9.6 2.490 2.828 2.490 2.828 multiply_cannon_sync_h2d 8220 15.4 2.489 2.768 2.489 2.768 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.382 2.650 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.364 2.601 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.248 2.492 potential_pw2rs 110 12.3 0.015 0.017 2.121 2.347 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.884 2.316 cp_fm_cholesky_decompose 22 10.9 1.951 2.160 1.951 2.160 mp_alltoall_d11v 2046 13.8 1.972 2.118 1.972 2.118 dbcsr_complete_redistribute 325 12.2 0.597 0.661 1.647 1.913 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.703 1.892 qs_env_update_s_mstruct 11 6.9 0.003 0.008 1.589 1.874 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.635 1.803 make_basis_sm 11 9.8 0.000 0.000 1.566 1.735 parallel_gemm_fm 81 9.0 0.000 0.000 1.550 1.728 parallel_gemm_fm_cosma 81 10.0 1.550 1.728 1.550 1.728 mp_allgather_i34 2055 14.4 0.530 1.669 0.530 1.669 multiply_cannon_metrocomm1 8220 15.4 0.024 0.026 1.041 1.667 acc_transpose_blocks 8220 15.4 0.042 0.045 1.471 1.650 qs_create_task_list 11 7.9 0.001 0.002 1.284 1.504 generate_qs_task_list 11 8.9 0.394 0.508 1.283 1.503 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.322 1.478 transfer_rs2pw 451 10.6 0.005 0.006 1.281 1.457 mp_alltoall_z22v 1111 15.6 1.296 1.447 1.296 1.447 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=69.281000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=812.909091, yerr=11.820280 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/47690d86814a0e25b6e15635e68dedc68a09033e_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 1.357193E+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 0.061 0.104 87.425 87.873 qs_mol_dyn_low 1 2.0 0.003 0.003 86.770 87.212 qs_forces 11 3.9 0.008 0.008 86.370 86.812 qs_energies 11 4.9 0.001 0.001 82.278 82.701 scf_env_do_scf 11 5.9 0.001 0.001 71.749 72.116 velocity_verlet 10 3.0 0.025 0.029 55.361 55.690 scf_env_do_scf_inner_loop 99 6.5 0.005 0.008 43.645 43.869 dbcsr_multiply_generic 2055 12.4 0.131 0.136 29.802 30.018 init_scf_loop 11 6.9 0.000 0.000 28.033 28.177 qs_scf_new_mos 99 7.5 0.001 0.001 26.731 26.852 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.730 26.851 prepare_preconditioner 11 7.9 0.000 0.000 26.052 26.182 make_preconditioner 11 8.9 0.000 0.000 26.052 26.182 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.233 25.561 ot_scf_mini 99 9.5 0.002 0.002 24.922 25.026 multiply_cannon 2055 13.4 0.335 0.363 22.812 23.547 multiply_cannon_loop 2055 14.4 0.820 0.843 21.031 21.385 cp_fm_upper_to_full 70 14.2 12.690 18.247 12.690 18.247 ot_mini 99 10.5 0.001 0.001 14.092 14.138 rebuild_ks_matrix 110 8.3 0.000 0.001 13.411 13.512 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 13.410 13.512 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.170 12.262 dbcsr_complete_redistribute 325 12.2 1.018 1.040 7.354 10.575 multiply_cannon_multrec 8220 15.4 4.075 4.228 10.078 10.202 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.363 9.594 qs_ot_get_derivative 99 11.5 0.001 0.001 9.525 9.549 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.803 9.008 mp_alltoall_i22 605 13.7 5.416 8.650 5.416 8.650 mp_waitall_1 84994 16.7 7.427 8.313 7.427 8.313 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.845 6.869 calculate_rho_elec 110 8.6 0.227 0.228 6.845 6.869 sum_up_and_integrate 110 10.3 0.002 0.002 6.575 6.600 integrate_v_rspace 110 11.3 0.004 0.004 6.546 6.570 init_scf_run 11 5.9 0.000 0.001 6.194 6.226 scf_env_initial_rho_setup 11 6.9 0.002 0.007 6.194 6.226 dbcsr_mm_accdrv_process 11614 15.7 3.872 4.178 5.854 6.071 make_m2s 4110 13.4 0.043 0.044 5.296 5.782 make_images 4110 14.4 0.883 0.939 5.106 5.591 qs_ot_get_p 110 10.4 0.007 0.012 5.556 5.580 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.709 5.194 apply_single 110 13.6 0.000 0.000 4.709 5.193 multiply_cannon_metrocomm3 8220 15.4 0.019 0.020 4.838 5.183 cp_fm_cholesky_invert 11 10.9 5.067 5.093 5.067 5.093 ot_diis_step 99 11.5 0.015 0.015 4.552 4.575 fft_wrap_pw1pw2 1111 11.6 0.016 0.016 4.465 4.481 qs_ot_p2m_diag 48 11.0 0.151 0.155 3.992 4.017 fft_wrap_pw1pw2_140 451 12.1 0.364 0.366 3.900 3.912 qs_energies_init_hamiltonians 11 5.9 0.003 0.004 3.732 3.752 grid_integrate_task_list 110 12.3 3.684 3.746 3.684 3.746 hybrid_alltoall_any 4261 16.3 0.268 0.565 2.787 3.515 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.477 3.495 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.026 3.473 make_images_data 4110 15.4 0.045 0.048 2.770 3.400 calculate_dm_sparse 110 9.5 0.001 0.001 3.330 3.369 fft3d_ps 1111 13.6 1.833 1.847 3.322 3.334 density_rs2pw 110 9.6 0.004 0.004 3.306 3.319 wfi_extrapolate 11 7.9 0.001 0.001 3.279 3.296 multiply_cannon_sync_h2d 8220 15.4 3.138 3.162 3.138 3.162 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.126 3.141 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.895 2.910 cp_fm_diag_elpa_base 48 14.0 2.356 2.553 2.893 2.908 calculate_first_density_matrix 1 7.0 0.000 0.000 2.802 2.813 grid_collocate_task_list 110 9.6 2.673 2.689 2.673 2.689 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.624 2.648 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.619 2.632 qs_env_update_s_mstruct 11 6.9 0.003 0.007 2.261 2.311 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.123 2.177 potential_pw2rs 110 12.3 0.021 0.021 2.105 2.111 jit_kernel_multiply 10 15.2 1.786 2.079 1.786 2.079 mp_alltoall_d11v 2046 13.8 1.992 2.062 1.992 2.062 cp_fm_cholesky_decompose 22 10.9 1.974 2.000 1.974 2.000 qs_create_task_list 11 7.9 0.001 0.001 1.885 1.941 generate_qs_task_list 11 8.9 0.733 0.787 1.884 1.940 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.842 1.855 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.817 1.841 acc_transpose_blocks 8220 15.4 0.042 0.042 1.795 1.824 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.715 1.758 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=87.873000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1231.818182, yerr=48.026164 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/47690d86814a0e25b6e15635e68dedc68a09033e_performance_tests/15/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 1.094965E+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 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 628.867072E+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 3175954870160 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 57905. MP_Allreduce 11059 797. MP_Sync 87 MP_Alltoall 2226 1773752. 