=== 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: 1aa3ad7d27726d25ce9c9edce445974c62615587 ################# 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-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/1aa3ad7d27726d25ce9c9edce445974c62615587_performance_tests/01 job id: 51339391 --- 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/1aa3ad7d27726d25ce9c9edce445974c62615587_performance_tests/02 job id: 51339392 --- Point --- name: 20 plot: h2o_32_ri_rpa_mp2 regex: Total MP2 Time= label: RI-MP2 (8n/6r/2t) --- Point --- name: 21 plot: h2o_32_ri_rpa_mp2_mem regex: Estimated peak process memory label: RI-MP2 (8n/6r/2t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1aa3ad7d27726d25ce9c9edce445974c62615587_performance_tests/03 job id: 51339393 --- Point --- name: 100 plot: h2o_64_md regex: CP2K label: (8n/12r/1t) --- Point --- name: 101 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/12r/1t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 6 # threads/rank = 2 nrepeat = 1 time[min] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1aa3ad7d27726d25ce9c9edce445974c62615587_performance_tests/04 job id: 51339394 --- Point --- name: 102 plot: h2o_64_md regex: CP2K label: (8n/6r/2t) --- Point --- name: 103 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/6r/2t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 4 # threads/rank = 3 nrepeat = 1 time[min] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1aa3ad7d27726d25ce9c9edce445974c62615587_performance_tests/05 job id: 51339396 --- Point --- name: 104 plot: h2o_64_md regex: CP2K label: (8n/4r/3t) --- Point --- name: 105 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/4r/3t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 3 # threads/rank = 4 nrepeat = 1 time[min] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1aa3ad7d27726d25ce9c9edce445974c62615587_performance_tests/06 job id: 51339397 --- Point --- name: 106 plot: h2o_64_md regex: CP2K label: (8n/3r/4t) --- Point --- name: 107 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/3r/4t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1aa3ad7d27726d25ce9c9edce445974c62615587_performance_tests/07 job id: 51339398 --- Point --- name: 108 plot: h2o_64_md regex: CP2K label: (8n/2r/6t) --- Point --- name: 109 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/2r/6t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 1 # threads/rank = 12 nrepeat = 1 time[min] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1aa3ad7d27726d25ce9c9edce445974c62615587_performance_tests/08 job id: 51339399 --- 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/1aa3ad7d27726d25ce9c9edce445974c62615587_performance_tests/09 job id: 51339400 --- 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/1aa3ad7d27726d25ce9c9edce445974c62615587_performance_tests/10 job id: 51339401 --- 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/1aa3ad7d27726d25ce9c9edce445974c62615587_performance_tests/11 job id: 51339402 --- 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/1aa3ad7d27726d25ce9c9edce445974c62615587_performance_tests/12 job id: 51339403 --- 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/1aa3ad7d27726d25ce9c9edce445974c62615587_performance_tests/13 job id: 51339404 --- 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/1aa3ad7d27726d25ce9c9edce445974c62615587_performance_tests/14 job id: 51339405 --- 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/1aa3ad7d27726d25ce9c9edce445974c62615587_performance_tests/15 job id: 51339406 --- 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/1aa3ad7d27726d25ce9c9edce445974c62615587_performance_tests/16 job id: 51339407 --- 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/1aa3ad7d27726d25ce9c9edce445974c62615587_performance_tests/17 job id: 51339408 --- 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/1aa3ad7d27726d25ce9c9edce445974c62615587_performance_tests/18 job id: 51339410 --- 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/1aa3ad7d27726d25ce9c9edce445974c62615587_performance_tests/19 job id: 51339411 --- 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/1aa3ad7d27726d25ce9c9edce445974c62615587_performance_tests/20 job id: 51339412 --- Point --- name: 410 plot: h2o_256_md regex: CP2K label: (8n/1r/12t) --- Point --- name: 411 plot: h2o_256_md_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1aa3ad7d27726d25ce9c9edce445974c62615587_performance_tests/21 job id: 51339413 --- Point --- name: 500 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/12r/1t) --- Point --- name: 501 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/12r/1t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 6 # threads/rank = 2 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1aa3ad7d27726d25ce9c9edce445974c62615587_performance_tests/22 job id: 51339414 --- Point --- name: 502 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/6r/2t) --- Point --- name: 503 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/6r/2t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 4 # threads/rank = 3 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1aa3ad7d27726d25ce9c9edce445974c62615587_performance_tests/23 job id: 51339415 --- Point --- name: 504 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/4r/3t) --- Point --- name: 505 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/4r/3t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 3 # threads/rank = 4 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1aa3ad7d27726d25ce9c9edce445974c62615587_performance_tests/24 job id: 51339416 --- Point --- name: 506 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/3r/4t) --- Point --- name: 507 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/3r/4t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1aa3ad7d27726d25ce9c9edce445974c62615587_performance_tests/25 job id: 51339417 --- Point --- name: 508 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/2r/6t) --- Point --- name: 509 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/2r/6t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 1 # threads/rank = 12 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1aa3ad7d27726d25ce9c9edce445974c62615587_performance_tests/26 job id: 51339418 --- Point --- name: 510 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/1r/12t) --- Point --- name: 511 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: 512 H2O (4 NVE MD steps on 64 nodes) input file: benchmarks/QS/00512_H2O/H2O-512_md.inp required files: [] output file: result.log # nodes = 64 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1aa3ad7d27726d25ce9c9edce445974c62615587_performance_tests/27 job id: 51339419 --- 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/1aa3ad7d27726d25ce9c9edce445974c62615587_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 0.018 0.033 133.701 133.701 farming_run 1 2.0 133.229 133.240 133.673 133.676 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.480593E+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 0.005 0.024 115.044 115.044 qs_energies 1 2.0 0.001 0.001 114.860 114.862 mp2_main 1 3.0 0.000 0.000 112.795 112.797 mp2_gpw_main 1 4.0 0.027 0.033 111.796 111.798 mp2_ri_gpw_compute_in 1 5.0 0.173 0.174 92.856 93.153 mp2_ri_gpw_compute_in_loop 1 6.0 0.003 0.004 54.884 55.182 mp2_eri_3c_integrate_gpw 272 7.0 0.152 0.166 41.308 46.385 get_2c_integrals 1 6.0 0.008 0.009 37.167 37.798 integrate_v_rspace 273 8.0 0.442 0.455 24.767 29.637 pw_transfer 6555 10.6 0.385 0.404 27.002 27.508 fft_wrap_pw1pw2 5465 11.4 0.045 0.047 25.634 26.168 grid_integrate_task_list 273 9.0 20.633 26.008 20.633 26.008 fft_wrap_pw1pw2_100 2178 12.4 0.028 0.029 23.174 23.677 compute_2c_integrals 1 7.0 0.002 0.002 19.323 19.325 rpa_ri_compute_en 1 5.0 0.019 0.020 18.836 19.143 compute_2c_integrals_loop_lm 1 8.0 0.002 0.003 18.656 19.046 mp2_eri_2c_integrate_gpw 1 9.0 2.391 2.433 18.655 19.045 cp_fm_cholesky_decompose 12 8.2 17.786 18.417 17.786 18.417 cholesky_decomp 1 7.0 0.000 0.000 16.688 17.315 fft3d_s 5443 13.4 16.217 16.613 16.239 16.632 ao_to_mo_and_store_B_mult_1 272 7.0 10.757 15.320 10.757 15.320 calculate_wavefunction 272 8.0 5.453 5.596 12.442 13.124 rpa_num_int 1 6.0 0.000 0.004 10.662 10.662 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.646 10.660 calc_mat_Q 8 8.0 0.000 0.000 9.500 9.591 contract_S_to_Q 8 9.0 0.000 0.000 8.924 9.015 calc_potential_gpw 544 9.5 0.004 0.005 8.244 8.807 parallel_gemm_fm 14 9.1 0.000 0.000 8.506 8.618 parallel_gemm_fm_cosma 14 10.1 8.506 8.618 8.506 8.618 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 8.111 8.425 potential_pw2rs 545 10.0 0.107 0.110 7.552 8.156 create_integ_mat 1 6.0 0.024 0.028 7.783 7.784 collocate_single_gaussian 272 10.0 0.039 0.041 7.348 7.601 array2fm 1 7.0 0.000 0.000 6.662 7.236 pw_scatter_s 2720 13.7 4.356 4.522 4.356 4.522 pw_gather_s 2722 13.2 3.479 3.676 3.479 3.676 array2fm_buffer_send 1 8.0 2.997 3.162 2.997 3.162 pw_poisson_solve 545 10.5 1.115 1.194 2.183 2.466 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.795345, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2799.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1aa3ad7d27726d25ce9c9edce445974c62615587_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 0.025 0.035 521.857 521.859 farming_run 1 2.0 520.993 520.997 521.813 521.816 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827141120 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788822 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.238565E+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 0.011 0.033 319.092 319.093 qs_energies 1 2.0 0.003 0.024 318.848 318.885 mp2_main 1 3.0 0.007 0.057 203.419 203.434 mp2_gpw_main 1 4.0 0.053 0.123 202.524 202.532 mp2_ri_gpw_compute_en 1 5.0 0.045 0.088 128.405 129.918 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.801 1.965 126.776 126.782 scf_env_do_scf 1 3.0 0.000 0.000 115.140 115.140 qs_ks_update_qs_env 5 5.0 0.000 0.000 114.241 114.248 rebuild_ks_matrix 4 6.0 0.000 0.000 114.240 114.247 qs_ks_build_kohn_sham_matrix 4 7.0 0.054 0.061 114.240 114.247 hfx_ks_matrix 4 8.0 0.001 0.001 113.886 113.890 integrate_four_center 4 9.0 0.152 0.466 113.885 113.889 mp2_ri_gpw_compute_en_expansio 172 7.0 0.469 0.501 111.321 111.987 local_gemm 172 8.0 110.852 111.506 110.852 111.506 integrate_four_center_main 4 10.0 0.110 0.659 101.819 105.579 integrate_four_center_bin 264 11.0 101.709 105.573 101.709 105.573 init_scf_loop 1 4.0 0.000 0.000 97.197 97.197 mp2_ri_gpw_compute_in 1 5.0 0.073 0.142 73.955 74.915 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 53.777 54.734 mp2_eri_3c_integrate_gpw 91 7.0 0.145 0.166 41.597 46.400 integrate_v_rspace 95 8.0 0.398 0.567 28.066 32.724 pw_transfer 2240 10.6 0.148 0.168 29.471 29.952 ao_to_mo_and_store_B_mult_1 91 7.0 10.512 29.096 10.512 29.096 fft_wrap_pw1pw2 1868 11.4 0.017 0.020 28.486 28.982 grid_integrate_task_list 95 9.0 23.480 28.340 23.480 28.340 fft_wrap_pw1pw2_100 730 12.4 0.012 0.013 26.241 26.702 get_2c_integrals 1 6.0 0.002 0.017 20.090 20.115 compute_2c_integrals 1 7.0 0.004 0.014 19.064 19.082 compute_2c_integrals_loop_lm 1 8.0 0.002 0.011 18.580 18.915 mp2_eri_2c_integrate_gpw 1 9.0 1.742 1.848 18.578 18.915 fft3d_s 1823 13.4 18.479 18.850 18.493 18.863 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 17.942 17.942 calculate_wavefunction 91 8.0 2.024 2.057 9.628 9.820 mp2_ri_gpw_compute_en_comm 22 7.0 0.500 0.522 7.562 8.996 potential_pw2rs 186 10.0 0.033 0.034 8.410 8.973 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.013 8.130 8.532 calc_potential_gpw 182 9.5 0.002 0.002 7.895 8.127 collocate_single_gaussian 91 10.0 0.017 0.031 7.730 7.975 mp_sync 37 10.5 4.031 7.672 4.031 7.672 mp_sendrecv_dm3 2068 8.0 5.594 7.071 5.594 7.071 integrate_four_center_load 4 10.0 0.000 0.000 6.756 6.760 hfx_load_balance 1 11.0 0.000 0.000 6.756 6.760 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=202.522096, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1496.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1aa3ad7d27726d25ce9c9edce445974c62615587_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 444.661760E+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 592243. 