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 0.042 0.121 210.413 213.810 qs_mol_dyn_low 1 2.0 0.021 0.109 209.813 213.215 qs_forces 11 3.9 0.005 0.005 209.519 212.862 qs_energies 11 4.9 0.012 0.020 203.916 207.166 scf_env_do_scf 11 5.9 0.001 0.001 186.685 189.715 scf_env_do_scf_inner_loop 117 6.6 0.006 0.016 166.125 168.818 dbcsr_multiply_generic 2507 12.6 0.183 0.187 127.060 129.176 qs_scf_new_mos 117 7.6 0.001 0.001 126.722 128.676 qs_scf_loop_do_ot 117 8.6 0.001 0.001 126.721 128.675 velocity_verlet 10 3.0 0.043 0.138 124.929 126.924 ot_scf_mini 117 9.6 0.003 0.003 119.994 121.824 multiply_cannon 2507 13.6 0.242 0.251 102.411 104.743 multiply_cannon_loop 2507 14.6 2.410 2.453 100.082 102.104 ot_mini 117 10.6 0.001 0.001 67.287 68.216 multiply_cannon_multrec 60168 15.6 31.643 33.119 41.794 43.314 qs_ot_get_derivative 117 11.6 0.001 0.001 42.531 43.040 rebuild_ks_matrix 128 8.3 0.001 0.001 33.817 34.507 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 33.816 34.507 mp_waitall_1 267128 16.5 29.382 31.792 29.382 31.792 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.403 31.007 qs_ot_get_p 128 10.4 0.004 0.010 30.372 30.933 multiply_cannon_sync_h2d 60168 15.6 26.313 27.716 26.313 27.716 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.323 24.912 apply_single 128 13.6 0.001 0.001 24.322 24.912 ot_diis_step 117 11.6 0.008 0.008 24.476 24.873 qs_ot_p2m_diag 83 11.4 0.079 0.093 23.698 24.071 cp_dbcsr_syevd 83 12.4 0.005 0.005 20.709 21.045 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 20.598 20.862 init_scf_loop 11 6.9 0.000 0.001 20.480 20.813 cp_fm_diag_elpa 83 13.4 0.000 0.000 17.676 17.966 cp_fm_diag_elpa_base 83 14.4 17.601 17.895 17.672 17.963 multiply_cannon_metrocomm3 60168 15.6 0.121 0.128 16.330 17.653 prepare_preconditioner 11 7.9 0.000 0.000 15.852 16.143 make_preconditioner 11 8.9 0.000 0.000 15.852 16.143 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.079 15.388 make_m2s 5014 13.6 0.106 0.114 13.942 14.228 sum_up_and_integrate 128 10.3 0.002 0.004 13.898 14.121 integrate_v_rspace 128 11.3 0.004 0.004 13.840 14.063 make_images 5014 14.6 0.407 0.427 13.756 14.052 init_scf_run 11 5.9 0.000 0.005 12.966 13.176 scf_env_initial_rho_setup 11 6.9 0.010 0.026 12.966 13.176 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.877 13.131 calculate_rho_elec 128 8.7 0.045 0.066 12.876 13.131 mp_sum_l 7950 12.9 9.456 10.878 9.456 10.878 dbcsr_mm_accdrv_process 124484 16.2 4.777 4.913 9.720 10.217 wfi_extrapolate 11 7.9 0.001 0.001 9.096 9.243 cp_fm_cholesky_invert 11 10.9 9.051 9.196 9.051 9.196 calculate_dm_sparse 128 9.5 0.001 0.001 8.571 8.723 qs_ot_get_orbitals 117 10.6 0.001 0.001 8.285 8.448 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.230 8.435 multiply_cannon_metrocomm1 60168 15.6 0.094 0.098 6.428 8.214 make_images_data 5014 15.6 0.068 0.074 6.829 7.869 grid_integrate_task_list 128 12.3 7.105 7.488 7.105 7.488 hybrid_alltoall_any 5200 16.5 0.297 2.269 5.967 7.417 density_rs2pw 128 9.7 0.006 0.007 6.497 7.326 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.969 7.080 fft_wrap_pw1pw2 1291 11.7 0.018 0.026 6.631 6.939 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.958 6.133 fft_wrap_pw1pw2_140 523 12.2 0.233 0.245 5.641 5.861 fft3d_ps 1291 13.7 2.095 2.600 5.458 5.693 grid_collocate_task_list 128 9.7 4.848 5.357 4.848 5.357 mp_alltoall_d11v 2415 14.1 4.302 5.314 4.302 5.314 cp_fm_cholesky_decompose 22 10.9 4.636 4.704 4.636 4.704 potential_pw2rs 128 12.3 0.009 0.011 4.389 4.490 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=213.810000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=595.636364, yerr=6.004131 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/47690d86814a0e25b6e15635e68dedc68a09033e_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 842.612736E+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 4020 57958. MP_Allreduce 11128 958. MP_Sync 87 MP_Alltoall 1969 5059968. 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 0.043 0.096 191.239 222.980 qs_mol_dyn_low 1 2.0 0.019 0.047 190.506 222.114 qs_forces 11 3.9 0.008 0.019 190.328 221.942 qs_energies 11 4.9 0.007 0.042 183.546 214.047 scf_env_do_scf 11 5.9 0.001 0.001 165.995 193.529 scf_env_do_scf_inner_loop 117 6.6 0.008 0.047 133.063 155.142 velocity_verlet 10 3.0 0.024 0.048 119.224 139.009 dbcsr_multiply_generic 2507 12.6 0.189 0.217 98.125 115.507 qs_scf_new_mos 117 7.6 0.001 0.001 94.562 109.755 qs_scf_loop_do_ot 117 8.6 0.001 0.001 94.561 109.755 ot_scf_mini 117 9.6 0.004 0.005 89.746 104.021 multiply_cannon 2507 13.6 0.508 0.564 78.101 92.923 multiply_cannon_loop 2507 14.6 1.563 1.790 74.720 89.007 ot_mini 117 10.6 0.001 0.001 50.636 58.509 mp_waitall_1 214728 16.6 23.921 41.772 23.921 41.772 rebuild_ks_matrix 128 8.3 0.001 0.001 32.584 38.643 qs_ks_build_kohn_sham_matrix 128 9.3 0.020 0.049 32.583 38.643 init_scf_loop 11 6.9 0.004 0.034 32.843 38.294 multiply_cannon_multrec 30084 15.6 21.138 25.455 32.054 36.728 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.336 34.827 prepare_preconditioner 11 7.9 0.000 0.000 28.397 33.050 make_preconditioner 11 8.9 0.003 0.019 28.397 33.050 qs_ot_get_derivative 117 11.6 0.001 0.002 28.602 32.755 multiply_cannon_metrocomm3 30084 15.6 0.097 0.115 15.644 31.775 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.101 30.703 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.095 26.892 apply_single 128 13.6 0.001 0.001 22.095 26.892 ot_diis_step 117 11.6 0.014 0.016 21.859 25.483 qs_ot_get_p 128 10.4 0.002 0.005 21.053 23.789 multiply_cannon_sync_h2d 30084 15.6 18.271 21.327 18.271 21.327 cp_fm_cholesky_invert 11 10.9 16.605 19.354 16.605 19.354 qs_ot_p2m_diag 83 11.4 0.188 0.216 16.084 18.697 cp_dbcsr_syevd 83 12.4 0.005 0.006 14.875 17.343 make_m2s 5014 13.6 0.087 0.098 14.049 16.347 make_images 5014 14.6 1.163 1.373 13.831 16.097 sum_up_and_integrate 128 10.3 0.002 0.007 13.752 16.049 integrate_v_rspace 128 11.3 0.003 0.004 13.692 15.980 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.064 15.186 calculate_rho_elec 128 8.7 0.093 0.125 13.064 15.185 init_scf_run 11 5.9 0.002 0.012 12.489 14.560 scf_env_initial_rho_setup 11 6.9 0.001 0.002 12.487 14.559 cp_fm_diag_elpa 83 13.4 0.000 0.000 11.626 13.581 cp_fm_diag_elpa_base 83 14.4 11.370 13.266 11.620 13.569 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 11.298 12.709 dbcsr_mm_accdrv_process 62242 16.2 5.634 7.017 10.356 11.282 multiply_cannon_metrocomm4 27577 15.6 0.108 0.124 3.800 10.773 mp_irecv_dv 69486 16.3 3.589 10.362 3.589 10.362 make_images_data 5014 15.6 0.066 0.076 8.273 10.315 hybrid_alltoall_any 5200 16.5 0.352 1.570 7.209 9.778 wfi_extrapolate 11 7.9 0.001 0.001 8.361 9.748 fft_wrap_pw1pw2 1291 11.7 0.019 0.024 7.402 8.592 grid_integrate_task_list 128 12.3 7.218 8.429 7.218 8.429 cp_fm_cholesky_decompose 22 10.9 6.885 8.024 6.885 8.024 calculate_dm_sparse 128 9.5 0.001 0.001 6.564 7.803 fft_wrap_pw1pw2_140 523 12.2 0.253 0.286 6.489 7.513 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.412 7.470 density_rs2pw 128 9.7 0.006 0.006 6.533 7.317 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.356 7.119 fft3d_ps 1291 13.7 2.772 3.172 5.845 6.803 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.416 6.507 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.393 6.278 grid_collocate_task_list 128 9.7 5.026 6.157 5.026 6.157 mp_sum_l 7950 12.9 4.181 5.943 4.181 5.943 mp_alltoall_d11v 2415 14.1 4.522 5.246 4.522 5.246 potential_pw2rs 128 12.3 0.015 0.017 4.191 4.887 mp_allgather_i34 2507 14.6 1.895 4.885 1.895 4.885 calculate_first_density_matrix 1 7.0 0.002 0.010 3.873 4.519 dbcsr_complete_redistribute 395 12.7 0.796 0.954 3.219 4.511 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=222.980000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=802.909091, yerr=2.842999 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/47690d86814a0e25b6e15635e68dedc68a09033e_performance_tests/17/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 2.928533E+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 3984192 0.0% 0.0% 100.0% average stack size 0.0 0.0 1695.