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 0.008 0.027 53.354 53.355 qs_mol_dyn_low 1 2.0 0.004 0.008 53.109 53.116 qs_forces 11 3.9 0.003 0.003 53.046 53.048 qs_energies 11 4.9 0.002 0.003 51.572 51.579 scf_env_do_scf 11 5.9 0.000 0.001 45.515 45.516 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 43.457 43.457 qs_scf_new_mos 108 7.5 0.000 0.000 33.593 33.876 qs_scf_loop_do_ot 108 8.5 0.000 0.001 33.593 33.876 dbcsr_multiply_generic 2286 12.5 0.094 0.097 33.457 33.858 ot_scf_mini 108 9.5 0.002 0.002 31.929 32.120 multiply_cannon 2286 13.5 0.185 0.190 26.213 27.676 multiply_cannon_loop 2286 14.5 1.828 1.929 25.571 27.058 velocity_verlet 10 3.0 0.002 0.005 25.875 25.876 ot_mini 108 10.5 0.001 0.001 19.337 19.599 qs_ot_get_derivative 108 11.5 0.001 0.001 16.396 16.576 mp_waitall_1 245248 16.5 8.241 14.442 8.241 14.442 multiply_cannon_metrocomm3 54864 15.5 0.074 0.079 5.976 12.529 multiply_cannon_multrec 54864 15.5 3.656 5.725 7.742 11.222 qs_ot_get_p 119 10.4 0.001 0.001 7.995 8.311 rebuild_ks_matrix 119 8.3 0.000 0.000 7.732 7.879 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.012 7.731 7.878 mp_sum_l 7287 12.8 5.385 7.179 5.385 7.179 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.804 6.940 multiply_cannon_sync_h2d 54864 15.5 5.202 6.339 5.202 6.339 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.763 6.204 dbcsr_mm_accdrv_process 76910 16.1 1.844 2.891 4.002 5.687 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.272 5.395 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.266 5.314 init_scf_run 11 5.9 0.000 0.001 4.759 4.759 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.759 4.759 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.477 4.477 sum_up_and_integrate 119 10.3 0.001 0.002 4.348 4.354 integrate_v_rspace 119 11.3 0.002 0.003 4.337 4.344 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.291 4.292 cp_fm_redistribute_end 50 14.0 2.189 4.268 2.195 4.271 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.058 4.220 calculate_rho_elec 119 8.7 0.011 0.017 4.058 4.220 cp_fm_diag_elpa_base 50 14.0 2.071 4.171 2.075 4.179 calculate_dm_sparse 119 9.5 0.000 0.000 2.987 3.141 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.912 3.099 apply_single 119 13.6 0.000 0.000 2.911 3.099 multiply_cannon_metrocomm1 54864 15.5 0.056 0.061 1.670 2.808 acc_transpose_blocks 54864 15.5 0.219 0.242 2.237 2.776 jit_kernel_multiply 13 15.8 2.097 2.732 2.097 2.732 ot_diis_step 108 11.5 0.006 0.006 2.665 2.666 calculate_first_density_matrix 1 7.0 0.002 0.009 2.582 2.589 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.415 2.479 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.258 2.260 density_rs2pw 119 9.7 0.004 0.004 2.053 2.155 grid_integrate_task_list 119 12.3 2.010 2.106 2.010 2.106 wfi_extrapolate 11 7.9 0.001 0.003 2.106 2.106 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.999 2.066 init_scf_loop 11 6.9 0.000 0.000 2.043 2.043 mp_sum_d 4137 12.0 1.281 1.887 1.281 1.887 potential_pw2rs 119 12.3 0.004 0.004 1.744 1.753 pw_transfer 1439 11.6 0.051 0.058 1.578 1.643 make_m2s 4572 13.5 0.053 0.056 1.547 1.598 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.502 1.569 make_images 4572 14.5 0.132 0.138 1.466 1.517 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.483 1.501 acc_transpose_blocks_sync 164592 16.5 1.196 1.434 1.196 1.434 mp_alltoall_d11v 2130 13.8 1.287 1.434 1.287 1.434 transfer_rs2pw 487 10.6 0.005 0.006 1.317 1.413 grid_collocate_task_list 119 9.7 1.350 1.410 1.350 1.410 mp_waitany 12084 13.8 1.253 1.388 1.253 1.388 transfer_pw2rs 487 13.2 0.006 0.007 1.284 1.292 fft3d_ps 1201 14.6 0.369 0.473 1.215 1.277 fft_wrap_pw1pw2_140 487 13.2 0.108 0.114 1.167 1.238 dbcsr_dot_sd 1205 11.9 0.049 0.060 0.703 1.098 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=53.355000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=424.090909, yerr=1.239835 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1aa3ad7d27726d25ce9c9edce445974c62615587_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 481.177600E+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 1032117. 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 0.013 0.031 38.624 38.626 qs_mol_dyn_low 1 2.0 0.020 0.072 38.235 38.244 qs_forces 11 3.9 0.011 0.016 38.154 38.172 qs_energies 11 4.9 0.002 0.002 36.482 36.502 scf_env_do_scf 11 5.9 0.000 0.001 31.441 31.442 scf_env_do_scf_inner_loop 108 6.5 0.004 0.010 29.023 29.023 dbcsr_multiply_generic 2286 12.5 0.101 0.104 21.878 22.221 qs_scf_new_mos 108 7.5 0.001 0.001 20.471 20.700 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.471 20.700 ot_scf_mini 108 9.5 0.002 0.003 19.553 19.719 multiply_cannon 2286 13.5 0.207 0.216 16.788 18.239 velocity_verlet 10 3.0 0.003 0.010 18.174 18.176 multiply_cannon_loop 2286 14.5 1.192 1.258 15.658 17.166 ot_mini 108 10.5 0.001 0.001 12.065 12.299 mp_waitall_1 200699 16.5 5.433 10.686 5.433 10.686 qs_ot_get_derivative 108 11.5 0.001 0.001 9.716 9.883 multiply_cannon_metrocomm3 27432 15.5 0.072 0.074 3.975 9.450 multiply_cannon_multrec 27432 15.5 1.827 4.083 6.149 8.847 rebuild_ks_matrix 119 8.3 0.000 0.000 6.748 6.885 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.748 6.884 dbcsr_mm_accdrv_process 47894 16.0 3.577 6.075 4.243 6.333 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.936 6.060 qs_ot_get_p 119 10.4 0.001 0.001 4.595 4.817 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.623 4.541 mp_sum_l 7287 12.8 2.082 4.084 2.082 4.084 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.063 4.053 apply_single 119 13.6 0.000 0.000 3.062 4.053 init_scf_run 11 5.9 0.000 0.001 3.789 3.790 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.789 3.789 sum_up_and_integrate 119 10.3 0.001 0.001 3.673 3.679 integrate_v_rspace 119 11.3 0.002 0.003 3.659 3.665 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.575 3.601 calculate_rho_elec 119 8.7 0.021 0.024 3.575 3.600 qs_ot_p2m_diag 50 11.0 0.009 0.013 2.957 2.977 make_m2s 4572 13.5 0.051 0.053 2.564 2.892 make_images 4572 14.5 0.204 0.242 2.475 2.806 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.559 2.559 init_scf_loop 11 6.9 0.000 0.000 2.398 2.398 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.305 2.392 ot_diis_step 108 11.5 0.011 0.011 2.303 2.303 calculate_first_density_matrix 1 7.0 0.000 0.000 2.281 2.284 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.247 2.247 cp_fm_redistribute_end 50 14.0 1.139 2.223 1.142 2.225 multiply_cannon_sync_h2d 27432 15.5 1.717 2.201 1.717 2.201 cp_fm_diag_elpa_base 50 14.0 1.052 2.139 1.080 2.173 calculate_dm_sparse 119 9.5 0.000 0.001 2.054 2.133 acc_transpose_blocks 27432 15.5 0.112 0.118 1.610 1.953 density_rs2pw 119 9.7 0.004 0.004 1.877 1.948 grid_integrate_task_list 119 12.3 1.839 1.934 1.839 1.934 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.858 1.860 pw_transfer 1439 11.6 0.065 0.068 1.823 1.851 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.731 1.763 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.667 1.706 jit_kernel_multiply 9 16.5 0.610 1.686 0.610 1.686 make_images_data 4572 15.5 0.047 0.054 1.222 1.566 prepare_preconditioner 11 7.9 0.000 0.000 1.480 1.507 make_preconditioner 11 8.9 0.000 0.000 1.480 1.507 wfi_extrapolate 11 7.9 0.001 0.001 1.450 1.450 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.387 1.443 hybrid_alltoall_any 4725 16.4 0.053 0.115 1.086 1.430 potential_pw2rs 119 12.3 0.006 0.006 1.418 1.422 fft_wrap_pw1pw2_140 487 13.2 0.127 0.132 1.383 1.416 fft3d_ps 1201 14.6 0.521 0.576 1.362 1.391 grid_collocate_task_list 119 9.7 1.287 1.354 1.287 1.354 mp_alltoall_d11v 2130 13.8 1.182 1.349 1.182 1.349 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.241 1.250 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.194 1.242 transfer_rs2pw 487 10.6 0.005 0.005 1.081 1.147 mp_sum_d 4137 12.0 0.547 1.053 0.547 1.053 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.994 1.012 acc_transpose_blocks_sync 82296 16.5 0.816 0.944 0.816 0.944 mp_allgather_i34 2286 14.5 0.557 0.943 0.557 0.943 transfer_pw2rs 487 13.2 0.004 0.005 0.891 0.898 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.869 0.882 acc_transpose_blocks_kernels 27432 16.5 0.190 0.278 0.657 0.864 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=38.626000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=457.636364, yerr=1.871933 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1aa3ad7d27726d25ce9c9edce445974c62615587_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 512.188416E+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 0.053 0.075 32.007 32.008 qs_mol_dyn_low 1 2.0 0.003 0.003 31.634 31.644 qs_forces 11 3.9 0.002 0.003 31.371 31.371 qs_energies 11 4.9 0.002 0.002 29.839 29.840 scf_env_do_scf 11 5.9 0.000 0.001 25.269 25.269 scf_env_do_scf_inner_loop 108 6.5 0.005 0.009 22.819 22.819 dbcsr_multiply_generic 2286 12.5 0.096 0.099 16.782 16.842 qs_scf_new_mos 108 7.5 0.001 0.001 15.251 15.268 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.250 15.267 velocity_verlet 10 3.0 0.002 0.002 15.180 15.188 ot_scf_mini 108 9.5 0.002 0.002 14.492 14.509 multiply_cannon 2286 13.5 0.193 0.195 13.582 14.372 multiply_cannon_loop 2286 14.5 0.862 0.929 12.842 13.645 ot_mini 108 10.5 0.001 0.001 8.993 9.010 qs_ot_get_derivative 108 11.5 0.001 0.001 7.516 7.534 multiply_cannon_multrec 18288 15.5 1.872 2.862 6.972 7.330 rebuild_ks_matrix 119 8.3 0.000 0.000 5.943 5.960 qs_ks_build_kohn_sham_matrix 119 9.3 0.018 0.021 5.942 5.959 dbcsr_mm_accdrv_process 38222 16.0 4.916 5.756 5.007 5.819 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.237 5.252 mp_waitall_1 158411 16.6 2.547 3.717 2.547 3.717 sum_up_and_integrate 119 10.3 0.001 0.001 3.490 3.494 integrate_v_rspace 119 11.3 0.003 0.003 3.478 3.483 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.797 3.416 init_scf_run 11 5.9 0.000 0.001 3.413 3.413 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.413 3.413 qs_ot_get_p 119 10.4 0.001 0.001 3.286 3.313 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.208 3.213 calculate_rho_elec 119 8.7 0.031 0.031 3.207 3.213 multiply_cannon_metrocomm3 18288 15.5 0.049 0.050 1.466 2.466 init_scf_loop 11 6.9 0.000 0.000 2.434 2.434 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.018 2.341 apply_single 119 13.6 0.000 0.000 2.018 2.341 calculate_first_density_matrix 1 7.0 0.000 0.000 2.207 2.208 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.181 2.187 make_m2s 4572 13.5 0.044 0.045 1.771 1.900 grid_integrate_task_list 119 12.3 1.807 1.900 1.807 1.900 density_rs2pw 119 9.7 0.004 0.004 1.784 1.891 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.886 1.887 calculate_dm_sparse 119 9.5 0.000 0.001 1.847 1.855 pw_transfer 1439 11.6 0.065 0.069 1.825 1.833 make_images 4572 14.5 0.190 0.202 1.685 1.813 acc_transpose_blocks 18288 15.5 0.077 0.079 1.622 1.762 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.732 1.742 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.693 1.704 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.650 1.660 cp_fm_diag_elpa_base 50 14.0 1.627 1.641 1.648 1.659 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.648 1.650 prepare_preconditioner 11 7.9 0.000 0.000 1.628 1.635 make_preconditioner 11 8.9 0.000 0.000 1.628 1.635 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.490 1.569 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.499 1.506 ot_diis_step 108 11.5 0.011 0.011 1.464 1.464 fft_wrap_pw1pw2_140 487 13.2 0.179 0.183 1.406 1.414 mp_sum_l 7287 12.8 1.049 1.375 1.049 1.375 grid_collocate_task_list 119 9.7 1.232 1.327 1.232 1.327 potential_pw2rs 119 12.3 0.007 0.008 1.301 1.303 fft3d_ps 1201 14.6 0.546 0.564 1.286 1.296 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.184 1.187 multiply_cannon_sync_h2d 18288 15.5 1.034 1.185 1.034 1.185 wfi_extrapolate 11 7.9 0.001 0.001 1.154 1.154 transfer_rs2pw 487 10.6 0.005 0.005 0.940 1.061 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.945 0.963 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.936 0.937 make_images_data 4572 15.5 0.047 0.051 0.770 0.928 acc_transpose_blocks_sync 54864 16.5 0.740 0.874 0.740 0.874 hybrid_alltoall_any 4725 16.4 0.058 0.114 0.671 0.860 mp_alltoall_d11v 2130 13.8 0.637 0.811 0.637 0.811 acc_transpose_blocks_kernels 18288 16.5 0.219 0.225 0.787 0.795 transfer_pw2rs 487 13.2 0.004 0.004 0.784 0.786 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.775 0.776 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.672 0.739 cp_fm_cholesky_invert 11 10.9 0.687 0.690 0.687 0.690 mp_alltoall_z22v 1201 16.6 0.594 0.660 0.594 0.660 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=32.008000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=487.909091, yerr=0.668043 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1aa3ad7d27726d25ce9c9edce445974c62615587_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 548.712448E+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 0.017 0.037 35.882 35.883 qs_mol_dyn_low 1 2.0 0.003 0.003 35.667 35.675 qs_forces 11 3.9 0.002 0.003 35.610 35.611 qs_energies 11 4.9 0.002 0.002 33.910 33.915 scf_env_do_scf 11 5.9 0.000 0.001 28.970 28.972 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 25.651 25.651 dbcsr_multiply_generic 2286 12.5 0.107 0.110 19.730 19.837 velocity_verlet 10 3.0 0.001 0.002 18.249 18.251 qs_scf_new_mos 108 7.5 0.001 0.001 17.647 17.703 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.646 17.703 ot_scf_mini 108 9.5 0.002 0.003 16.619 16.674 multiply_cannon 2286 13.5 0.218 0.227 16.094 16.412 multiply_cannon_loop 2286 14.5 1.534 1.612 15.162 15.505 ot_mini 108 10.5 0.001 0.001 10.259 10.322 multiply_cannon_multrec 27432 15.5 2.452 3.107 9.024 9.325 qs_ot_get_derivative 108 11.5 0.001 0.001 8.394 8.451 dbcsr_mm_accdrv_process 47916 15.9 6.079 7.770 6.469 7.831 rebuild_ks_matrix 119 8.3 0.000 0.000 6.267 6.311 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.267 6.311 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.550 5.589 qs_ot_get_p 119 10.4 0.001 0.001 3.526 3.603 init_scf_run 11 5.9 0.000 0.001 3.596 3.597 scf_env_initial_rho_setup 11 6.9 0.000 0.001 3.596 3.596 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.116 3.521 sum_up_and_integrate 119 10.3 0.001 0.001 3.439 3.448 integrate_v_rspace 119 11.3 0.003 0.003 3.427 3.437 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.332 3.369 calculate_rho_elec 119 8.7 0.040 0.046 3.331 3.369 init_scf_loop 11 6.9 0.000 0.000 3.300 3.301 acc_transpose_blocks 27432 15.5 0.117 0.120 2.398 2.504 prepare_preconditioner 11 7.9 0.000 0.000 2.479 2.487 make_preconditioner 11 8.9 0.000 0.000 2.479 2.487 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.091 2.416 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.324 2.352 make_m2s 4572 13.5 0.054 0.056 2.239 2.345 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.120 2.243 apply_single 119 13.6 0.000 0.000 2.120 2.243 make_images 4572 14.5 0.272 0.334 2.130 2.234 calculate_first_density_matrix 1 7.0 0.000 0.000 2.227 2.229 mp_waitall_1 137007 16.6 1.701 2.213 1.701 2.213 calculate_dm_sparse 119 9.5 0.000 0.000 2.125 2.180 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.137 2.147 pw_transfer 1439 11.6 0.065 0.069 1.960 1.996 grid_integrate_task_list 119 12.3 1.831 1.922 1.831 1.922 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.867 1.906 density_rs2pw 119 9.7 0.004 0.004 1.774 1.864 ot_diis_step 108 11.5 0.012 0.012 1.826 1.826 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.823 1.824 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.736 1.736 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.650 1.663 fft_wrap_pw1pw2_140 487 13.2 0.227 0.235 1.557 1.598 acc_transpose_blocks_sync 82296 16.5 1.433 1.531 1.433 1.531 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.504 1.513 cp_fm_diag_elpa_base 50 14.0 1.471 1.486 1.502 1.512 fft3d_ps 1201 14.6 0.590 0.651 1.356 1.379 grid_collocate_task_list 119 9.7 1.248 1.329 1.248 1.329 wfi_extrapolate 11 7.9 0.001 0.001 1.318 1.318 multiply_cannon_metrocomm3 27432 15.5 0.040 0.042 0.766 1.310 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.281 1.301 mp_sum_l 7287 12.8 0.963 1.268 0.963 1.268 potential_pw2rs 119 12.3 0.008 0.009 1.252 1.254 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.225 1.233 cp_fm_upper_to_full 72 14.2 0.814 1.157 0.814 1.157 qs_energies_init_hamiltonians 11 5.9 0.018 0.021 1.087 1.088 jit_kernel_multiply 6 16.5 0.325 1.082 0.325 1.082 dbcsr_complete_redistribute 329 12.2 0.120 0.145 0.750 1.031 transfer_rs2pw 487 10.6 0.004 0.005 0.865 0.958 make_images_data 4572 15.5 0.048 0.053 0.834 0.949 hybrid_alltoall_any 4725 16.4 0.065 0.155 0.713 0.896 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.801 0.878 mp_alltoall_d11v 2130 13.8 0.724 0.871 0.724 0.871 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.561 0.832 acc_transpose_blocks_kernels 27432 16.5 0.272 0.281 0.819 0.829 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.804 0.808 cp_fm_cholesky_invert 11 10.9 0.722 0.725 0.722 0.725 mp_alltoall_i22 627 13.8 0.423 0.723 0.423 0.723 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=35.883000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=521.636364, yerr=2.496278 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1aa3ad7d27726d25ce9c9edce445974c62615587_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 607.068160E+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 0.021 0.062 28.423 28.424 qs_mol_dyn_low 1 2.0 0.003 0.003 28.172 28.180 qs_forces 11 3.9 0.002 0.002 28.116 28.117 qs_energies 11 4.9 0.002 0.002 26.422 26.424 scf_env_do_scf 11 5.9 0.000 0.001 21.697 21.697 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 19.232 19.232 velocity_verlet 10 3.0 0.002 0.002 14.510 14.512 dbcsr_multiply_generic 2286 12.5 0.096 0.099 12.942 13.015 qs_scf_new_mos 108 7.5 0.001 0.001 11.585 11.609 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.585 11.608 ot_scf_mini 108 9.5 0.002 0.002 10.884 10.911 multiply_cannon 2286 13.5 0.224 0.228 10.425 10.779 multiply_cannon_loop 2286 14.5 0.641 0.656 9.545 9.721 ot_mini 108 10.5 0.001 0.001 6.271 6.303 multiply_cannon_multrec 9144 15.5 1.703 1.978 5.976 6.237 rebuild_ks_matrix 119 8.3 0.000 0.000 5.804 5.827 qs_ks_build_kohn_sham_matrix 119 9.3 0.016 0.018 5.804 5.826 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.171 5.192 qs_ot_get_derivative 108 11.5 0.001 0.001 4.992 5.018 dbcsr_mm_accdrv_process 12550 15.8 3.571 4.191 4.164 4.247 sum_up_and_integrate 119 10.3 0.001 0.001 3.410 3.416 integrate_v_rspace 119 11.3 0.003 0.003 3.400 3.405 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.338 3.346 calculate_rho_elec 119 8.7 0.060 0.061 3.338 3.345 init_scf_run 11 5.9 0.000 0.001 3.249 3.249 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.248 3.248 qs_ot_get_p 119 10.4 0.001 0.001 2.758 2.786 init_scf_loop 11 6.9 0.000 0.000 2.446 2.447 calculate_first_density_matrix 1 7.0 0.000 0.000 2.123 2.124 pw_transfer 1439 11.6 0.066 0.068 2.109 2.118 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.015 2.028 grid_integrate_task_list 119 12.3 1.865 1.923 1.865 1.923 density_rs2pw 119 9.7 0.003 0.003 1.759 1.857 make_m2s 4572 13.5 0.033 0.034 1.717 1.842 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.787 1.790 calculate_dm_sparse 119 9.5 0.000 0.000 1.745 1.763 mp_waitall_1 115863 16.7 1.313 1.762 1.313 1.762 make_images 4572 14.5 0.268 0.299 1.628 1.751 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.680 1.703 fft_wrap_pw1pw2_140 487 13.2 0.325 0.335 1.681 1.693 prepare_preconditioner 11 7.9 0.000 0.000 1.682 1.687 make_preconditioner 11 8.9 0.000 0.000 1.682 1.687 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.571 1.599 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.556 1.557 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.548 1.549 acc_transpose_blocks 9144 15.5 0.042 0.043 1.443 1.478 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.393 1.400 fft3d_ps 1201 14.6 0.651 0.662 1.366 1.376 grid_collocate_task_list 119 9.7 1.297 1.372 1.297 1.372 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.327 1.339 jit_kernel_multiply 8 15.8 0.554 1.337 0.554 1.337 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.275 1.283 cp_fm_diag_elpa_base 50 14.0 1.248 1.265 1.274 1.282 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.265 1.266 ot_diis_step 108 11.5 0.013 0.013 1.266 1.266 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.216 1.248 apply_single 119 13.6 0.000 0.000 1.216 1.248 potential_pw2rs 119 12.3 0.010 0.011 1.224 1.226 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.141 1.146 wfi_extrapolate 11 7.9 0.001 0.001 1.074 1.074 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.875 0.925 hybrid_alltoall_any 4725 16.4 0.065 0.177 0.742 0.915 transfer_rs2pw 487 10.6 0.004 0.005 0.782 0.872 make_images_data 4572 15.5 0.042 0.046 0.729 0.869 cp_fm_cholesky_invert 11 10.9 0.838 0.841 0.838 0.841 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.810 0.816 mp_alltoall_d11v 2130 13.8 0.733 0.802 0.733 0.802 multiply_cannon_metrocomm3 9144 15.5 0.020 0.021 0.372 0.762 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.753 0.754 acc_transpose_blocks_sync 27432 16.5 0.719 0.743 0.719 0.743 qs_env_update_s_mstruct 11 6.9 0.001 0.002 0.685 0.725 acc_transpose_blocks_kernels 9144 16.5 0.118 0.119 0.666 0.687 transfer_pw2rs 487 13.2 0.003 0.004 0.626 0.627 mp_alltoall_z22v 1201 16.6 0.589 0.622 0.589 0.622 mp_allgather_i34 2286 14.5 0.197 0.586 0.197 0.586 jit_kernel_transpose 5 15.6 0.548 0.570 0.548 0.570 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=28.424000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=573.090909, yerr=7.391082 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1aa3ad7d27726d25ce9c9edce445974c62615587_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 753.246208E+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 0.011 0.030 42.459 42.460 qs_mol_dyn_low 1 2.0 0.003 0.003 42.262 42.269 qs_forces 11 3.9 0.003 0.006 42.200 42.200 qs_energies 11 4.9 0.002 0.002 40.194 40.198 scf_env_do_scf 11 5.9 0.001 0.001 34.412 34.413 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 26.602 26.604 velocity_verlet 10 3.0 0.002 0.002 23.971 23.977 dbcsr_multiply_generic 2286 12.5 0.104 0.107 18.718 18.956 qs_scf_new_mos 108 7.5 0.001 0.001 16.969 17.060 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.968 17.059 ot_scf_mini 108 9.5 0.002 0.002 15.813 15.911 multiply_cannon 2286 13.5 0.299 0.304 14.965 15.890 multiply_cannon_loop 2286 14.5 0.867 0.891 13.731 14.686 ot_mini 108 10.5 0.001 0.001 9.627 9.740 multiply_cannon_multrec 9144 15.5 3.338 4.599 8.789 8.881 init_scf_loop 11 6.9 0.000 0.000 7.785 7.788 qs_ot_get_derivative 108 11.5 0.001 0.001 7.596 7.694 rebuild_ks_matrix 119 8.3 0.000 0.000 7.245 7.383 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.245 7.383 prepare_preconditioner 11 7.9 0.000 0.000 6.827 6.839 make_preconditioner 11 8.9 0.000 0.000 6.826 6.839 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.316 6.719 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.522 6.648 dbcsr_mm_accdrv_process 12550 15.8 4.739 6.512 5.318 6.566 cp_fm_upper_to_full 72 14.2 3.210 4.666 3.210 4.666 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.163 4.169 calculate_rho_elec 119 8.7 0.117 0.120 4.163 4.169 sum_up_and_integrate 119 10.3 0.001 0.001 3.847 3.853 integrate_v_rspace 119 11.3 0.004 0.004 3.837 3.843 init_scf_run 11 5.9 0.000 0.001 3.735 3.735 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.735 3.735 qs_ot_get_p 119 10.4 0.001 0.001 3.282 3.419 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.780 3.176 mp_waitall_1 94719 16.7 2.119 3.138 2.119 3.138 pw_transfer 1439 11.6 0.069 0.070 2.973 2.977 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.873 2.877 dbcsr_complete_redistribute 329 12.2 0.283 0.288 2.016 2.848 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.729 2.553 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.173 2.489 apply_single 119 13.6 0.000 0.000 2.173 2.488 fft_wrap_pw1pw2_140 487 13.2 0.651 0.654 2.461 2.467 make_m2s 4572 13.5 0.037 0.037 2.241 2.384 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.505 2.324 mp_alltoall_i22 627 13.8 1.478 2.323 1.478 2.323 calculate_first_density_matrix 1 7.0 0.000 0.000 2.285 2.289 calculate_dm_sparse 119 9.5 0.000 0.000 2.268 2.284 make_images 4572 14.5 0.353 0.387 2.122 2.264 density_rs2pw 119 9.7 0.003 0.004 2.197 2.220 multiply_cannon_metrocomm3 9144 15.5 0.021 0.021 1.269 2.193 grid_integrate_task_list 119 12.3 2.081 2.147 2.081 2.147 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.038 2.088 ot_diis_step 108 11.5 0.014 0.014 2.005 2.005 qs_ot_p2m_diag 50 11.0 0.043 0.043 1.931 1.933 acc_transpose_blocks 9144 15.5 0.044 0.045 1.831 1.879 fft3d_ps 1201 14.6 0.866 0.885 1.793 1.797 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.777 1.777 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.769 1.770 mp_sum_l 7287 12.8 0.962 1.750 0.962 1.750 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.578 1.578 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.534 1.578 grid_collocate_task_list 119 9.7 1.522 1.562 1.522 1.562 wfi_extrapolate 11 7.9 0.001 0.001 1.385 1.385 cp_fm_cholesky_invert 11 10.9 1.379 1.382 1.379 1.382 potential_pw2rs 119 12.3 0.014 0.014 1.335 1.336 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.301 1.321 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.303 1.303 cp_fm_diag_elpa_base 50 14.0 1.158 1.211 1.301 1.301 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.275 1.283 hybrid_alltoall_any 4725 16.4 0.091 0.151 1.016 1.211 mp_alltoall_d11v 2130 13.8 1.149 1.175 1.149 1.175 acc_transpose_blocks_sync 27432 16.5 1.111 1.152 1.111 1.152 qs_env_update_s_mstruct 11 6.9 0.002 0.002 1.098 1.114 make_images_data 4572 15.5 0.046 0.050 0.932 1.103 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.995 1.025 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.931 0.944 qs_create_task_list 11 7.9 0.001 0.001 0.927 0.938 generate_qs_task_list 11 8.9 0.369 0.388 0.926 0.938 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=42.460000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=704.181818, yerr=13.749831 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1aa3ad7d27726d25ce9c9edce445974c62615587_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 493.506560E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66212. MP_Allreduce 9776 488. MP_Sync 52 MP_Alltoall 1938 1383689. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.032 80.114 80.116 qs_mol_dyn_low 1 2.0 0.005 0.017 79.812 79.822 qs_forces 11 3.9 0.003 0.003 79.730 79.732 qs_energies 11 4.9 0.003 0.010 76.891 76.903 scf_env_do_scf 11 5.9 0.000 0.001 68.091 68.094 scf_env_do_scf_inner_loop 99 6.5 0.003 0.010 62.510 62.511 dbcsr_multiply_generic 2055 12.4 0.104 0.106 49.687 50.163 qs_scf_new_mos 99 7.5 0.000 0.001 45.662 45.795 qs_scf_loop_do_ot 99 8.5 0.001 0.001 45.661 45.794 ot_scf_mini 99 9.5 0.002 0.003 43.348 43.480 multiply_cannon 2055 13.4 0.181 0.185 42.033 43.211 multiply_cannon_loop 2055 14.4 1.773 1.816 41.074 42.317 velocity_verlet 10 3.0 0.002 0.007 41.938 41.939 ot_mini 99 10.5 0.002 0.009 25.575 25.701 qs_ot_get_derivative 99 11.5 0.001 0.001 18.772 18.898 multiply_cannon_multrec 49320 15.4 11.583 12.220 17.476 18.232 rebuild_ks_matrix 110 8.3 0.001 0.001 14.432 14.627 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.019 14.432 14.627 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.653 12.831 mp_waitall_1 220248 16.4 9.969 11.075 9.969 11.075 multiply_cannon_sync_h2d 49320 15.4 9.840 10.388 9.840 10.388 qs_ot_get_p 110 10.4 0.001 0.002 9.520 9.628 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.229 7.703 apply_single 110 13.6 0.000 0.000 7.228 7.702 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.088 7.642 multiply_cannon_metrocomm3 49320 15.4 0.083 0.087 6.072 7.544 sum_up_and_integrate 110 10.3 0.002 0.003 6.951 6.966 integrate_v_rspace 110 11.3 0.003 0.003 6.925 6.948 init_scf_run 11 5.9 0.000 0.001 6.710 6.711 scf_env_initial_rho_setup 11 6.9 0.001 0.003 6.710 6.710 ot_diis_step 99 11.5 0.005 0.006 6.625 6.625 qs_ot_p2m_diag 48 11.0 0.013 0.019 6.345 6.371 qs_rho_update_rho_low 110 7.6 0.001 0.002 6.173 6.288 calculate_rho_elec 110 8.6 0.021 0.025 6.173 6.287 dbcsr_mm_accdrv_process 87628 16.1 2.953 3.083 5.761 6.083 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.563 5.563 init_scf_loop 11 6.9 0.007 0.058 5.554 5.559 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.163 5.231 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.000 5.022 cp_fm_diag_elpa_base 48 14.0 4.988 5.010 4.998 5.021 mp_sum_l 6594 12.7 3.384 4.163 3.384 4.163 wfi_extrapolate 11 7.9 0.002 0.008 4.039 4.039 make_m2s 4110 13.4 0.060 0.064 3.825 3.958 make_images 4110 14.4 0.177 0.190 3.731 3.867 calculate_dm_sparse 110 9.5 0.001 0.001 3.675 3.778 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.519 3.523 multiply_cannon_metrocomm1 49320 15.4 0.065 0.068 2.105 3.464 grid_integrate_task_list 110 12.3 3.263 3.426 3.263 3.426 density_rs2pw 110 9.6 0.004 0.004 3.264 3.407 prepare_preconditioner 11 7.9 0.000 0.000 3.248 3.266 make_preconditioner 11 8.9 0.000 0.001 3.248 3.266 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.186 3.237 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.046 3.107 pw_transfer 1331 11.6 0.055 0.068 3.017 3.099 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.033 3.080 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 2.929 3.010 fft_wrap_pw1pw2_140 451 13.1 0.309 0.332 2.517 2.608 calculate_first_density_matrix 1 7.0 0.001 0.004 2.571 2.576 acc_transpose_blocks 49320 15.4 0.199 0.207 2.485 2.571 potential_pw2rs 110 12.3 0.005 0.006 2.549 2.565 jit_kernel_multiply 13 15.9 2.534 2.556 2.534 2.556 mp_alltoall_d11v 2046 13.8 1.997 2.429 1.997 2.429 fft3d_ps 1111 14.6 0.797 0.887 2.223 2.284 grid_collocate_task_list 110 9.6 2.155 2.263 2.155 2.263 mp_waitany 14300 13.8 1.825 2.094 1.825 2.094 transfer_rs2pw 451 10.6 0.005 0.006 1.851 1.989 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.911 1.932 make_images_data 4110 15.4 0.043 0.048 1.734 1.857 mp_sum_d 3891 11.9 1.332 1.855 1.332 1.855 hybrid_alltoall_any 4261 16.3 0.084 0.484 1.507 1.809 cp_fm_cholesky_invert 11 10.9 1.751 1.755 1.751 1.755 transfer_pw2rs 451 13.1 0.006 0.007 1.726 1.732 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.637 1.664 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=80.116000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=468.090909, yerr=2.502891 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1aa3ad7d27726d25ce9c9edce445974c62615587_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 583.684096E+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 1714660. 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 0.008 0.026 68.248 68.249 qs_mol_dyn_low 1 2.0 0.004 0.008 68.039 68.050 qs_forces 11 3.9 0.003 0.003 67.947 67.949 qs_energies 11 4.9 0.002 0.002 64.635 64.640 scf_env_do_scf 11 5.9 0.000 0.001 56.110 56.114 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 48.701 48.702 dbcsr_multiply_generic 2055 12.4 0.115 0.119 37.780 38.013 velocity_verlet 10 3.0 0.002 0.006 35.733 35.735 qs_scf_new_mos 99 7.5 0.001 0.001 32.939 33.062 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.939 33.061 multiply_cannon 2055 13.4 0.224 0.242 31.302 32.399 ot_scf_mini 99 9.5 0.003 0.003 31.272 31.400 multiply_cannon_loop 2055 14.4 1.164 1.183 30.095 30.805 ot_mini 99 10.5 0.001 0.001 18.259 18.401 multiply_cannon_multrec 24660 15.4 6.982 8.511 14.072 15.527 rebuild_ks_matrix 110 8.3 0.000 0.000 13.469 13.568 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 13.469 13.568 qs_ot_get_derivative 99 11.5 0.001 0.001 12.501 12.632 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.846 11.943 mp_waitall_1 176588 16.5 7.431 9.827 7.431 9.827 multiply_cannon_sync_h2d 24660 15.4 6.438 8.015 6.438 8.015 multiply_cannon_metrocomm3 24660 15.4 0.074 0.075 5.096 7.934 init_scf_loop 11 6.9 0.000 0.000 7.375 7.376 dbcsr_mm_accdrv_process 52282 16.1 5.572 6.465 6.920 7.215 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.437 7.065 apply_single 110 13.6 0.000 0.001 6.437 7.065 qs_ot_get_p 110 10.4 0.001 0.001 6.372 6.537 sum_up_and_integrate 110 10.3 0.001 0.003 6.322 6.332 integrate_v_rspace 110 11.3 0.003 0.003 6.295 6.306 init_scf_run 11 5.9 0.000 0.001 6.137 6.138 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.137 6.137 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.839 5.851 calculate_rho_elec 110 8.6 0.039 0.047 5.839 5.851 ot_diis_step 99 11.5 0.010 0.010 5.712 5.712 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.697 5.413 prepare_preconditioner 11 7.9 0.000 0.000 5.379 5.399 make_preconditioner 11 8.9 0.000 0.000 5.379 5.399 make_full_inverse_cholesky 11 9.9 0.000 0.000 4.981 5.134 make_m2s 4110 13.4 0.056 0.059 4.133 4.543 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.503 4.522 make_images 4110 14.4 0.404 0.460 4.022 4.431 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.040 4.041 pw_transfer 1331 11.6 0.066 0.072 3.508 3.641 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.401 3.537 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.487 3.500 cp_fm_diag_elpa_base 48 14.0 3.441 3.452 3.484 3.497 wfi_extrapolate 11 7.9 0.001 0.001 3.477 3.477 density_rs2pw 110 9.6 0.004 0.004 3.093 3.383 grid_integrate_task_list 110 12.3 3.168 3.360 3.168 3.360 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.257 3.328 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.213 3.215 fft_wrap_pw1pw2_140 451 13.1 0.359 0.379 2.924 3.060 calculate_dm_sparse 110 9.5 0.001 0.001 3.009 3.033 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.882 2.927 make_images_data 4110 15.4 0.048 0.053 2.251 2.691 hybrid_alltoall_any 4261 16.3 0.105 0.454 1.934 2.683 fft3d_ps 1111 14.6 1.115 1.349 2.492 2.648 calculate_first_density_matrix 1 7.0 0.000 0.000 2.564 2.567 cp_fm_cholesky_invert 11 10.9 2.489 2.495 2.489 2.495 mp_sum_l 6594 12.7 1.710 2.450 1.710 2.450 grid_collocate_task_list 110 9.6 2.168 2.293 2.168 2.293 potential_pw2rs 110 12.3 0.008 0.009 2.209 2.221 mp_alltoall_d11v 2046 13.8 1.672 2.034 1.672 2.034 acc_transpose_blocks 24660 15.4 0.113 0.118 1.966 1.989 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.951 1.968 jit_kernel_multiply 10 16.4 0.984 1.931 0.984 1.931 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 1.888 1.889 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.777 1.790 transfer_rs2pw 451 10.6 0.006 0.007 1.442 1.724 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.575 1.587 multiply_cannon_metrocomm4 22605 15.4 0.074 0.078 0.769 1.549 mp_allgather_i34 2055 14.4 0.517 1.520 0.517 1.520 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.389 1.493 mp_waitany 10164 13.8 1.214 1.411 1.214 1.411 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=68.249000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=552.545455, yerr=6.959149 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1aa3ad7d27726d25ce9c9edce445974c62615587_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 657.338368E+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 0.008 0.028 59.190 59.191 qs_mol_dyn_low 1 2.0 0.003 0.003 58.945 58.955 qs_forces 11 3.9 0.003 0.003 58.880 58.880 qs_energies 11 4.9 0.001 0.002 55.692 55.695 scf_env_do_scf 11 5.9 0.000 0.001 47.772 47.773 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 39.418 39.419 velocity_verlet 10 3.0 0.001 0.002 32.134 32.135 dbcsr_multiply_generic 2055 12.4 0.110 0.114 28.512 28.756 qs_scf_new_mos 99 7.5 0.001 0.001 24.893 24.984 qs_scf_loop_do_ot 99 8.5 0.001 0.001 24.892 24.983 ot_scf_mini 99 9.5 0.002 0.002 23.648 23.766 multiply_cannon 2055 13.4 0.212 0.222 22.416 23.647 multiply_cannon_loop 2055 14.4 0.812 0.843 21.280 22.427 ot_mini 99 10.5 0.001 0.001 13.562 13.680 rebuild_ks_matrix 110 8.3 0.000 0.000 12.069 12.207 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.068 12.206 multiply_cannon_multrec 16440 15.4 3.738 4.653 9.935 10.809 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.622 10.746 mp_waitall_1 139946 16.5 6.728 9.831 6.728 9.831 qs_ot_get_derivative 99 11.5 0.001 0.001 9.175 9.289 init_scf_loop 11 6.9 0.000 0.000 8.322 8.322 multiply_cannon_metrocomm3 16440 15.4 0.046 0.048 4.232 7.204 prepare_preconditioner 11 7.9 0.000 0.000 6.590 6.611 make_preconditioner 11 8.9 0.000 0.000 6.590 6.611 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.944 6.288 sum_up_and_integrate 110 10.3 0.001 0.002 6.218 6.235 integrate_v_rspace 110 11.3 0.003 0.003 6.192 6.210 dbcsr_mm_accdrv_process 34862 16.1 5.331 5.710 6.042 6.151 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.679 5.688 calculate_rho_elec 110 8.6 0.059 0.059 5.678 5.688 init_scf_run 11 5.9 0.000 0.001 5.546 5.546 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.545 5.546 qs_ot_get_p 110 10.4 0.001 0.001 5.276 5.414 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.843 5.327 apply_single 110 13.6 0.000 0.000 4.843 5.327 make_m2s 4110 13.4 0.049 0.051 4.110 4.474 make_images 4110 14.4 0.399 0.524 3.993 4.359 ot_diis_step 99 11.5 0.011 0.011 4.359 4.359 multiply_cannon_sync_h2d 16440 15.4 3.272 4.029 3.272 4.029 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.188 3.762 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.679 3.683 pw_transfer 1331 11.6 0.066 0.073 3.563 3.574 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.455 3.470 grid_integrate_task_list 110 12.3 3.201 3.397 3.201 3.397 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.313 3.313 density_rs2pw 110 9.6 0.004 0.004 2.936 3.133 fft_wrap_pw1pw2_140 451 13.1 0.461 0.477 3.007 3.025 wfi_extrapolate 11 7.9 0.001 0.001 2.924 2.924 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.790 2.798 cp_fm_diag_elpa_base 48 14.0 2.725 2.754 2.789 2.796 hybrid_alltoall_any 4261 16.3 0.108 0.380 2.088 2.789 make_images_data 4110 15.4 0.046 0.051 2.320 2.758 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.738 2.740 cp_fm_cholesky_invert 11 10.9 2.574 2.580 2.574 2.580 calculate_dm_sparse 110 9.5 0.001 0.001 2.512 2.540 calculate_first_density_matrix 1 7.0 0.000 0.000 2.531 2.533 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.410 2.470 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.435 2.469 fft3d_ps 1111 14.6 1.122 1.140 2.370 2.381 multiply_cannon_metrocomm4 14385 15.4 0.047 0.050 0.864 2.378 grid_collocate_task_list 110 9.6 2.212 2.369 2.212 2.369 mp_irecv_dv 48980 15.7 0.789 2.240 0.789 2.240 potential_pw2rs 110 12.3 0.011 0.011 2.077 2.085 mp_sum_l 6594 12.7 1.359 2.050 1.359 2.050 mp_alltoall_d11v 2046 13.8 1.731 1.987 1.731 1.987 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.948 1.948 dbcsr_complete_redistribute 325 12.2 0.323 0.355 1.375 1.824 cp_fm_upper_to_full 70 14.2 1.367 1.728 1.367 1.728 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.637 1.654 acc_transpose_blocks 16440 15.4 0.077 0.079 1.571 1.648 cp_fm_cholesky_decompose 22 10.9 1.535 1.552 1.535 1.552 mp_allgather_i34 2055 14.4 0.452 1.506 0.452 1.506 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.377 1.506 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.451 1.462 transfer_rs2pw 451 10.6 0.005 0.006 1.236 1.426 mp_waitany 17072 13.8 1.124 1.372 1.124 1.372 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.927 1.367 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.265 1.273 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=59.191000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=624.090909, yerr=7.500964 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1aa3ad7d27726d25ce9c9edce445974c62615587_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 723.566592E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66419. MP_Allreduce 