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 948.428800E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1042912 MPI messages size (bytes): total size 2.716210E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.604448E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 281856 36943429632 131072 < size <= 4194304 660064 996105256960 4194304 < size <= 16777216 65632 931531265168 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4010 58094. MP_Allreduce 11102 999. MP_Sync 87 MP_Alltoall 1712 9388896. MP_SendRecv 7936 75008. MP_ISendRecv 7936 75008. MP_Wait 21820 MP_ISend 11748 275205. MP_IRecv 11748 275205. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.036 174.368 177.009 qs_mol_dyn_low 1 2.0 0.007 0.035 173.755 176.475 qs_forces 11 3.9 0.004 0.005 173.573 176.202 qs_energies 11 4.9 0.005 0.017 167.035 169.573 scf_env_do_scf 11 5.9 0.001 0.001 151.063 153.349 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 116.570 118.333 velocity_verlet 10 3.0 0.021 0.080 111.087 112.711 dbcsr_multiply_generic 2507 12.6 0.186 0.192 82.586 84.965 qs_scf_new_mos 117 7.6 0.001 0.001 80.038 81.509 qs_scf_loop_do_ot 117 8.6 0.001 0.001 80.037 81.509 ot_scf_mini 117 9.6 0.003 0.004 75.797 77.244 multiply_cannon 2507 13.6 0.509 0.532 62.891 69.121 multiply_cannon_loop 2507 14.6 1.127 1.195 59.974 62.930 ot_mini 117 10.6 0.001 0.001 43.035 44.053 init_scf_loop 11 6.9 0.000 0.000 34.397 34.919 mp_waitall_1 170520 16.6 24.734 34.491 24.734 34.491 rebuild_ks_matrix 128 8.3 0.001 0.001 30.234 31.108 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 30.233 31.107 prepare_preconditioner 11 7.9 0.000 0.000 30.390 30.846 make_preconditioner 11 8.9 0.000 0.000 30.390 30.846 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.077 29.847 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.266 28.066 multiply_cannon_multrec 20056 15.6 13.220 16.148 22.960 25.883 multiply_cannon_metrocomm3 20056 15.6 0.061 0.064 15.323 24.818 qs_ot_get_derivative 117 11.6 0.001 0.002 23.507 24.233 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.626 20.959 apply_single 128 13.6 0.001 0.001 19.626 20.959 ot_diis_step 117 11.6 0.018 0.018 19.416 19.710 qs_ot_get_p 128 10.4 0.003 0.005 18.167 18.874 make_m2s 5014 13.6 0.078 0.085 14.587 15.734 make_images 5014 14.6 1.141 1.256 14.350 15.495 multiply_cannon_sync_h2d 20056 15.6 13.669 15.080 13.669 15.080 cp_fm_cholesky_invert 11 10.9 14.055 14.266 14.055 14.266 qs_ot_p2m_diag 83 11.4 0.265 0.272 13.855 14.061 sum_up_and_integrate 128 10.3 0.002 0.003 13.586 13.806 integrate_v_rspace 128 11.3 0.003 0.004 13.526 13.746 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.937 13.147 calculate_rho_elec 128 8.7 0.132 0.147 12.936 13.146 cp_dbcsr_syevd 83 12.4 0.005 0.006 12.805 12.999 init_scf_run 11 5.9 0.001 0.026 10.939 11.105 scf_env_initial_rho_setup 11 6.9 0.023 0.034 10.938 11.090 make_images_data 5014 15.6 0.062 0.073 8.740 10.321 hybrid_alltoall_any 5200 16.5 0.450 2.068 7.643 9.956 cp_fm_diag_elpa 83 13.4 0.000 0.000 9.669 9.807 cp_fm_diag_elpa_base 83 14.4 9.260 9.538 9.666 9.804 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 9.111 9.522 dbcsr_mm_accdrv_process 41502 16.2 5.899 6.383 9.192 9.443 multiply_cannon_metrocomm4 17549 15.6 0.067 0.079 3.449 9.133 mp_irecv_dv 50230 16.2 3.317 8.861 3.317 8.861 grid_integrate_task_list 128 12.3 7.321 7.653 7.321 7.653 fft_wrap_pw1pw2 1291 11.7 0.019 0.025 7.448 7.595 wfi_extrapolate 11 7.9 0.001 0.001 7.345 7.455 cp_fm_upper_to_full 105 14.8 5.797 7.315 5.797 7.315 cp_fm_cholesky_decompose 22 10.9 7.166 7.283 7.166 7.283 fft_wrap_pw1pw2_140 523 12.2 0.282 0.290 6.580 6.732 density_rs2pw 128 9.7 0.005 0.006 6.287 6.650 dbcsr_complete_redistribute 395 12.7 1.190 1.262 4.541 6.448 calculate_dm_sparse 128 9.5 0.001 0.001 5.891 6.076 fft3d_ps 1291 13.7 2.766 2.996 5.752 5.851 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.675 5.760 grid_collocate_task_list 128 9.7 5.169 5.686 5.169 5.686 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.661 5.319 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.414 5.298 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.736 4.903 mp_sum_l 7950 12.9 3.275 4.830 3.275 4.830 mp_alltoall_d11v 2415 14.1 4.243 4.690 4.243 4.690 mp_allgather_i34 2507 14.6 1.399 4.675 1.399 4.675 transfer_fm_to_dbcsr 11 9.9 0.019 0.023 2.294 4.083 potential_pw2rs 128 12.3 0.020 0.022 3.990 4.051 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.901 3.975 mp_alltoall_i22 716 14.1 1.919 3.819 1.919 3.819 qs_energies_init_hamiltonians 11 5.9 0.004 0.008 3.720 3.777 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.478 3.560 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=177.009000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=896.545455, yerr=8.228152 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/47690d86814a0e25b6e15635e68dedc68a09033e_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 4.320339E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 5927808 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.2 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.157960E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1133160 MPI messages size (bytes): total size 2.008142E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.772161E+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 315952 35695099904 131072 < size <= 4194304 709496 778939400192 4194304 < size <= 16777216 69840 660837692480 16777216 < size 30480 532676608000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4007 58132. MP_Allreduce 11095 1083. MP_Sync 86 MP_Alltoall 1700 12496371. MP_SendRecv 5842 75008. MP_ISendRecv 5842 75008. MP_Wait 22272 MP_ISend 14840 244848. MP_IRecv 14840 244848. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.126 0.330 196.427 198.958 qs_mol_dyn_low 1 2.0 0.010 0.063 195.088 197.597 qs_forces 11 3.9 0.005 0.007 194.568 197.076 qs_energies 11 4.9 0.014 0.087 187.358 189.776 scf_env_do_scf 11 5.9 0.001 0.009 169.743 171.943 velocity_verlet 10 3.0 0.059 0.319 127.818 129.476 scf_env_do_scf_inner_loop 116 6.6 0.009 0.055 121.213 122.772 qs_scf_new_mos 116 7.6 0.001 0.001 85.060 85.908 qs_scf_loop_do_ot 116 8.6 0.001 0.001 85.059 85.908 dbcsr_multiply_generic 2485 12.5 0.192 0.197 81.414 83.268 ot_scf_mini 116 9.6 0.003 0.004 80.318 81.127 multiply_cannon 2485 13.5 0.547 0.577 55.812 58.732 multiply_cannon_loop 2485 14.5 1.826 1.884 51.962 54.492 init_scf_loop 11 6.9 0.009 0.069 48.396 49.026 prepare_preconditioner 11 7.9 0.000 0.000 43.823 44.387 make_preconditioner 11 8.9 0.004 0.031 43.823 44.387 ot_mini 116 10.6 0.001 0.001 43.553 44.008 make_full_inverse_cholesky 11 9.9 0.011 0.022 37.349 42.586 multiply_cannon_multrec 29820 15.5 13.361 18.717 26.290 31.306 rebuild_ks_matrix 127 8.3 0.001 0.001 29.650 30.350 qs_ks_build_kohn_sham_matrix 127 9.3 0.020 0.048 29.650 30.350 mp_waitall_1 146592 16.7 17.753 27.971 17.753 27.971 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.750 27.379 qs_ot_get_derivative 116 11.6 0.001 0.002 24.004 24.383 qs_ot_get_p 127 10.4 0.004 0.011 23.018 23.411 make_m2s 4970 13.5 0.093 0.096 20.993 22.033 make_images 4970 14.5 1.929 2.235 20.682 21.723 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 18.977 19.856 apply_single 127 13.6 0.001 0.001 18.977 19.856 ot_diis_step 116 11.6 0.017 0.018 19.427 19.679 qs_ot_p2m_diag 82 11.4 0.338 0.386 18.432 18.719 cp_fm_upper_to_full 104 14.8 11.728 17.293 11.728 17.293 cp_fm_cholesky_invert 11 10.9 16.978 17.206 16.978 17.206 cp_dbcsr_syevd 82 12.4 0.005 0.005 16.988 17.206 multiply_cannon_metrocomm3 29820 15.5 0.051 0.054 6.449 15.127 sum_up_and_integrate 127 10.3 0.003 0.007 13.904 14.079 integrate_v_rspace 127 11.3 0.003 0.004 13.841 14.020 cp_fm_diag_elpa 82 13.4 0.000 0.000 13.583 13.756 cp_fm_diag_elpa_base 82 14.4 12.584 12.975 13.575 13.751 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.515 13.685 calculate_rho_elec 127 8.7 0.175 0.207 13.515 13.685 dbcsr_complete_redistribute 393 12.7 1.500 1.616 9.433 13.225 dbcsr_mm_accdrv_process 61748 16.2 8.438 9.135 12.498 13.135 make_images_data 4970 15.5 0.065 0.069 11.287 12.658 hybrid_alltoall_any 5155 16.4 0.524 2.186 9.994 11.763 copy_fm_to_dbcsr 208 11.6 0.002 0.007 7.971 11.761 init_scf_run 11 5.9 0.000 0.003 11.467 11.616 scf_env_initial_rho_setup 11 6.9 0.002 0.002 11.467 11.616 multiply_cannon_sync_h2d 29820 15.5 10.344 11.236 10.344 11.236 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 10.105 10.357 transfer_fm_to_dbcsr 11 9.9 0.001 0.009 6.448 10.167 mp_alltoall_i22 712 14.1 5.804 9.688 5.804 9.688 fft_wrap_pw1pw2 1281 11.7 0.019 0.024 8.059 8.238 cp_fm_cholesky_decompose 22 10.9 7.882 7.979 7.882 7.979 grid_integrate_task_list 127 12.3 7.461 7.920 7.461 7.920 wfi_extrapolate 11 7.9 0.001 0.001 7.804 7.905 multiply_cannon_metrocomm4 24850 15.5 0.080 0.094 2.841 7.493 fft_wrap_pw1pw2_140 519 12.2 0.326 0.341 7.053 7.236 mp_irecv_dv 75445 16.2 2.687 7.210 2.687 7.210 density_rs2pw 127 9.7 0.012 0.058 6.499 6.829 calculate_dm_sparse 127 9.5 0.001 0.001 6.472 6.565 fft3d_ps 1281 13.7 2.927 2.996 6.259 6.408 mp_alltoall_d11v 2401 14.1 5.307 6.137 5.307 6.137 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.696 5.773 grid_collocate_task_list 127 9.7 5.256 5.691 5.256 5.691 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.647 4.744 qs_energies_init_hamiltonians 11 5.9 0.002 0.004 4.560 4.625 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.480 4.574 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.307 4.403 potential_pw2rs 127 12.3 0.023 0.030 4.219 4.283 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=198.958000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1089.636364, yerr=15.697555 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/47690d86814a0e25b6e15635e68dedc68a09033e_performance_tests/19/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 5.865089E+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 1960712 0.0% 0.0% 100.0% average stack size 0.0 0.0 3445.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.541517E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 240672 MPI messages size (bytes): total size 1.331455E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.532238E+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 113904 59718500352 4194304 < size <= 16777216 104976 550376570880 16777216 < size 20208 721350232272 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8931 51. MP_Alltoall 9654 799394. MP_ISend 40068 2102573. MP_IRecv 40068 2101676. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4020 57947. MP_Allreduce 11127 1165. MP_Sync 87 MP_Alltoall 1712 18838222. MP_SendRecv 3840 122880. MP_ISendRecv 3840 122880. MP_Wait 16122 MP_ISend 10680 423556. MP_IRecv 10680 423556. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.070 0.093 187.435 215.764 qs_mol_dyn_low 1 2.0 0.003 0.003 186.541 214.733 qs_forces 11 3.9 0.004 0.005 186.426 214.602 qs_energies 11 4.9 0.002 0.003 178.508 205.486 scf_env_do_scf 11 5.9 0.001 0.001 158.972 182.998 velocity_verlet 10 3.0 0.013 0.017 121.325 139.667 scf_env_do_scf_inner_loop 117 6.6 0.004 0.009 119.701 137.793 qs_scf_new_mos 117 7.6 0.001 0.001 80.632 92.755 qs_scf_loop_do_ot 117 8.6 0.001 0.001 80.631 92.754 dbcsr_multiply_generic 2507 12.6 0.197 0.222 78.059 90.001 ot_scf_mini 117 9.6 0.004 0.004 75.882 87.333 multiply_cannon 2507 13.6 0.590 0.685 57.511 66.023 multiply_cannon_loop 2507 14.6 0.855 1.000 54.140 62.506 ot_mini 117 10.6 0.001 0.001 41.232 47.429 init_scf_loop 11 6.9 0.000 0.000 39.102 45.012 prepare_preconditioner 11 7.9 0.000 0.000 34.909 40.196 make_preconditioner 11 8.9 0.000 0.000 34.909 40.196 mp_waitall_1 125778 16.7 27.630 38.791 27.630 38.791 make_full_inverse_cholesky 11 9.9 0.016 0.033 32.584 37.400 rebuild_ks_matrix 128 8.3 0.001 0.001 31.193 35.957 qs_ks_build_kohn_sham_matrix 128 9.3 0.018 0.021 31.192 35.956 qs_ks_update_qs_env 128 7.6 0.001 0.001 28.438 32.776 qs_ot_get_derivative 117 11.6 0.002 0.002 22.745 26.154 multiply_cannon_metrocomm3 10028 15.6 0.027 0.032 14.232 24.671 qs_ot_get_p 128 10.4 0.020 0.052 21.030 24.155 cp_fm_cholesky_invert 11 10.9 20.069 23.094 20.069 23.094 multiply_cannon_multrec 10028 15.6 10.787 14.845 18.882 22.523 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 18.447 21.345 apply_single 128 13.6 0.001 0.001 18.447 21.344 ot_diis_step 117 11.6 0.021 0.024 18.389 21.168 qs_ot_p2m_diag 83 11.4 0.523 0.609 16.710 19.232 make_m2s 5014 13.6 0.068 0.081 16.018 18.869 make_images 5014 14.6 2.268 2.782 15.692 18.499 cp_dbcsr_syevd 83 12.4 0.006 0.006 15.416 17.744 sum_up_and_integrate 128 10.3 0.002 0.003 14.872 17.081 integrate_v_rspace 128 11.3 0.004 0.005 14.809 17.010 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.602 16.824 calculate_rho_elec 128 8.7 0.271 0.319 14.601 16.823 cp_fm_diag_elpa 83 13.4 0.000 0.000 12.186 14.023 init_scf_run 11 5.9 0.000 0.003 12.175 14.015 scf_env_initial_rho_setup 11 6.9 0.002 0.003 12.175 14.015 cp_fm_diag_elpa_base 83 14.4 11.924 13.697 12.179 14.013 multiply_cannon_sync_h2d 10028 15.6 11.275 13.714 11.275 13.714 make_images_data 5014 15.6 0.059 0.069 9.253 11.833 hybrid_alltoall_any 5200 16.5 0.876 3.761 9.023 11.480 qs_ot_get_derivative_diag 77 12.4 0.003 0.003 8.967 10.296 fft_wrap_pw1pw2 1291 11.7 0.019 0.023 8.947 10.293 cp_fm_cholesky_decompose 22 10.9 8.751 10.051 8.751 10.051 grid_integrate_task_list 128 12.3 8.211 9.923 8.211 9.923 dbcsr_mm_accdrv_process 20762 16.1 3.357 4.535 7.732 9.664 wfi_extrapolate 11 7.9 0.001 0.001 7.845 9.031 fft_wrap_pw1pw2_140 523 12.2 0.457 0.531 7.814 8.983 multiply_cannon_metrocomm1 10028 15.6 0.032 0.037 4.858 8.548 density_rs2pw 128 9.7 0.005 0.006 6.847 7.977 fft3d_ps 1291 13.7 3.289 3.818 6.809 7.848 calculate_dm_sparse 128 9.5 0.001 0.001 6.549 7.465 grid_collocate_task_list 128 9.7 5.871 6.955 5.871 6.955 dbcsr_complete_redistribute 395 12.7 2.227 2.543 5.788 6.717 qs_energies_init_hamiltonians 11 5.9 0.020 0.036 5.621 6.474 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.588 6.434 multiply_cannon_metrocomm4 7521 15.6 0.027 0.033 1.929 6.067 mp_irecv_dv 28860 15.9 1.886 5.961 1.886 5.961 mp_alltoall_d11v 2415 14.1 5.262 5.871 5.262 5.871 potential_pw2rs 128 12.3 0.027 0.031 4.390 5.057 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.348 4.989 mp_allgather_i34 2507 14.6 1.470 4.903 1.470 4.903 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.041 4.714 calculate_first_density_matrix 1 7.0 0.000 0.000 4.077 4.690 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.955 4.574 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.805 4.552 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.773 4.433 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=215.764000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1444.636364, yerr=42.268133 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/47690d86814a0e25b6e15635e68dedc68a09033e_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430454546432 0.0% 0.0% 100.0% flops 32 x 32 x 32 1979979923456 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 94.250208E+12 0.0% 0.0% 100.0% flops max/rank 11.788211E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806645728 0.0% 0.0% 100.0% number of processed stacks 1983232 0.0% 0.0% 100.0% average stack size 0.0 0.0 3432.1 marketing flops 145.668111E+12 ------------------------------------------------------------------------------- # multiplications 2537 max memory usage/rank 3.206472E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101480 MPI messages size (bytes): total size 1.145238E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.285357E+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 45968 35701915648 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592540656 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4042 58572. MP_Allreduce 11182 1500. MP_Sync 87 MP_Alltoall 1724 36993632. 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 0.059 0.128 315.361 315.361 qs_mol_dyn_low 1 2.0 0.003 0.003 314.337 314.367 qs_forces 11 3.9 0.006 0.006 314.216 314.220 qs_energies 11 4.9 0.002 0.003 305.179 305.193 scf_env_do_scf 11 5.9 0.001 0.001 281.476 281.484 velocity_verlet 10 3.0 0.019 0.021 224.542 224.552 scf_env_do_scf_inner_loop 118 6.6 0.011 0.015 150.372 150.374 init_scf_loop 11 6.9 0.000 0.000 130.806 130.809 prepare_preconditioner 11 7.9 0.000 0.000 125.850 125.877 make_preconditioner 11 8.9 0.000 0.000 125.850 125.877 make_full_inverse_cholesky 11 9.9 0.037 0.039 102.846 122.957 qs_scf_new_mos 118 7.6 0.001 0.001 104.197 104.387 qs_scf_loop_do_ot 118 8.6 0.001 0.001 104.196 104.386 ot_scf_mini 118 9.6 0.004 0.004 99.107 99.259 dbcsr_multiply_generic 2537 12.6 0.220 0.228 86.796 87.506 cp_fm_upper_to_full 105 14.8 52.976 74.881 52.976 74.881 multiply_cannon 2537 13.6 0.703 0.761 61.672 65.148 multiply_cannon_loop 2537 14.6 1.064 1.077 57.533 58.998 ot_mini 118 10.6 0.001 0.001 48.399 48.633 dbcsr_complete_redistribute 395 12.7 3.959 4.007 29.246 41.278 cp_fm_cholesky_invert 11 10.9 38.747 38.754 38.747 38.754 copy_fm_to_dbcsr 209 11.7 0.001 0.002 25.568 37.584 rebuild_ks_matrix 129 8.3 0.001 0.001 35.907 36.249 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.018 35.906 36.248 qs_ot_get_p 129 10.4 0.010 0.010 35.253 35.522 transfer_fm_to_dbcsr 11 9.9 0.030 0.030 22.962 34.869 qs_ks_update_qs_env 129 7.6 0.001 0.001 33.397 33.705 mp_waitall_1 104900 16.8 31.052 33.628 31.052 33.628 mp_alltoall_i22 718 14.1 20.765 32.946 20.765 32.946 qs_ot_p2m_diag 83 11.4 0.879 0.884 30.277 30.289 cp_dbcsr_syevd 83 12.4 0.006 0.006 28.323 28.326 qs_ot_get_derivative 118 11.6 0.002 0.002 27.213 27.361 cp_fm_diag_elpa 83 13.4 0.000 0.000 24.366 24.367 cp_fm_diag_elpa_base 83 14.4 19.921 21.448 24.361 24.361 multiply_cannon_metrocomm3 10148 15.6 0.025 0.026 20.792 22.098 ot_diis_step 118 11.6 0.023 0.023 21.060 21.061 make_m2s 5074 13.6 0.076 0.078 19.446 20.814 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.884 20.573 apply_single 129 13.6 0.001 0.001 19.884 20.573 make_images 5074 14.6 3.091 3.286 18.957 20.327 multiply_cannon_multrec 10148 15.6 10.502 12.533 18.594 18.662 qs_rho_update_rho_low 129 7.7 0.001 0.001 17.430 17.446 calculate_rho_elec 129 8.7 0.486 0.487 17.429 17.445 sum_up_and_integrate 129 10.3 0.002 0.002 17.268 17.361 integrate_v_rspace 129 11.3 0.004 0.004 17.204 17.299 multiply_cannon_sync_h2d 10148 15.6 14.420 14.438 14.420 14.438 make_images_data 5074 15.6 0.064 0.068 11.125 13.356 hybrid_alltoall_any 5260 16.5 1.332 3.096 11.202 13.291 init_scf_run 11 5.9 0.001 0.007 13.202 13.203 scf_env_initial_rho_setup 11 6.9 0.006 0.007 13.201 13.203 fft_wrap_pw1pw2 1301 11.7 0.021 0.021 11.319 11.367 cp_fm_cholesky_decompose 22 10.9 10.984 11.015 10.984 11.015 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 10.742 10.848 fft_wrap_pw1pw2_140 527 12.2 0.831 0.832 9.869 9.925 dbcsr_mm_accdrv_process 20966 16.1 4.430 6.275 7.833 9.828 wfi_extrapolate 11 7.9 0.001 0.001 9.526 9.526 grid_integrate_task_list 129 12.3 8.623 8.826 8.623 8.826 fft3d_ps 1301 13.7 4.025 4.042 8.599 8.634 qs_energies_init_hamiltonians 11 5.9 0.002 0.003 8.288 8.289 density_rs2pw 129 9.7 0.005 0.005 8.117 8.174 mp_alltoall_d11v 2423 14.1 7.998 8.167 7.998 8.167 calculate_dm_sparse 129 9.5 0.001 0.001 7.068 7.101 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.604 6.619 copy_dbcsr_to_fm 186 11.8 0.004 0.004 6.461 6.568 grid_collocate_task_list 129 9.7 6.470 6.555 6.470 6.555 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=315.361000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2879.909091, yerr=174.222666 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/47690d86814a0e25b6e15635e68dedc68a09033e_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 1.263247E+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 234420. 