9774 603. MP_Sync 52 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.009 0.027 65.160 65.161 qs_mol_dyn_low 1 2.0 0.003 0.003 64.947 64.958 qs_forces 11 3.9 0.003 0.003 64.748 64.749 qs_energies 11 4.9 0.002 0.002 61.358 61.361 scf_env_do_scf 11 5.9 0.000 0.001 53.206 53.209 scf_env_do_scf_inner_loop 99 6.5 0.014 0.023 41.661 41.661 velocity_verlet 10 3.0 0.003 0.003 37.043 37.050 dbcsr_multiply_generic 2055 12.4 0.117 0.123 30.398 30.633 qs_scf_new_mos 99 7.5 0.001 0.001 27.155 27.252 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.155 27.251 ot_scf_mini 99 9.5 0.002 0.003 25.475 25.565 multiply_cannon 2055 13.4 0.236 0.253 23.483 24.373 multiply_cannon_loop 2055 14.4 1.406 1.469 22.126 22.810 ot_mini 99 10.5 0.001 0.001 14.703 14.798 multiply_cannon_multrec 24660 15.4 4.100 6.749 13.065 14.131 rebuild_ks_matrix 110 8.3 0.000 0.000 11.870 11.982 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.869 11.981 init_scf_loop 11 6.9 0.000 0.000 11.507 11.508 qs_ot_get_derivative 99 11.5 0.001 0.001 10.544 10.629 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.480 10.582 prepare_preconditioner 11 7.9 0.000 0.000 9.768 9.783 make_preconditioner 11 8.9 0.000 0.000 9.768 9.783 dbcsr_mm_accdrv_process 52304 16.0 7.845 9.427 8.807 9.762 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.015 9.477 sum_up_and_integrate 110 10.3 0.001 0.002 6.172 6.186 integrate_v_rspace 110 11.3 0.003 0.003 6.146 6.159 qs_ot_get_p 110 10.4 0.001 0.001 5.724 5.873 mp_waitall_1 121746 16.5 4.065 5.818 4.065 5.818 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.793 5.804 calculate_rho_elec 110 8.6 0.078 0.081 5.793 5.804 make_m2s 4110 13.4 0.059 0.061 5.310 5.642 make_images 4110 14.4 0.574 0.694 5.168 5.497 init_scf_run 11 5.9 0.000 0.001 5.432 5.432 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.432 5.432 cp_fm_upper_to_full 70 14.2 3.379 4.882 3.379 4.882 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.037 4.129 apply_single 110 13.6 0.000 0.000 4.037 4.129 ot_diis_step 99 11.5 0.011 0.011 4.122 4.122 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.866 3.884 pw_transfer 1331 11.6 0.066 0.073 3.727 3.763 dbcsr_complete_redistribute 325 12.2 0.416 0.460 2.603 3.751 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.620 3.659 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.499 3.545 grid_integrate_task_list 110 12.3 3.265 3.430 3.265 3.430 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.337 3.384 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.300 3.301 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.089 3.225 fft_wrap_pw1pw2_140 451 13.1 0.531 0.548 3.130 3.173 density_rs2pw 110 9.6 0.004 0.004 2.932 3.114 calculate_dm_sparse 110 9.5 0.001 0.001 2.989 3.026 multiply_cannon_metrocomm3 24660 15.4 0.038 0.039 1.343 3.016 hybrid_alltoall_any 4261 16.3 0.122 0.463 2.312 2.988 wfi_extrapolate 11 7.9 0.001 0.001 2.962 2.962 make_images_data 4110 15.4 0.049 0.052 2.611 2.950 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.744 2.867 mp_alltoall_i22 605 13.7 1.644 2.801 1.644 2.801 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.779 2.781 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.768 2.780 cp_fm_diag_elpa_base 48 14.0 2.620 2.680 2.766 2.779 cp_fm_cholesky_invert 11 10.9 2.584 2.593 2.584 2.593 multiply_cannon_sync_h2d 24660 15.4 2.354 2.523 2.354 2.523 acc_transpose_blocks 24660 15.4 0.109 0.113 2.411 2.517 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.426 2.466 fft3d_ps 1111 14.6 1.153 1.198 2.445 2.461 grid_collocate_task_list 110 9.6 2.264 2.397 2.264 2.397 calculate_first_density_matrix 1 7.0 0.000 0.000 2.376 2.379 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 2.234 2.234 potential_pw2rs 110 12.3 0.013 0.013 2.015 2.019 mp_alltoall_d11v 2046 13.8 1.725 1.937 1.725 1.937 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.784 1.824 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.617 1.722 cp_fm_cholesky_decompose 22 10.9 1.619 1.664 1.619 1.664 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.627 1.637 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.549 1.565 multiply_cannon_metrocomm4 20550 15.4 0.059 0.062 0.860 1.522 mp_sum_l 6594 12.7 0.912 1.507 0.912 1.507 acc_transpose_blocks_sync 73980 16.4 1.390 1.490 1.390 1.490 mp_irecv_dv 62702 16.1 0.754 1.440 0.754 1.440 transfer_rs2pw 451 10.6 0.005 0.006 1.149 1.378 mp_waitany 13376 13.8 1.022 1.317 1.022 1.317 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=65.161000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=686.363636, yerr=9.167856 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1aa3ad7d27726d25ce9c9edce445974c62615587_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 855.597056E+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 0.009 0.028 54.761 54.761 qs_mol_dyn_low 1 2.0 0.003 0.003 54.538 54.547 qs_forces 11 3.9 0.003 0.003 54.473 54.473 qs_energies 11 4.9 0.001 0.002 50.827 50.831 scf_env_do_scf 11 5.9 0.000 0.001 42.671 42.671 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 35.058 35.059 velocity_verlet 10 3.0 0.002 0.002 30.902 30.905 dbcsr_multiply_generic 2055 12.4 0.107 0.109 23.228 23.416 qs_scf_new_mos 99 7.5 0.001 0.001 20.522 20.557 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.521 20.556 ot_scf_mini 99 9.5 0.002 0.002 19.271 19.295 multiply_cannon 2055 13.4 0.237 0.250 17.647 19.143 multiply_cannon_loop 2055 14.4 0.603 0.625 16.387 16.811 rebuild_ks_matrix 110 8.3 0.000 0.000 11.622 11.652 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.622 11.652 ot_mini 99 10.5 0.001 0.001 10.838 10.855 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.335 10.362 multiply_cannon_multrec 8220 15.4 3.168 4.406 7.642 8.563 init_scf_loop 11 6.9 0.000 0.000 7.568 7.570 mp_waitall_1 103326 16.6 5.766 7.456 5.766 7.456 qs_ot_get_derivative 99 11.5 0.001 0.001 7.153 7.177 sum_up_and_integrate 110 10.3 0.001 0.002 6.185 6.199 integrate_v_rspace 110 11.3 0.003 0.003 6.158 6.172 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.966 5.982 calculate_rho_elec 110 8.6 0.113 0.113 5.966 5.981 prepare_preconditioner 11 7.9 0.000 0.000 5.930 5.937 make_preconditioner 11 8.9 0.000 0.000 5.930 5.937 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.498 5.574 dbcsr_mm_accdrv_process 17442 15.9 3.091 4.194 4.334 5.239 init_scf_run 11 5.9 0.000 0.001 5.040 5.040 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.039 5.040 qs_ot_get_p 110 10.4 0.001 0.001 4.500 4.516 make_m2s 4110 13.4 0.037 0.039 4.156 4.450 multiply_cannon_metrocomm3 8220 15.4 0.018 0.019 3.034 4.445 make_images 4110 14.4 0.644 0.705 4.027 4.321 pw_transfer 1331 11.6 0.066 0.071 4.048 4.056 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.941 3.952 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.811 3.860 apply_single 110 13.6 0.000 0.000 3.811 3.859 ot_diis_step 99 11.5 0.012 0.015 3.663 3.663 grid_integrate_task_list 110 12.3 3.383 3.526 3.383 3.526 fft_wrap_pw1pw2_140 451 13.1 0.721 0.732 3.430 3.444 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.172 3.175 density_rs2pw 110 9.6 0.004 0.004 2.989 3.110 cp_dbcsr_syevd 48 12.0 0.003 0.003 2.836 2.837 cp_fm_cholesky_invert 11 10.9 2.831 2.835 2.831 2.835 hybrid_alltoall_any 4261 16.3 0.199 0.856 2.227 2.745 wfi_extrapolate 11 7.9 0.001 0.001 2.695 2.695 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 2.667 2.667 make_images_data 4110 15.4 0.040 0.046 2.237 2.611 grid_collocate_task_list 110 9.6 2.364 2.555 2.364 2.555 calculate_dm_sparse 110 9.5 0.001 0.001 2.481 2.516 multiply_cannon_sync_h2d 8220 15.4 2.385 2.508 2.385 2.508 fft3d_ps 1111 14.6 1.290 1.297 2.481 2.499 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.494 2.495 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.341 2.348 cp_fm_diag_elpa_base 48 14.0 2.284 2.305 2.339 2.346 calculate_first_density_matrix 1 7.0 0.000 0.000 2.250 2.251 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.065 2.188 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.139 2.155 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.782 2.004 potential_pw2rs 110 12.3 0.016 0.016 1.991 1.995 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.883 1.895 mp_alltoall_d11v 2046 13.8 1.625 1.797 1.625 1.797 cp_fm_cholesky_decompose 22 10.9 1.634 1.652 1.634 1.652 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.614 1.616 qs_env_update_s_mstruct 11 6.9 0.002 0.006 1.491 1.607 mp_allgather_i34 2055 14.4 0.433 1.519 0.433 1.519 dbcsr_complete_redistribute 325 12.2 0.554 0.574 1.407 1.508 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.432 1.444 jit_kernel_multiply 8 15.7 0.928 1.421 0.928 1.421 acc_transpose_blocks 8220 15.4 0.040 0.041 1.358 1.395 qs_create_task_list 11 7.9 0.001 0.001 1.204 1.303 generate_qs_task_list 11 8.9 0.374 0.442 1.203 1.302 transfer_rs2pw 451 10.6 0.005 0.005 1.093 1.241 mp_waitany 9240 13.8 1.045 1.196 1.045 1.196 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.135 1.162 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=54.761000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=805.454545, yerr=12.830104 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1aa3ad7d27726d25ce9c9edce445974c62615587_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 1.364308E+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 0.013 0.031 89.640 89.641 qs_mol_dyn_low 1 2.0 0.003 0.003 89.344 89.355 qs_forces 11 3.9 0.003 0.003 89.278 89.278 qs_energies 11 4.9 0.004 0.004 85.096 85.098 scf_env_do_scf 11 5.9 0.000 0.001 75.003 75.003 velocity_verlet 10 3.0 0.002 0.002 57.118 57.124 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 46.160 46.162 dbcsr_multiply_generic 2055 12.4 0.122 0.127 30.184 30.327 init_scf_loop 11 6.9 0.000 0.000 28.770 28.773 qs_scf_new_mos 99 7.5 0.001 0.001 27.946 28.036 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.945 28.035 prepare_preconditioner 11 7.9 0.000 0.000 26.707 26.720 make_preconditioner 11 8.9 0.000 0.000 26.707 26.720 ot_scf_mini 99 9.5 0.002 0.002 26.136 26.218 make_full_inverse_cholesky 11 9.9 0.000 0.000 21.077 26.176 multiply_cannon 2055 13.4 0.333 0.357 22.736 23.384 multiply_cannon_loop 2055 14.4 0.822 0.831 20.887 21.301 cp_fm_upper_to_full 70 14.2 12.735 18.085 12.735 18.085 ot_mini 99 10.5 0.001 0.001 14.700 14.777 rebuild_ks_matrix 110 8.3 0.000 0.001 14.185 14.276 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 14.184 14.276 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.853 12.936 dbcsr_complete_redistribute 325 12.2 1.010 1.027 7.231 10.290 qs_ot_get_derivative 99 11.5 0.001 0.001 9.897 9.977 multiply_cannon_multrec 8220 15.4 4.085 4.216 9.773 9.959 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.202 9.267 mp_waitall_1 84994 16.7 7.889 8.992 7.889 8.992 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.614 8.650 mp_alltoall_i22 605 13.7 5.254 8.331 5.254 8.331 qs_rho_update_rho_low 110 7.6 0.000 0.001 7.605 7.645 calculate_rho_elec 110 8.6 0.222 0.223 7.605 7.644 sum_up_and_integrate 110 10.3 0.002 0.002 7.044 7.062 integrate_v_rspace 110 11.3 0.004 0.004 7.016 7.034 qs_ot_get_p 110 10.4 0.001 0.001 6.062 6.158 make_m2s 4110 13.4 0.043 0.043 5.505 6.029 pw_transfer 1331 11.6 0.075 0.075 5.953 5.964 fft_wrap_pw1pw2 1111 12.6 0.009 0.010 5.836 5.847 make_images 4110 14.4 0.877 0.934 5.315 5.838 init_scf_run 11 5.9 0.000 0.001 5.809 5.809 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.809 5.809 dbcsr_mm_accdrv_process 11614 15.7 3.860 4.125 5.539 5.729 cp_fm_cholesky_invert 11 10.9 5.703 5.709 5.703 5.709 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.907 5.414 apply_single 110 13.6 0.000 0.000 4.907 5.414 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 5.078 5.404 fft_wrap_pw1pw2_140 451 13.1 1.366 1.368 5.099 5.109 ot_diis_step 99 11.5 0.015 0.016 4.781 4.781 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.378 4.385 density_rs2pw 110 9.6 0.004 0.004 4.038 4.054 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.850 3.850 grid_integrate_task_list 110 12.3 3.687 3.756 3.687 3.756 hybrid_alltoall_any 4261 16.3 0.263 0.561 2.933 3.687 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 3.668 3.668 make_images_data 4110 15.4 0.045 0.048 2.926 3.625 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.097 3.570 fft3d_ps 1111 14.6 1.884 1.895 3.541 3.553 wfi_extrapolate 11 7.9 0.001 0.001 3.432 3.432 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.223 3.223 cp_fm_diag_elpa_base 48 14.0 2.687 2.881 3.222 3.222 calculate_dm_sparse 110 9.5 0.001 0.001 3.187 3.202 multiply_cannon_sync_h2d 8220 15.4 3.134 3.162 3.134 3.162 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.933 2.936 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.749 2.794 grid_collocate_task_list 110 9.6 2.675 2.685 2.675 2.685 potential_pw2rs 110 12.3 0.021 0.022 2.537 2.543 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.397 2.426 calculate_first_density_matrix 1 7.0 0.000 0.000 2.253 2.257 qs_env_update_s_mstruct 11 6.9 0.001 0.001 2.192 2.241 mp_alltoall_d11v 2046 13.8 2.166 2.228 2.166 2.228 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.096 2.191 cp_fm_cholesky_decompose 22 10.9 2.094 2.121 2.094 2.121 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.930 1.939 qs_create_task_list 11 7.9 0.001 0.001 1.882 1.926 generate_qs_task_list 11 8.9 0.733 0.786 1.882 1.925 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.815 1.863 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=89.641000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1230.363636, yerr=58.487408 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1aa3ad7d27726d25ce9c9edce445974c62615587_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 1.086553E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 11851392 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.8 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 625.242112E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10258080 MPI messages size (bytes): total size 4.456715E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.459031E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3542056 94711185408 32768 < size <= 131072 1282176 73356279808 131072 < size <= 4194304 5107038 3151762421624 4194304 < size <= 16777216 259842 1136842803272 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4007 57688. MP_Allreduce 11097 796. MP_Sync 86 MP_Alltoall 2210 2272107. MP_SendRecv 24130 18752. MP_ISendRecv 24130 18752. MP_Wait 42150 MP_ISend 15900 108037. MP_IRecv 15900 108037. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.045 205.256 205.259 qs_mol_dyn_low 1 2.0 0.007 0.012 204.500 204.565 qs_forces 11 3.9 0.005 0.006 204.350 204.354 qs_energies 11 4.9 0.004 0.005 198.763 198.782 scf_env_do_scf 11 5.9 0.001 0.001 182.369 182.373 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 162.085 162.087 dbcsr_multiply_generic 2485 12.5 0.177 0.182 125.429 126.691 qs_scf_new_mos 116 7.6 0.001 0.001 123.377 123.607 qs_scf_loop_do_ot 116 8.6 0.001 0.001 123.376 123.606 velocity_verlet 10 3.0 0.005 0.009 121.906 121.909 ot_scf_mini 116 9.6 0.003 0.004 116.719 116.980 multiply_cannon 2485 13.5 0.232 0.241 101.264 104.016 multiply_cannon_loop 2485 14.5 2.381 2.432 99.186 101.932 ot_mini 116 10.6 0.001 0.002 66.547 66.809 multiply_cannon_multrec 59640 15.5 31.662 33.662 41.371 43.551 qs_ot_get_derivative 116 11.6 0.001 0.002 41.716 42.000 rebuild_ks_matrix 127 8.3 0.001 0.001 33.125 33.621 qs_ks_build_kohn_sham_matrix 127 9.3 0.015 0.023 33.124 33.620 mp_waitall_1 264810 16.5 28.960 31.689 28.960 31.689 qs_ks_update_qs_env 127 7.6 0.001 0.001 29.699 30.105 multiply_cannon_sync_h2d 59640 15.5 26.161 28.885 26.161 28.885 qs_ot_get_p 127 10.4 0.001 0.001 28.117 28.512 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 24.264 25.696 apply_single 127 13.6 0.001 0.001 24.263 25.696 ot_diis_step 116 11.6 0.008 0.012 24.563 24.564 qs_ot_p2m_diag 82 11.4 0.079 0.092 21.613 21.682 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 20.175 20.351 init_scf_loop 11 6.9 0.000 0.001 20.211 20.212 cp_dbcsr_syevd 82 12.4 0.005 0.005 18.813 18.814 multiply_cannon_metrocomm3 59640 15.5 0.114 0.119 16.281 18.687 cp_fm_diag_elpa 82 13.4 0.000 0.000 15.943 15.980 cp_fm_diag_elpa_base 82 14.4 15.878 15.909 15.939 15.974 prepare_preconditioner 11 7.9 0.000 0.000 15.619 15.668 make_preconditioner 11 8.9 0.000 0.000 15.619 15.668 make_full_inverse_cholesky 11 9.9 0.000 0.000 14.839 15.015 make_m2s 4970 13.5 0.103 0.112 13.607 14.001 make_images 4970 14.5 0.394 0.413 13.424 13.823 sum_up_and_integrate 127 10.3 0.002 0.004 13.714 13.729 integrate_v_rspace 127 11.3 0.003 0.004 13.656 13.676 qs_rho_update_rho_low 127 7.7 0.001 0.002 12.829 12.927 calculate_rho_elec 127 8.7 0.045 0.063 12.828 12.926 init_scf_run 11 5.9 0.000 0.001 12.229 12.230 scf_env_initial_rho_setup 11 6.9 0.001 0.002 12.229 12.230 mp_sum_l 7884 12.9 9.327 11.140 9.327 11.140 dbcsr_mm_accdrv_process 123452 16.2 4.693 4.839 9.280 9.809 wfi_extrapolate 11 7.9 0.001 0.005 9.015 9.015 cp_fm_cholesky_invert 11 10.9 8.865 8.873 8.865 8.873 calculate_dm_sparse 127 9.5 0.001 0.001 8.464 8.606 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.159 8.317 qs_ot_get_orbitals 116 10.6 0.001 0.001 8.025 8.104 multiply_cannon_metrocomm1 59640 15.5 0.091 0.094 6.378 7.962 make_images_data 4970 15.5 0.066 0.072 6.666 7.589 pw_transfer 1535 11.6 0.073 0.086 7.420 7.589 grid_integrate_task_list 127 12.3 6.990 7.407 6.990 7.407 fft_wrap_pw1pw2 1281 12.7 0.010 0.012 7.221 7.386 hybrid_alltoall_any 5155 16.4 0.290 2.235 5.851 7.364 density_rs2pw 127 9.7 0.006 0.006 6.538 7.205 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.653 6.665 fft_wrap_pw1pw2_140 519 13.2 0.852 0.892 6.384 6.553 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.761 5.869 fft3d_ps 1281 14.7 2.176 2.812 5.242 5.555 mp_alltoall_d11v 2401 14.1 4.257 5.428 4.257 5.428 grid_collocate_task_list 127 9.7 4.798 5.223 4.798 5.223 cp_fm_cholesky_decompose 22 10.9 4.566 4.581 4.566 4.581 potential_pw2rs 127 12.3 0.009 0.010 4.308 4.320 mp_sum_d 4456 12.1 3.480 4.247 3.480 4.247 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=205.259000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=589.727273, yerr=6.209830 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1aa3ad7d27726d25ce9c9edce445974c62615587_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 2.166472E+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 5925696 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.6 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 834.007040E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2385600 MPI messages size (bytes): total size 4.069300E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.705776E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70188 2295595008 32768 < size <= 131072 716032 54973693952 131072 < size <= 4194304 1363760 1386318135296 4194304 < size <= 16777216 153648 1453842923456 16777216 < size 67056 1171888537600 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4003 58199. MP_Allreduce 11086 960. MP_Sync 86 MP_Alltoall 1955 5671270. MP_SendRecv 11938 47072. MP_ISendRecv 11938 47072. MP_Wait 25718 MP_ISend 11660 212488. MP_IRecv 11660 212488. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.062 0.118 188.143 188.146 qs_mol_dyn_low 1 2.0 0.007 0.010 187.530 187.544 qs_forces 11 3.9 0.004 0.005 187.428 187.433 qs_energies 11 4.9 0.003 0.006 180.662 180.675 scf_env_do_scf 11 5.9 0.001 0.003 164.288 164.298 scf_env_do_scf_inner_loop 116 6.6 0.003 0.009 131.619 131.622 velocity_verlet 10 3.0 0.006 0.012 117.754 117.756 dbcsr_multiply_generic 2485 12.5 0.187 0.193 96.609 98.004 qs_scf_new_mos 116 7.6 0.001 0.001 93.449 93.998 qs_scf_loop_do_ot 116 8.6 0.001 0.002 93.448 93.998 ot_scf_mini 116 9.6 0.004 0.005 88.676 89.205 multiply_cannon 2485 13.5 0.499 0.551 76.632 81.255 multiply_cannon_loop 2485 14.5 1.549 1.630 73.312 75.986 ot_mini 116 10.6 0.001 0.001 49.921 50.489 mp_waitall_1 212858 16.6 23.736 39.237 23.736 39.237 multiply_cannon_multrec 29820 15.5 20.850 25.137 31.370 35.892 rebuild_ks_matrix 127 8.3 0.001 0.001 32.054 32.751 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.024 32.053 32.750 init_scf_loop 11 6.9 0.002 0.010 32.580 32.581 qs_ks_update_qs_env 127 7.6 0.001 0.001 28.804 29.436 multiply_cannon_metrocomm3 29820 15.5 0.098 0.104 15.432 29.064 qs_ot_get_derivative 116 11.6 0.001 0.002 28.108 28.625 prepare_preconditioner 11 7.9 0.000 0.000 28.258 28.322 make_preconditioner 11 8.9 0.001 0.003 28.258 28.322 make_full_inverse_cholesky 11 9.9 0.000 0.001 26.922 27.460 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 21.915 23.171 apply_single 127 13.6 0.001 0.001 21.914 23.171 ot_diis_step 116 11.6 0.014 0.015 21.639 21.641 qs_ot_get_p 127 10.4 0.001 0.001 20.977 21.637 multiply_cannon_sync_h2d 29820 15.5 17.954 21.188 17.954 21.188 cp_fm_cholesky_invert 11 10.9 16.489 16.502 16.489 16.502 qs_ot_p2m_diag 82 11.4 0.185 0.213 16.073 16.107 make_m2s 4970 13.5 0.085 0.091 14.057 15.682 make_images 4970 14.5 1.147 1.340 13.843 15.468 cp_dbcsr_syevd 82 12.4 0.010 0.040 14.885 14.886 sum_up_and_integrate 127 10.3 0.002 0.004 13.909 13.936 integrate_v_rspace 127 11.3 0.003 0.004 13.849 13.878 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.256 13.288 calculate_rho_elec 127 8.7 0.086 0.104 13.255 13.287 cp_fm_diag_elpa 82 13.4 0.001 0.002 11.694 11.723 cp_fm_diag_elpa_base 82 14.4 11.437 11.528 11.684 11.709 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 11.099 11.493 init_scf_run 11 5.9 0.000 0.001 11.458 11.459 scf_env_initial_rho_setup 11 6.9 0.001 0.002 11.457 11.459 multiply_cannon_metrocomm4 27335 15.5 0.098 0.114 3.808 11.018 mp_irecv_dv 68888 16.3 3.606 10.622 3.606 10.622 dbcsr_mm_accdrv_process 61726 16.2 5.410 6.035 9.972 10.497 make_images_data 4970 15.5 0.065 0.075 8.260 10.201 hybrid_alltoall_any 5155 16.4 0.343 1.479 6.959 9.613 pw_transfer 1535 11.6 0.084 0.094 8.509 8.561 wfi_extrapolate 11 7.9 0.001 0.001 8.352 8.352 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.288 8.342 grid_integrate_task_list 127 12.3 7.146 7.503 7.146 7.503 fft_wrap_pw1pw2_140 519 13.2 0.925 0.953 7.337 7.400 density_rs2pw 127 9.7 0.006 0.006 7.007 7.351 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.264 7.039 cp_fm_cholesky_decompose 22 10.9 6.895 6.984 6.895 6.984 calculate_dm_sparse 127 9.5 0.001 0.001 6.422 6.572 mp_sum_l 7884 12.9 4.167 6.219 4.167 6.219 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.084 6.092 fft3d_ps 1281 14.7 2.804 2.981 5.865 5.924 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 5.197 5.412 qs_ot_get_orbitals 116 10.6 0.001 0.001 5.320 5.397 grid_collocate_task_list 127 9.7 4.971 5.278 4.971 5.278 mp_allgather_i34 2485 14.5 1.864 4.956 1.864 4.956 mp_alltoall_d11v 2401 14.1 4.126 4.744 4.126 4.744 potential_pw2rs 127 12.3 0.015 0.018 4.452 4.466 mp_sum_d 4454 12.1 2.687 4.060 2.687 4.060 dbcsr_complete_redistribute 393 12.7 0.765 0.850 3.112 3.938 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=188.146000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=795.454545, yerr=1.724879 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1aa3ad7d27726d25ce9c9edce445974c62615587_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 957.014016E+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 4002 58208. MP_Allreduce 11082 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 0.010 0.028 172.882 172.883 qs_mol_dyn_low 1 2.0 0.004 0.011 172.505 172.519 qs_forces 11 3.9 0.004 0.005 172.410 172.412 qs_energies 11 4.9 0.002 0.002 165.878 165.888 scf_env_do_scf 11 5.9 0.001 0.001 150.527 150.528 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 116.042 116.042 velocity_verlet 10 3.0 0.003 0.011 111.092 111.094 dbcsr_multiply_generic 2507 12.6 0.186 0.190 81.533 82.659 qs_scf_new_mos 117 7.6 0.001 0.001 79.285 79.618 qs_scf_loop_do_ot 117 8.6 0.001 0.001 79.284 79.617 ot_scf_mini 117 9.6 0.003 0.004 75.060 75.458 multiply_cannon 2507 13.6 0.507 0.528 62.128 67.342 multiply_cannon_loop 2507 14.6 1.126 1.187 59.216 61.802 ot_mini 117 10.6 0.001 0.001 42.445 42.862 mp_waitall_1 170520 16.6 24.350 34.552 24.350 34.552 init_scf_loop 11 6.9 0.000 0.000 34.390 34.390 rebuild_ks_matrix 128 8.3 0.001 0.001 29.969 30.488 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 29.968 30.487 prepare_preconditioner 11 7.9 0.000 0.000 30.435 30.475 make_preconditioner 11 8.9 0.000 0.000 30.435 30.475 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.092 29.456 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.006 27.472 multiply_cannon_multrec 20056 15.6 13.190 15.958 22.517 25.337 multiply_cannon_metrocomm3 20056 15.6 0.061 0.066 15.222 24.806 qs_ot_get_derivative 117 11.6 0.001 0.002 22.954 23.359 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.521 20.572 apply_single 128 13.6 0.001 0.001 19.521 20.572 ot_diis_step 117 11.6 0.018 0.018 19.377 19.378 qs_ot_get_p 128 10.4 0.001 0.001 18.041 18.537 make_m2s 5014 13.6 0.078 0.084 14.375 15.622 make_images 5014 14.6 1.137 1.219 14.139 15.382 multiply_cannon_sync_h2d 20056 15.6 13.655 15.101 13.655 15.101 cp_fm_cholesky_invert 11 10.9 14.173 14.182 14.173 14.182 sum_up_and_integrate 128 10.3 0.002 0.003 13.812 13.834 qs_ot_p2m_diag 83 11.4 0.265 0.272 13.769 13.779 integrate_v_rspace 128 11.3 0.003 0.004 13.753 13.773 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.422 13.465 calculate_rho_elec 128 8.7 0.129 0.143 13.422 13.464 cp_dbcsr_syevd 83 12.4 0.005 0.005 12.721 12.722 init_scf_run 11 5.9 0.000 0.001 10.345 10.346 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.345 10.345 make_images_data 5014 15.6 0.064 0.074 8.584 10.231 hybrid_alltoall_any 5200 16.5 0.445 2.023 7.400 9.890 cp_fm_diag_elpa 83 13.4 0.000 0.000 9.592 9.615 cp_fm_diag_elpa_base 83 14.4 9.188 9.344 9.589 9.612 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.042 9.316 multiply_cannon_metrocomm4 17549 15.6 0.065 0.073 3.422 9.193 dbcsr_mm_accdrv_process 41502 16.2 5.626 5.935 8.783 8.952 mp_irecv_dv 50230 16.2 3.293 8.935 3.293 8.935 pw_transfer 