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 0.039 0.102 90.627 98.735 qs_energies 1 2.0 0.000 0.000 89.595 97.602 ls_scf 1 3.0 0.000 0.000 88.642 96.564 dbcsr_multiply_generic 111 6.7 0.015 0.017 76.860 83.754 multiply_cannon 111 7.7 0.018 0.022 59.184 65.154 multiply_cannon_loop 111 8.7 0.241 0.287 55.632 61.536 ls_scf_main 1 4.0 0.000 0.000 55.445 60.409 density_matrix_trs4 2 5.0 0.002 0.003 49.522 53.947 ls_scf_init_scf 1 4.0 0.000 0.000 29.962 32.638 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.662 31.240 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.413 28.777 mp_waitall_1 11031 10.9 23.754 28.658 23.754 28.658 multiply_cannon_multrec 2664 9.7 8.602 10.159 16.451 19.891 multiply_cannon_sync_h2d 2664 9.7 14.401 18.203 14.401 18.203 make_m2s 222 7.7 0.009 0.012 13.889 15.565 make_images 222 8.7 0.106 0.125 13.866 15.542 multiply_cannon_metrocomm1 2664 9.7 0.010 0.013 10.210 13.708 make_images_data 222 9.7 0.005 0.005 8.117 9.286 dbcsr_mm_accdrv_process 4760 10.4 0.632 0.795 7.444 9.234 multiply_cannon_metrocomm3 2664 9.7 0.010 0.012 5.787 8.827 hybrid_alltoall_any 227 10.6 0.224 1.846 7.076 8.492 dbcsr_mm_accdrv_process_sort 4760 11.4 6.592 8.194 6.592 8.194 calculate_norms 4752 9.8 5.840 7.158 5.840 7.158 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.355 5.941 mp_sum_l 887 5.1 3.272 4.932 3.272 4.932 multiply_cannon_metrocomm4 2442 9.7 0.013 0.016 2.197 4.030 mp_irecv_dv 6231 10.9 2.180 3.994 2.180 3.994 arnoldi_extremal 4 6.8 0.000 0.000 3.429 3.746 arnoldi_normal_ev 4 7.8 0.001 0.003 3.429 3.746 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.410 3.658 build_subspace 16 8.4 0.010 0.014 3.330 3.627 make_images_sizes 222 9.7 0.000 0.000 0.721 3.569 mp_alltoall_i44 222 10.7 0.721 3.569 0.721 3.569 ls_scf_post 1 4.0 0.000 0.000 3.234 3.521 ls_scf_store_result 1 5.0 0.000 0.000 3.038 3.330 dbcsr_special_finalize 555 9.7 0.006 0.007 2.551 3.112 dbcsr_merge_single_wm 555 10.7 0.483 0.683 2.542 3.102 make_images_pack 222 9.7 2.344 3.044 2.345 3.046 dbcsr_matrix_vector_mult 304 9.0 0.007 0.015 2.457 2.867 dbcsr_sort_data 658 11.4 2.333 2.809 2.333 2.809 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.451 2.776 dbcsr_matrix_vector_mult_local 304 10.0 2.184 2.740 2.186 2.742 buffer_matrices_ensure_size 222 8.7 1.848 2.411 1.848 2.411 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.983 2.157 rebuild_ks_matrix 3 7.3 0.000 0.000 1.973 2.146 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.973 2.146 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=98.735000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1136.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/47690d86814a0e25b6e15635e68dedc68a09033e_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 2.168410E+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 0.045 0.087 92.504 99.762 qs_energies 1 2.0 0.000 0.000 91.926 99.121 ls_scf 1 3.0 0.003 0.025 90.495 97.581 dbcsr_multiply_generic 111 6.7 0.015 0.016 75.789 81.947 ls_scf_main 1 4.0 0.001 0.009 55.853 60.239 multiply_cannon 111 7.7 0.028 0.042 53.572 59.941 multiply_cannon_loop 111 8.7 0.135 0.152 50.799 57.689 density_matrix_trs4 2 5.0 0.002 0.003 49.899 53.850 ls_scf_init_scf 1 4.0 0.000 0.002 30.818 33.233 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.479 31.874 mp_waitall_1 9105 10.9 21.694 30.996 21.694 30.996 matrix_sqrt_Newton_Schulz 2 6.5 0.004 0.011 26.951 29.064 multiply_cannon_multrec 1332 9.7 13.368 17.369 22.762 27.814 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.975 21.423 make_m2s 222 7.7 0.007 0.008 15.479 17.151 make_images 222 8.7 1.387 1.689 15.447 17.119 make_images_data 222 9.7 0.004 0.005 9.092 10.535 dbcsr_mm_accdrv_process 4041 10.4 0.337 0.541 8.986 10.440 dbcsr_mm_accdrv_process_sort 4041 11.4 8.492 9.918 8.492 9.918 hybrid_alltoall_any 227 10.6 0.543 2.566 8.535 9.788 mp_sum_l 887 5.1 5.321 9.235 5.321 9.235 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.252 7.697 mp_irecv_dv 3311 11.0 3.231 7.642 3.231 7.642 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.898 7.168 calculate_norms 2376 9.8 6.107 6.787 6.107 6.787 multiply_cannon_sync_h2d 1332 9.7 4.877 6.655 4.877 6.655 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.168 5.719 arnoldi_extremal 4 6.8 0.000 0.000 4.988 5.395 arnoldi_normal_ev 4 7.8 0.008 0.026 4.987 5.395 build_subspace 16 8.4 0.014 0.021 4.701 5.070 ls_scf_post 1 4.0 0.004 0.032 3.821 4.109 ls_scf_store_result 1 5.0 0.000 0.000 3.526 3.847 dbcsr_matrix_vector_mult 304 9.0 0.010 0.023 3.289 3.530 dbcsr_matrix_vector_mult_local 304 10.0 2.806 3.265 2.808 3.266 mp_allgather_i34 111 8.7 0.737 2.949 0.737 2.949 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.677 2.938 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.273 2.794 dbcsr_data_new 4174 10.1 2.142 2.609 2.142 2.609 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.060 2.218 rebuild_ks_matrix 3 7.3 0.000 0.000 2.046 2.203 qs_ks_build_kohn_sham_matrix 3 8.3 0.007 0.056 2.046 2.203 make_images_pack 222 9.7 1.843 2.130 1.846 2.133 dbcsr_sort_data 436 11.2 1.829 2.122 1.829 2.122 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=99.762000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1796.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/47690d86814a0e25b6e15635e68dedc68a09033e_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 2.958234E+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 0.069 0.123 94.392 100.569 qs_energies 1 2.0 0.001 0.009 93.576 99.681 ls_scf 1 3.0 0.000 0.004 92.010 98.024 dbcsr_multiply_generic 111 6.7 0.017 0.025 76.045 80.765 ls_scf_main 1 4.0 0.008 0.074 57.101 60.841 multiply_cannon 111 7.7 0.029 0.052 52.473 58.042 multiply_cannon_loop 111 8.7 0.116 0.132 49.637 55.133 density_matrix_trs4 2 5.0 0.006 0.039 51.109 54.528 mp_waitall_1 7281 11.0 23.764 33.502 23.764 33.502 ls_scf_init_scf 1 4.0 0.000 0.001 31.234 33.282 ls_scf_init_matrix_S 1 5.0 0.001 0.008 29.814 31.654 matrix_sqrt_Newton_Schulz 2 6.5 0.015 0.062 27.389 29.202 multiply_cannon_multrec 888 9.7 12.747 16.216 21.414 26.090 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.981 23.065 make_m2s 222 7.7 0.006 0.007 16.619 18.168 make_images 222 8.7 1.596 1.952 16.580 18.128 make_images_data 222 9.7 0.004 0.005 9.730 10.660 hybrid_alltoall_any 227 10.6 0.650 3.051 9.190 10.488 dbcsr_mm_accdrv_process 3754 10.4 0.342 0.516 8.173 9.354 mp_sum_l 887 5.1 5.250 8.899 5.250 8.899 dbcsr_mm_accdrv_process_sort 3754 11.4 7.728 8.839 7.728 8.839 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.497 7.549 mp_irecv_dv 2335 11.1 2.481 7.484 2.481 7.484 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.709 7.384 multiply_cannon_sync_h2d 888 9.7 6.101 7.075 6.101 7.075 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.045 6.913 arnoldi_extremal 4 6.8 0.000 0.000 5.326 5.709 arnoldi_normal_ev 4 7.8 0.016 0.032 5.326 5.709 build_subspace 16 8.4 0.014 0.020 4.888 5.196 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.768 5.187 calculate_norms 1584 9.8 4.444 4.978 4.444 4.978 mp_allgather_i34 111 8.7 0.939 3.976 0.939 3.976 dbcsr_matrix_vector_mult 304 9.0 0.010 0.024 3.495 3.929 ls_scf_post 1 4.0 0.004 0.028 3.675 3.901 dbcsr_matrix_vector_mult_local 304 10.0 3.073 3.650 3.076 3.652 ls_scf_store_result 1 5.0 0.000 0.000 3.396 3.610 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.883 3.165 dbcsr_data_new 4116 9.9 2.130 2.515 2.130 2.515 make_images_sizes 222 9.7 0.000 0.000 0.994 2.248 mp_alltoall_i44 222 10.7 0.994 2.248 0.994 2.248 dbcsr_sort_data 325 11.1 1.913 2.240 1.913 2.240 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.051 2.180 rebuild_ks_matrix 3 7.3 0.000 0.000 2.032 2.160 qs_ks_build_kohn_sham_matrix 3 8.3 0.005 0.022 2.032 2.160 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=100.569000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2293.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/47690d86814a0e25b6e15635e68dedc68a09033e_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 3.410653E+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 0.073 0.148 97.591 98.204 qs_energies 1 2.0 0.000 0.000 96.773 97.376 ls_scf 1 3.0 0.002 0.017 95.060 95.660 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.682 79.360 ls_scf_main 1 4.0 0.000 0.002 58.740 59.111 multiply_cannon 111 7.7 0.045 0.102 51.778 55.464 density_matrix_trs4 2 5.0 0.002 0.003 52.633 53.034 multiply_cannon_loop 111 8.7 0.151 0.167 46.811 49.790 ls_scf_init_scf 1 4.0 0.000 0.003 33.043 33.256 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.771 31.994 mp_waitall_1 6369 11.0 22.695 29.805 22.695 29.805 matrix_sqrt_Newton_Schulz 2 6.5 0.005 0.021 29.064 29.264 multiply_cannon_multrec 1332 9.7 14.132 17.575 22.125 24.874 make_m2s 222 7.7 0.007 0.008 21.187 22.514 make_images 222 8.7 3.153 3.637 21.137 22.463 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.282 17.072 make_images_data 222 9.7 0.004 0.004 11.775 13.412 hybrid_alltoall_any 227 10.6 0.799 3.803 11.242 12.975 dbcsr_mm_accdrv_process 3641 10.4 0.292 0.514 7.634 9.073 dbcsr_mm_accdrv_process_sort 3641 11.4 7.138 8.577 7.138 8.577 mp_sum_l 887 5.1 4.169 7.166 4.169 7.166 multiply_cannon_sync_h2d 1332 9.7 5.441 6.051 5.441 6.051 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.104 6.043 mp_irecv_dv 3229 10.9 2.078 5.964 2.078 5.964 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.092 5.551 arnoldi_extremal 4 6.8 0.000 0.000 5.251 5.305 arnoldi_normal_ev 4 7.8 0.006 0.024 5.251 5.305 build_subspace 16 8.4 0.014 0.021 4.908 4.946 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.739 4.924 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.473 4.739 calculate_norms 2376 9.8 4.236 4.619 4.236 4.619 mp_allgather_i34 111 8.7 2.066 4.161 2.066 4.161 dbcsr_matrix_vector_mult 304 9.0 0.011 0.022 3.618 3.925 dbcsr_matrix_vector_mult_local 304 10.0 3.233 3.723 3.235 3.725 dbcsr_sort_data 658 11.4 3.136 3.567 3.136 3.567 ls_scf_post 1 4.0 0.003 0.023 3.276 3.295 dbcsr_special_finalize 555 9.7 0.006 0.007 2.887 3.244 dbcsr_merge_single_wm 555 10.7 0.547 0.665 2.878 3.235 ls_scf_store_result 1 5.0 0.000 0.000 3.008 3.090 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.984 3.022 dbcsr_data_release 10477 10.7 1.573 2.454 1.573 2.454 dbcsr_finalize 304 7.8 0.049 0.062 1.812 2.063 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.017 2.037 rebuild_ks_matrix 3 7.3 0.000 0.000 1.994 2.014 qs_ks_build_kohn_sham_matrix 3 8.3 0.005 0.040 1.994 2.014 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=98.204000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2795.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/47690d86814a0e25b6e15635e68dedc68a09033e_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 4.855235E+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 0.125 0.163 101.705 113.371 qs_energies 1 2.0 0.000 0.000 100.524 112.053 ls_scf 1 3.0 0.000 0.000 98.383 109.667 dbcsr_multiply_generic 111 6.7 0.019 0.020 78.524 87.413 ls_scf_main 1 4.0 0.000 0.000 60.793 67.766 multiply_cannon 111 7.7 0.081 0.147 59.231 65.877 multiply_cannon_loop 111 8.7 0.094 0.108 56.407 61.030 density_matrix_trs4 2 5.0 0.002 0.003 53.352 59.436 ls_scf_init_scf 1 4.0 0.000 0.000 33.936 37.828 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.560 36.281 matrix_sqrt_Newton_Schulz 2 6.5 0.002 0.005 30.303 33.791 mp_waitall_1 5436 11.0 26.421 32.822 26.421 32.822 multiply_cannon_multrec 444 9.7 14.654 17.827 22.254 24.276 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.026 17.847 make_m2s 222 7.7 0.005 0.006 14.360 16.061 make_images 222 8.7 2.180 2.764 14.288 15.985 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 7.480 15.890 hybrid_alltoall_any 227 10.6 0.840 3.815 8.691 10.744 make_images_data 222 9.7 0.003 0.004 8.771 10.571 dbcsr_mm_accdrv_process 3003 10.4 0.375 0.511 7.265 9.002 multiply_cannon_sync_h2d 444 9.7 7.245 8.543 7.245 8.543 dbcsr_mm_accdrv_process_sort 3003 11.4 6.861 8.542 6.861 8.542 arnoldi_extremal 4 6.8 0.000 0.000 6.301 7.038 arnoldi_normal_ev 4 7.8 0.007 0.033 6.301 7.038 build_subspace 16 8.4 0.016 0.020 5.836 6.508 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.702 5.247 dbcsr_matrix_vector_mult 304 9.0 0.012 0.022 4.514 4.998 mp_sum_l 887 5.1 2.832 4.771 2.832 4.771 dbcsr_matrix_vector_mult_local 304 10.0 4.015 4.713 4.017 4.715 calculate_norms 792 9.8 3.868 4.251 3.868 4.251 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.755 4.187 multiply_cannon_metrocomm4 333 9.7 0.002 0.002 1.580 4.102 mp_irecv_dv 1241 11.2 1.568 4.079 1.568 4.079 ls_scf_post 1 4.0 0.000 0.000 3.655 4.072 compute_matrix_preconditioner 1 6.0 0.001 0.002 3.435 3.834 mp_allgather_i34 111 8.7 1.274 3.754 1.274 3.754 ls_scf_store_result 1 5.0 0.000 0.000 3.410 3.723 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.972 3.524 acc_transpose_blocks 444 9.7 0.003 0.003 2.804 3.183 make_images_sizes 222 9.7 0.000 0.000 0.900 3.116 mp_alltoall_i44 222 10.7 0.900 3.116 0.900 3.116 acc_transpose_blocks_kernels 444 10.7 0.007 0.008 2.734 3.108 jit_kernel_transpose 1 13.0 2.727 3.101 2.727 3.101 dbcsr_finalize 304 7.8 0.066 0.087 2.355 2.607 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.248 2.505 dbcsr_data_new 4608 9.7 1.926 2.484 1.926 2.484 rebuild_ks_matrix 3 7.3 0.000 0.000 2.212 2.466 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.212 2.466 dbcsr_merge_all 275 8.9 0.515 0.592 2.202 2.444 qs_energies_init_hamiltonians 1 3.0 0.006 0.018 2.124 2.367 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=113.371000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3799.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/47690d86814a0e25b6e15635e68dedc68a09033e_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 8.999973E+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 0.086 0.098 105.723 106.163 qs_energies 1 2.0 0.000 0.000 104.128 104.568 ls_scf 1 3.0 0.000 0.000 101.073 101.500 dbcsr_multiply_generic 111 6.7 0.024 0.026 74.720 75.082 ls_scf_main 1 4.0 0.000 0.000 63.544 63.808 density_matrix_trs4 2 5.0 0.002 0.003 54.791 55.068 multiply_cannon 111 7.7 0.180 0.289 48.305 50.672 multiply_cannon_loop 111 8.7 0.101 0.119 45.334 45.975 ls_scf_init_scf 1 4.0 0.000 0.000 33.738 33.878 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.141 32.265 matrix_sqrt_Newton_Schulz 2 6.5 0.004 0.014 29.449 29.573 mp_waitall_1 4527 11.1 21.561 25.114 21.561 25.114 make_m2s 222 7.7 0.005 0.005 22.734 23.817 make_images 222 8.7 3.577 3.879 22.626 23.709 multiply_cannon_multrec 444 9.7 17.941 18.587 22.641 23.348 hybrid_alltoall_any 227 10.6 1.658 3.635 12.866 15.840 make_images_data 222 9.7 0.003 0.004 13.114 15.196 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 9.832 10.347 multiply_cannon_sync_h2d 444 9.7 8.783 8.855 8.783 8.855 arnoldi_extremal 4 6.8 0.000 0.000 7.492 7.528 arnoldi_normal_ev 4 7.8 0.005 0.016 7.492 7.528 build_subspace 16 8.4 0.026 0.036 6.925 6.950 dbcsr_matrix_vector_mult 304 9.0 0.018 0.035 5.567 5.730 dbcsr_matrix_vector_mult_local 304 10.0 5.138 5.484 5.140 5.486 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.837 5.116 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.905 5.003 dbcsr_mm_accdrv_process 1814 10.4 0.311 0.449 4.518 4.708 dbcsr_mm_accdrv_process_sort 1814 11.4 4.138 4.272 4.138 4.272 ls_scf_post 1 4.0 0.000 0.000 3.791 3.814 make_images_sizes 222 9.7 0.000 0.000 1.461 3.596 mp_alltoall_i44 222 10.7 1.461 3.595 1.461 3.595 ls_scf_store_result 1 5.0 0.000 0.000 3.508 3.538 mp_allgather_i34 111 8.7 1.080 3.459 1.080 3.459 calculate_norms 792 9.8 3.261 3.312 3.261 3.312 dbcsr_finalize 304 7.8 0.082 0.090 3.102 3.204 qs_energies_init_hamiltonians 1 3.0 0.002 0.003 3.025 3.037 dbcsr_merge_all 275 8.9 0.884 0.921 2.868 2.960 dbcsr_complete_redistribute 5 7.6 1.423 1.487 2.746 2.870 dbcsr_sort_data 325 11.1 2.444 2.515 2.444 2.515 matrix_ls_to_qs 2 6.0 0.000 0.000 2.399 2.513 dbcsr_data_new 6591 9.6 1.842 2.444 1.842 2.444 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.325 2.334 dbcsr_new_transposed 4 7.5 0.243 0.254 2.299 2.313 rebuild_ks_matrix 3 7.3 0.000 0.000 2.258 2.267 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.258 2.267 dbcsr_frobenius_norm 74 6.6 2.064 2.147 2.219 2.256 dbcsr_data_release 12724 10.6 2.022 2.243 2.022 2.243 dbcsr_add_d 103 6.2 0.000 0.000 2.151 2.239 dbcsr_add_anytype 103 7.2 0.861 0.897 2.150 2.239 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=106.163000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=7160.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/47690d86814a0e25b6e15635e68dedc68a09033e_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 585.707520E+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 993462630. 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 0.157 0.230 282.671 290.396 qs_mol_dyn_low 1 2.0 0.016 0.136 281.183 288.865 qs_forces 5 3.8 0.005 0.008 280.880 288.568 qs_energies 5 4.8 0.018 0.176 277.237 284.823 scf_env_do_scf 5 5.8 0.000 0.001 260.836 267.965 scf_env_do_scf_inner_loop 105 6.6 0.003 0.032 226.291 232.472 qs_scf_new_mos 105 7.6 0.000 0.001 174.806 179.478 qs_scf_loop_do_ot 105 8.6 0.001 0.001 174.805 179.477 ot_scf_mini 105 9.6 0.003 0.005 164.550 169.125 dbcsr_multiply_generic 1445 12.2 0.205 0.218 137.641 141.448 velocity_verlet 4 3.0 0.016 0.116 118.784 122.036 multiply_cannon 1445 13.2 0.278 0.290 116.362 120.036 multiply_cannon_loop 1445 14.2 2.857 3.036 114.164 117.853 qs_ot_get_p 112 10.4 0.001 0.001 70.905 72.839 ot_mini 105 10.6 0.001 0.001 62.847 64.519 qs_ot_p2m_diag 40 11.0 0.020 0.030 58.868 60.552 cp_dbcsr_syevd 40 12.0 0.002 0.003 55.350 56.863 cp_fm_syevd 40 13.0 0.000 0.000 49.003 50.351 mp_waitall_1 488190 16.1 39.629 47.403 39.629 47.403 multiply_cannon_multrec 69360 15.2 29.796 34.467 40.416 45.217 cp_fm_redistribute_end 40 14.0 21.815 44.650 21.828 44.662 cp_fm_syevd_base 40 14.0 21.639 44.468 21.639 44.468 qs_ot_get_derivative 55 11.6 0.001 0.001 40.908 41.992 multiply_cannon_metrocomm3 69360 15.2 0.207 0.219 26.719 36.924 init_scf_loop 7 6.6 0.002 0.096 34.493 35.436 rebuild_ks_matrix 110 8.4 0.000 0.000 34.249 35.245 qs_ks_build_kohn_sham_matrix 110 9.4 0.014 0.113 34.249 35.245 multiply_cannon_sync_h2d 69360 15.2 29.168 33.574 29.168 33.574 qs_ks_update_qs_env 112 7.6 0.001 0.001 31.345 32.253 prepare_preconditioner 7 7.6 0.000 0.000 28.597 29.365 make_preconditioner 7 8.6 0.001 0.050 28.597 29.365 qs_rho_update_rho_low 110 7.6 0.001 0.001 24.676 25.391 calculate_rho_elec 110 8.6 0.029 0.038 24.675 25.390 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 23.181 23.744 apply_single 62 13.6 0.000 0.000 23.181 23.744 ot_new_cg_direction 55 11.6 0.001 0.001 21.167 21.745 make_full_inverse_cholesky 7 9.6 0.000 0.000 19.933 20.490 density_rs2pw 110 9.6 0.004 0.005 18.445 19.050 qs_ot_get_orbitals 105 10.6 0.001 0.001 15.429 15.895 transfer_rs2pw 445 10.6 0.008 0.010 14.610 15.211 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 14.295 14.709 cp_fm_cholesky_invert 7 10.6 13.992 14.396 13.992 14.396 mp_sum_l 4764 12.2 13.105 14.050 13.105 14.050 fft_wrap_pw1pw2 1425 12.5 0.019 0.026 13.270 13.966 init_scf_run 5 5.8 0.000 0.008 12.941 13.294 scf_env_initial_rho_setup 5 6.8 0.009 0.012 12.940 13.294 sum_up_and_integrate 60 10.3 0.002 0.029 12.575 12.936 integrate_v_rspace 60 11.3 0.002 0.002 12.557 12.920 calculate_dm_sparse 110 9.5 0.000 0.001 12.399 12.798 fft_wrap_pw1pw2_240 915 14.0 0.194 0.219 11.514 12.055 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 11.369 11.661 dbcsr_mm_accdrv_process 154766 15.8 6.291 6.595 10.487 11.620 qs_vxc_create 110 10.4 0.003 0.019 11.086 11.376 check_diag 80 13.5 8.747 9.062 11.034 11.366 fft3d_pb 915 15.0 2.220 2.437 10.523 11.033 transfer_rs2pw_30 110 11.6 1.267 1.344 7.553 9.382 multiply_cannon_metrocomm1 69360 15.2 0.096 0.102 5.062 9.217 mp_sendrecv_dv 168740 12.6 8.650 8.987 8.650 8.987 potential_pw2rs 60 12.3 0.002 0.003 8.476 8.739 make_full_single_inverse 7 9.6 0.002 0.040 8.276 8.502 make_m2s 2890 13.2 0.078 0.086 7.675 8.351 mp_alltoall_z22v 2340 16.7 7.785 8.255 7.785 8.255 make_images 2890 14.2 0.241 0.262 7.566 8.242 xc_rho_set_and_dset_create 110 12.4 0.074 0.101 7.740 8.102 xc_vxc_pw_create 60 11.3 0.036 0.039 7.469 7.661 calculate_first_density_matrix 1 7.0 0.000 0.005 7.438 7.640 acc_transpose_blocks 69360 15.2 0.356 0.382 6.829 7.417 transfer_pw2rs 245 13.2 0.003 0.004 7.077 7.302 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.001 6.944 7.104 xc_pw_derive 510 13.4 0.005 0.006 6.829 7.054 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 6.189 6.393 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=290.396000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=558.000000, yerr=2.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 47690d86814a0e25b6e15635e68dedc68a09033e Summary: empty Status: OK