1547 11.6 0.084 0.101 8.682 8.785 fft_wrap_pw1pw2 1291 12.7 0.010 0.010 8.459 8.566 grid_integrate_task_list 128 12.3 7.296 7.704 7.296 7.704 fft_wrap_pw1pw2_140 523 13.2 1.055 1.093 7.546 7.658 wfi_extrapolate 11 7.9 0.001 0.001 7.350 7.350 cp_fm_cholesky_decompose 22 10.9 7.206 7.275 7.206 7.275 density_rs2pw 128 9.7 0.005 0.006 6.785 7.226 cp_fm_upper_to_full 105 14.8 5.644 7.112 5.644 7.112 dbcsr_complete_redistribute 395 12.7 1.164 1.192 4.504 6.203 calculate_dm_sparse 128 9.5 0.001 0.001 5.847 5.959 fft3d_ps 1291 14.7 2.814 3.048 5.779 5.868 grid_collocate_task_list 128 9.7 5.163 5.641 5.163 5.641 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.371 5.376 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.671 5.336 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.372 5.066 mp_alltoall_d11v 2415 14.1 4.306 5.050 4.306 5.050 mp_sum_l 7950 12.9 3.189 4.808 3.189 4.808 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.498 4.630 mp_allgather_i34 2507 14.6 1.425 4.472 1.425 4.472 potential_pw2rs 128 12.3 0.020 0.022 4.240 4.249 transfer_fm_to_dbcsr 11 9.9 0.018 0.025 2.323 3.997 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.903 3.935 mp_alltoall_i22 716 14.1 1.913 3.802 1.913 3.802 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.729 3.730 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.459 3.513 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=172.883000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=909.272727, yerr=4.350359 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1aa3ad7d27726d25ce9c9edce445974c62615587_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 4.353788E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 5977344 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.143235E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1143192 MPI messages size (bytes): total size 2.023815E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.770319E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 319024 36042702848 131072 < size <= 4194304 715736 785529176064 4194304 < size <= 16777216 70320 665379241840 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58348. MP_Allreduce 11057 1083. MP_Sync 87 MP_Alltoall 1712 12503084. MP_SendRecv 5888 75008. MP_ISendRecv 5888 75008. MP_Wait 22442 MP_ISend 14952 244818. MP_IRecv 14952 244818. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.030 187.843 187.844 qs_mol_dyn_low 1 2.0 0.004 0.011 187.486 187.500 qs_forces 11 3.9 0.004 0.004 187.356 187.363 qs_energies 11 4.9 0.002 0.003 180.222 180.232 scf_env_do_scf 11 5.9 0.001 0.001 163.510 163.526 velocity_verlet 10 3.0 0.003 0.015 124.325 124.329 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 117.539 117.541 qs_scf_new_mos 117 7.6 0.001 0.001 81.090 81.380 qs_scf_loop_do_ot 117 8.6 0.001 0.001 81.090 81.379 dbcsr_multiply_generic 2507 12.6 0.193 0.200 80.471 81.148 ot_scf_mini 117 9.6 0.003 0.003 76.546 76.880 multiply_cannon 2507 13.6 0.553 0.587 55.904 58.930 multiply_cannon_loop 2507 14.6 1.821 1.915 52.098 54.000 init_scf_loop 11 6.9 0.000 0.000 45.849 45.850 ot_mini 117 10.6 0.001 0.001 42.964 43.282 prepare_preconditioner 11 7.9 0.000 0.000 41.781 41.820 make_preconditioner 11 8.9 0.000 0.000 41.781 41.820 make_full_inverse_cholesky 11 9.9 0.010 0.021 35.560 40.512 multiply_cannon_multrec 30084 15.6 13.600 18.955 26.519 31.432 rebuild_ks_matrix 128 8.3 0.001 0.001 29.374 29.679 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.018 29.373 29.678 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.488 26.757 mp_waitall_1 147882 16.7 17.060 26.159 17.060 26.159 qs_ot_get_derivative 117 11.6 0.001 0.002 23.431 23.768 make_m2s 5014 13.6 0.092 0.098 20.137 21.307 make_images 5014 14.6 1.922 2.254 19.827 20.995 qs_ot_get_p 128 10.4 0.001 0.001 19.795 20.150 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 18.992 19.461 apply_single 128 13.6 0.001 0.001 18.992 19.461 ot_diis_step 117 11.6 0.017 0.018 19.414 19.416 cp_fm_upper_to_full 105 14.8 11.527 16.907 11.527 16.907 cp_fm_cholesky_invert 11 10.9 15.906 15.915 15.906 15.915 qs_ot_p2m_diag 83 11.4 0.342 0.389 15.553 15.603 multiply_cannon_metrocomm3 30084 15.6 0.049 0.051 6.331 15.276 cp_dbcsr_syevd 83 12.4 0.005 0.005 14.150 14.152 sum_up_and_integrate 128 10.3 0.002 0.003 14.045 14.076 integrate_v_rspace 128 11.3 0.003 0.004 13.985 14.018 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.888 13.938 calculate_rho_elec 128 8.7 0.170 0.186 13.888 13.937 dbcsr_mm_accdrv_process 62264 16.2 8.547 9.207 12.487 12.998 dbcsr_complete_redistribute 395 12.7 1.483 1.620 8.933 12.670 make_images_data 5014 15.6 0.065 0.072 10.757 12.521 hybrid_alltoall_any 5200 16.5 0.528 2.190 9.582 11.751 copy_fm_to_dbcsr 209 11.7 0.001 0.002 7.558 11.315 multiply_cannon_sync_h2d 30084 15.6 10.475 11.202 10.475 11.202 cp_fm_diag_elpa 83 13.4 0.000 0.000 10.921 10.940 cp_fm_diag_elpa_base 83 14.4 9.935 10.264 10.913 10.932 init_scf_run 11 5.9 0.000 0.001 10.780 10.781 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.780 10.781 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.885 10.127 transfer_fm_to_dbcsr 11 9.9 0.001 0.007 6.199 9.880 pw_transfer 1547 11.6 0.086 0.102 9.306 9.367 mp_alltoall_i22 716 14.1 5.454 9.329 5.454 9.329 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 9.079 9.145 fft_wrap_pw1pw2_140 523 13.2 1.248 1.286 8.058 8.150 grid_integrate_task_list 128 12.3 7.539 7.854 7.539 7.854 wfi_extrapolate 11 7.9 0.001 0.001 7.686 7.687 cp_fm_cholesky_decompose 22 10.9 7.481 7.570 7.481 7.570 density_rs2pw 128 9.7 0.005 0.006 6.913 7.249 multiply_cannon_metrocomm4 25070 15.6 0.079 0.088 2.820 7.054 mp_irecv_dv 76098 16.2 2.666 6.769 2.666 6.769 calculate_dm_sparse 128 9.5 0.001 0.001 6.251 6.351 fft3d_ps 1291 14.7 2.999 3.052 6.128 6.188 grid_collocate_task_list 128 9.7 5.283 5.691 5.283 5.691 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.453 5.491 mp_alltoall_d11v 2415 14.1 4.913 5.379 4.913 5.379 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.495 4.592 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.384 4.466 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 4.449 4.450 potential_pw2rs 128 12.3 0.023 0.024 4.312 4.320 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.205 4.258 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=187.844000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1079.727273, yerr=20.863587 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1aa3ad7d27726d25ce9c9edce445974c62615587_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456039424 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613072052224 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176077312 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176077312 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.233020E+12 0.0% 0.0% 100.0% flops max/rank 5.910120E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383904 0.0% 0.0% 100.0% number of processed stacks 1976928 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.9 marketing flops 145.650931E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 1.549238E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 242784 MPI messages size (bytes): total size 1.341806E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.526748E+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 115008 60297314304 4194304 < size <= 16777216 105840 554906419200 16777216 < size 20352 726592466352 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 9010 51. MP_Alltoall 9724 794507. MP_ISend 40420 2100460. MP_IRecv 40420 2099564. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4045 57596. MP_Allreduce 11189 1162. MP_Sync 88 MP_Alltoall 1724 18848034. MP_SendRecv 3870 122880. MP_ISendRecv 3870 122880. MP_Wait 16244 MP_ISend 10760 423501. MP_IRecv 10760 423501. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.032 169.603 169.604 qs_mol_dyn_low 1 2.0 0.003 0.003 169.221 169.235 qs_forces 11 3.9 0.004 0.005 169.128 169.130 qs_energies 11 4.9 0.002 0.002 161.728 161.734 scf_env_do_scf 11 5.9 0.001 0.001 144.711 144.720 velocity_verlet 10 3.0 0.002 0.002 111.907 111.911 scf_env_do_scf_inner_loop 118 6.6 0.005 0.009 109.117 109.119 dbcsr_multiply_generic 2529 12.6 0.186 0.194 71.952 72.289 qs_scf_new_mos 118 7.6 0.001 0.001 71.475 71.568 qs_scf_loop_do_ot 118 8.6 0.001 0.001 71.474 71.567 ot_scf_mini 118 9.6 0.003 0.004 67.024 67.062 multiply_cannon 2529 13.6 0.564 0.601 53.372 56.875 multiply_cannon_loop 2529 14.6 0.809 0.846 50.399 51.061 ot_mini 118 10.6 0.001 0.001 37.376 37.407 init_scf_loop 11 6.9 0.000 0.000 35.446 35.448 prepare_preconditioner 11 7.9 0.000 0.000 31.609 31.638 make_preconditioner 11 8.9 0.000 0.000 31.609 31.638 mp_waitall_1 126876 16.7 24.582 31.095 24.582 31.095 make_full_inverse_cholesky 11 9.9 0.014 0.025 29.542 29.788 rebuild_ks_matrix 129 8.3 0.001 0.001 29.397 29.486 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.019 29.396 29.485 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.813 26.891 multiply_cannon_multrec 10116 15.6 10.450 14.945 18.066 21.133 qs_ot_get_derivative 118 11.6 0.002 0.002 20.357 20.390 multiply_cannon_metrocomm3 10116 15.6 0.024 0.026 12.516 19.496 cp_fm_cholesky_invert 11 10.9 18.071 18.076 18.071 18.076 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 17.033 17.147 apply_single 129 13.6 0.001 0.001 17.032 17.147 ot_diis_step 118 11.6 0.020 0.020 16.948 16.949 qs_ot_get_p 129 10.4 0.001 0.001 16.870 16.945 make_m2s 5058 13.6 0.065 0.069 14.758 15.656 make_images 5058 14.6 2.166 2.606 14.446 15.338 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.382 14.410 calculate_rho_elec 129 8.7 0.255 0.266 14.381 14.409 sum_up_and_integrate 129 10.3 0.002 0.002 14.216 14.270 integrate_v_rspace 129 11.3 0.004 0.004 14.156 14.212 qs_ot_p2m_diag 84 11.4 0.502 0.508 13.136 13.152 cp_dbcsr_syevd 84 12.4 0.005 0.006 11.963 11.963 multiply_cannon_sync_h2d 10116 15.6 11.001 11.516 11.001 11.516 init_scf_run 11 5.9 0.000 0.001 10.227 10.228 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.227 10.227 pw_transfer 1559 11.6 0.086 0.093 9.930 9.966 hybrid_alltoall_any 5245 16.5 0.847 3.794 8.372 9.880 make_images_data 5058 15.6 0.055 0.065 8.534 9.841 fft_wrap_pw1pw2 1301 12.7 0.010 0.010 9.706 9.747 cp_fm_diag_elpa 84 13.4 0.000 0.000 9.064 9.076 cp_fm_diag_elpa_base 84 14.4 8.826 8.905 9.061 9.073 fft_wrap_pw1pw2_140 527 13.2 1.690 1.726 8.619 8.660 grid_integrate_task_list 129 12.3 7.817 8.095 7.817 8.095 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 8.010 8.042 cp_fm_cholesky_decompose 22 10.9 7.921 8.033 7.921 8.033 dbcsr_mm_accdrv_process 20934 16.1 3.175 4.549 7.264 7.962 density_rs2pw 129 9.7 0.005 0.005 6.994 7.459 multiply_cannon_metrocomm1 10116 15.6 0.030 0.031 4.313 7.434 wfi_extrapolate 11 7.9 0.001 0.001 7.231 7.231 calculate_dm_sparse 129 9.5 0.001 0.001 6.088 6.157 fft3d_ps 1301 14.7 3.166 3.263 6.123 6.145 grid_collocate_task_list 129 9.7 5.604 6.007 5.604 6.007 dbcsr_complete_redistribute 397 12.7 2.100 2.178 5.109 5.370 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.243 5.244 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.072 5.081 mp_alltoall_d11v 2429 14.1 4.602 5.037 4.602 5.037 mp_allgather_i34 2529 14.6 1.160 4.657 1.160 4.657 potential_pw2rs 129 12.3 0.027 0.028 4.336 4.343 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.020 4.080 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.596 3.865 multiply_cannon_metrocomm4 7587 15.6 0.026 0.029 1.725 3.784 mp_irecv_dv 29102 15.9 1.684 3.711 1.684 3.711 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.559 3.622 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.578 3.610 copy_fm_to_dbcsr 210 11.7 0.002 0.002 3.290 3.592 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.491 3.503 copy_dbcsr_to_fm 187 11.8 0.004 0.004 3.382 3.468 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=169.604000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1450.727273, yerr=28.948515 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1aa3ad7d27726d25ce9c9edce445974c62615587_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963542011904 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019182452736 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019182452736 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 92.796564E+12 0.0% 0.0% 100.0% flops max/rank 11.606412E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499488 0.0% 0.0% 100.0% number of processed stacks 1947808 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.6 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 3.177820E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 99400 MPI messages size (bytes): total size 1.127422E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.342272E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 44768 34745614336 4194304 < size <= 16777216 43984 376564613120 16777216 < size 10032 716108490000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3991 59293. MP_Allreduce 11055 1504. MP_Sync 86 MP_Alltoall 1700 36954339. MP_SendRecv 1778 218624. MP_ISendRecv 1778 218624. MP_Wait 9728 MP_ISend 6360 1080477. MP_IRecv 6360 1080477. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.041 0.059 288.724 288.725 qs_mol_dyn_low 1 2.0 0.003 0.003 288.147 288.182 qs_forces 11 3.9 0.005 0.005 288.047 288.050 qs_energies 11 4.9 0.002 0.002 279.143 279.150 scf_env_do_scf 11 5.9 0.001 0.001 256.659 256.668 velocity_verlet 10 3.0 0.002 0.002 209.068 209.076 scf_env_do_scf_inner_loop 116 6.6 0.004 0.009 132.342 132.345 init_scf_loop 11 6.9 0.000 0.000 124.047 124.053 prepare_preconditioner 11 7.9 0.000 0.000 119.199 119.218 make_preconditioner 11 8.9 0.000 0.000 119.199 119.218 make_full_inverse_cholesky 11 9.9 0.038 0.039 95.159 116.362 qs_scf_new_mos 116 7.6 0.001 0.001 88.008 88.058 qs_scf_loop_do_ot 116 8.6 0.001 0.001 88.007 88.057 ot_scf_mini 116 9.6 0.004 0.004 83.171 83.199 dbcsr_multiply_generic 2485 12.5 0.215 0.225 80.740 81.119 cp_fm_upper_to_full 104 14.8 52.922 75.919 52.922 75.919 multiply_cannon 2485 13.5 0.672 0.702 58.397 59.400 multiply_cannon_loop 2485 14.5 1.032 1.043 54.511 55.661 ot_mini 116 10.6 0.001 0.001 43.559 43.588 dbcsr_complete_redistribute 393 12.7 3.985 4.022 29.965 43.020 copy_fm_to_dbcsr 208 11.6 0.001 0.002 26.472 39.488 transfer_fm_to_dbcsr 11 9.9 0.030 0.030 23.997 36.799 mp_alltoall_i22 712 14.1 21.760 34.885 21.760 34.885 rebuild_ks_matrix 127 8.3 0.001 0.001 34.419 34.450 qs_ks_build_kohn_sham_matrix 127 9.3 0.018 0.018 34.418 34.449 cp_fm_cholesky_invert 11 10.9 33.162 33.168 33.162 33.168 qs_ks_update_qs_env 127 7.6 0.001 0.001 31.922 31.962 mp_waitall_1 102768 16.8 27.411 31.599 27.411 31.599 qs_ot_get_p 127 10.4 0.001 0.001 24.755 24.807 qs_ot_get_derivative 116 11.6 0.002 0.002 24.182 24.212 qs_ot_p2m_diag 82 11.4 0.868 0.874 20.827 20.856 multiply_cannon_metrocomm3 9940 15.5 0.025 0.025 18.829 19.956 ot_diis_step 116 11.6 0.021 0.022 19.348 19.348 make_m2s 4970 13.5 0.072 0.075 17.857 19.259 cp_dbcsr_syevd 82 12.4 0.006 0.006 19.056 19.059 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 18.698 18.821 apply_single 127 13.6 0.001 0.001 18.698 18.821 make_images 4970 14.5 3.030 3.225 17.379 18.781 multiply_cannon_multrec 9940 15.5 10.167 12.009 17.911 17.982 qs_rho_update_rho_low 127 7.7 0.001 0.001 17.154 17.174 calculate_rho_elec 127 8.7 0.474 0.475 17.154 17.174 sum_up_and_integrate 127 10.3 0.002 0.002 16.335 16.424 integrate_v_rspace 127 11.3 0.005 0.005 16.273 16.361 cp_fm_diag_elpa 82 13.4 0.000 0.000 15.944 15.945 cp_fm_diag_elpa_base 82 14.4 11.616 13.172 15.941 15.941 multiply_cannon_sync_h2d 9940 15.5 14.207 14.251 14.207 14.251 pw_transfer 1535 11.6 0.095 0.096 13.068 13.074 fft_wrap_pw1pw2 1281 12.7 0.011 0.011 12.832 12.837 init_scf_run 11 5.9 0.000 0.001 12.172 12.173 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.172 12.173 hybrid_alltoall_any 5155 16.4 1.307 3.061 10.084 12.117 make_images_data 4970 15.5 0.063 0.068 9.888 11.943 fft_wrap_pw1pw2_140 519 13.2 3.203 3.246 11.433 11.440 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.493 9.519 dbcsr_mm_accdrv_process 20590 16.0 4.244 6.086 7.495 9.326 wfi_extrapolate 11 7.9 0.001 0.001 9.032 9.032 cp_fm_cholesky_decompose 22 10.9 8.843 8.865 8.843 8.865 grid_integrate_task_list 127 12.3 8.495 8.685 8.495 8.685 density_rs2pw 127 9.7 0.005 0.005 8.260 8.316 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 8.170 8.171 fft3d_ps 1281 14.7 3.949 3.963 7.323 7.392 mp_alltoall_d11v 2401 14.1 6.849 6.982 6.849 6.982 calculate_dm_sparse 127 9.5 0.001 0.001 6.716 6.777 grid_collocate_task_list 127 9.7 6.367 6.446 6.367 6.446 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.361 6.415 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.128 6.217 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=288.725000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2891.454545, yerr=157.397796 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1aa3ad7d27726d25ce9c9edce445974c62615587_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 1.251545E+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 8020411. MP_SendRecv 285 19200. MP_ISendRecv 285 19200. MP_Wait 1017 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.036 84.914 84.916 qs_energies 1 2.0 0.000 0.001 84.422 84.430 ls_scf 1 3.0 0.001 0.002 83.518 83.526 dbcsr_multiply_generic 111 6.7 0.015 0.015 72.203 72.385 multiply_cannon 111 7.7 0.017 0.020 55.651 56.655 multiply_cannon_loop 111 8.7 0.230 0.251 52.285 53.502 ls_scf_main 1 4.0 0.000 0.002 52.058 52.060 density_matrix_trs4 2 5.0 0.002 0.003 46.541 46.601 ls_scf_init_scf 1 4.0 0.000 0.001 28.414 28.417 ls_scf_init_matrix_S 1 5.0 0.000 0.000 26.968 27.017 mp_waitall_1 11031 10.9 22.165 25.292 22.165 25.292 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 24.903 24.921 multiply_cannon_multrec 2664 9.7 8.130 8.968 15.504 17.388 multiply_cannon_sync_h2d 2664 9.7 13.612 16.100 13.612 16.100 make_m2s 222 7.7 0.009 0.011 12.912 13.329 make_images 222 8.7 0.099 0.110 12.890 13.308 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 9.608 12.473 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.390 8.801 make_images_data 222 9.7 0.004 0.005 7.497 8.077 dbcsr_mm_accdrv_process 4760 10.4 0.592 0.703 6.991 8.005 hybrid_alltoall_any 227 10.6 0.216 1.842 6.484 7.607 dbcsr_mm_accdrv_process_sort 4760 11.4 6.198 7.117 6.198 7.117 calculate_norms 4752 9.8 5.500 6.198 5.500 6.198 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.973 5.124 mp_sum_l 887 5.1 3.109 4.781 3.109 4.781 multiply_cannon_metrocomm4 2442 9.7 0.012 0.016 2.019 4.210 mp_irecv_dv 6231 10.9 2.002 4.178 2.002 4.178 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.363 3.843 make_images_sizes 222 9.7 0.000 0.000 0.650 3.656 mp_alltoall_i44 222 10.7 0.650 3.656 0.650 3.656 arnoldi_extremal 4 6.8 0.000 0.000 3.239 3.270 arnoldi_normal_ev 4 7.8 0.001 0.003 3.239 3.270 build_subspace 16 8.4 0.009 0.012 3.143 3.145 ls_scf_post 1 4.0 0.000 0.001 3.046 3.054 ls_scf_store_result 1 5.0 0.000 0.000 2.847 2.897 dbcsr_special_finalize 555 9.7 0.005 0.006 2.411 2.766 dbcsr_merge_single_wm 555 10.7 0.457 0.576 2.403 2.758 make_images_pack 222 9.7 2.207 2.624 2.209 2.625 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.310 2.555 dbcsr_sort_data 658 11.4 2.203 2.519 2.203 2.519 dbcsr_matrix_vector_mult_local 304 10.0 2.068 2.463 2.070 2.465 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.313 2.420 buffer_matrices_ensure_size 222 8.7 1.745 2.058 1.745 2.058 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.777 1.781 rebuild_ks_matrix 3 7.3 0.000 0.001 1.766 1.770 qs_ks_build_kohn_sham_matrix 3 8.3 0.003 0.013 1.766 1.770 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=84.916000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1132.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1aa3ad7d27726d25ce9c9edce445974c62615587_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 2.137133E+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 0.029 0.069 90.078 90.080 qs_energies 1 2.0 0.000 0.002 89.545 89.548 ls_scf 1 3.0 0.000 0.002 88.200 88.203 dbcsr_multiply_generic 111 6.7 0.015 0.017 74.421 74.696 multiply_cannon 111 7.7 0.028 0.041 52.525 57.578 ls_scf_main 1 4.0 0.000 0.001 54.655 54.660 multiply_cannon_loop 111 8.7 0.135 0.146 49.726 53.311 density_matrix_trs4 2 5.0 0.002 0.003 48.928 49.144 ls_scf_init_scf 1 4.0 0.000 0.001 29.904 29.906 mp_waitall_1 9105 10.9 20.576 29.090 20.576 29.090 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.719 28.801 multiply_cannon_multrec 1332 9.7 13.342 17.046 22.610 27.402 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 26.362 26.374 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.318 20.389 make_m2s 222 7.7 0.006 0.008 15.036 15.577 make_images 222 8.7 1.370 1.693 15.006 15.548 dbcsr_mm_accdrv_process 4041 10.4 0.367 0.539 8.859 10.400 dbcsr_mm_accdrv_process_sort 4041 11.4 8.377 9.861 8.377 9.861 hybrid_alltoall_any 227 10.6 0.541 2.548 8.057 9.592 make_images_data 222 9.7 0.004 0.005 8.735 9.591 mp_sum_l 887 5.1 5.389 8.480 5.389 8.480 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.192 7.805 mp_irecv_dv 3311 11.0 3.171 7.749 3.171 7.749 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.184 6.826 calculate_norms 2376 9.8 6.062 6.722 6.062 6.722 multiply_cannon_sync_h2d 1332 9.7 4.884 5.940 4.884 5.940 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.933 5.160 arnoldi_extremal 4 6.8 0.000 0.000 4.605 4.622 arnoldi_normal_ev 4 7.8 0.001 0.005 4.605 4.622 build_subspace 16 8.4 0.014 0.021 4.353 4.356 ls_scf_post 1 4.0 0.000 0.001 3.641 3.645 ls_scf_store_result 1 5.0 0.000 0.000 3.367 3.464 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.139 3.362 dbcsr_matrix_vector_mult_local 304 10.0 2.745 3.226 2.747 3.227 mp_allgather_i34 111 8.7 0.796 3.062 0.796 3.062 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.632 2.740 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.169 2.514 dbcsr_data_new 4174 10.1 2.114 2.389 2.114 2.389 make_images_pack 222 9.7 1.825 2.140 1.828 2.142 dbcsr_sort_data 436 11.2 1.840 2.091 1.840 2.091 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.858 1.860 rebuild_ks_matrix 3 7.3 0.000 0.001 1.843 1.846 qs_ks_build_kohn_sham_matrix 3 8.3 0.005 0.013 1.843 1.845 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=90.080000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1799.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1aa3ad7d27726d25ce9c9edce445974c62615587_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 2.816164E+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 0.027 0.040 91.700 91.702 qs_energies 1 2.0 0.000 0.002 91.100 91.103 ls_scf 1 3.0 0.000 0.002 89.661 89.665 dbcsr_multiply_generic 111 6.7 0.016 0.016 74.358 74.671 ls_scf_main 1 4.0 0.000 0.001 55.975 55.979 multiply_cannon 111 7.7 0.034 0.079 51.571 55.676 multiply_cannon_loop 111 8.7 0.116 0.129 48.871 52.680 density_matrix_trs4 2 5.0 0.002 0.003 50.012 50.156 mp_waitall_1 7281 11.0 23.138 32.976 23.138 32.976 ls_scf_init_scf 1 4.0 0.000 0.001 30.000 30.003 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.781 28.873 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.461 26.473 multiply_cannon_multrec 888 9.7 12.673 15.272 21.212 24.328 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.896 22.626 make_m2s 222 7.7 0.006 0.007 16.371 17.026 make_images 222 8.7 1.582 1.858 16.333 16.989 make_images_data 222 9.7 0.004 0.004 9.515 10.455 hybrid_alltoall_any 227 10.6 0.642 2.951 8.952 10.143 dbcsr_mm_accdrv_process 3754 10.4 0.311 0.490 8.076 9.280 dbcsr_mm_accdrv_process_sort 3754 11.4 7.640 8.790 7.640 8.790 mp_sum_l 887 5.1 4.710 7.403 4.710 7.403 multiply_cannon_sync_h2d 888 9.7 6.055 7.387 6.055 7.387 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.462 7.320 mp_irecv_dv 2335 11.1 2.446 7.255 2.446 7.255 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.372 6.097 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.556 5.788 arnoldi_extremal 4 6.8 0.000 0.000 5.061 5.076 arnoldi_normal_ev 4 7.8 0.001 0.005 5.061 5.075 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.777 5.010 calculate_norms 1584 9.8 4.413 4.850 4.413 4.850 build_subspace 16 8.4 0.014 0.020 4.758 4.764 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.427 3.763 mp_allgather_i34 111 8.7 0.819 3.711 0.819 3.711 ls_scf_post 1 4.0 0.000 0.001 3.686 3.691 dbcsr_matrix_vector_mult_local 304 10.0 3.024 3.595 3.026 3.597 ls_scf_store_result 1 5.0 0.000 0.000 3.411 3.512 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.871 2.983 dbcsr_data_new 4116 9.9 2.099 2.445 2.099 2.445 dbcsr_sort_data 325 11.1 1.890 2.182 1.890 2.182 make_images_sizes 222 9.7 0.000 0.000 1.017 2.123 mp_alltoall_i44 222 10.7 1.016 2.122 1.016 2.122 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.896 1.899 dbcsr_finalize 304 7.8 0.026 0.032 1.619 1.888 make_images_pack 222 9.7 1.625 1.882 1.628 1.885 rebuild_ks_matrix 3 7.3 0.000 0.001 1.874 1.877 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.005 1.874 1.877 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=91.702000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2203.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1aa3ad7d27726d25ce9c9edce445974c62615587_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 3.343684E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266673. MP_Allreduce 3138 13030. MP_Sync 4 MP_Alltoall 47 30278988. MP_SendRecv 69 86400. MP_ISendRecv 69 86400. MP_Wait 531 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.028 0.045 97.000 97.001 qs_energies 1 2.0 0.000 0.000 96.414 96.419 ls_scf 1 3.0 0.000 0.000 94.755 94.761 dbcsr_multiply_generic 111 6.7 0.016 0.017 78.565 78.815 ls_scf_main 1 4.0 0.000 0.000 59.026 59.027 multiply_cannon 111 7.7 0.049 0.098 51.590 56.634 density_matrix_trs4 2 5.0 0.002 0.003 52.853 52.990 multiply_cannon_loop 111 8.7 0.151 0.167 46.550 50.112 ls_scf_init_scf 1 4.0 0.000 0.000 32.491 32.494 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.299 31.361 mp_waitall_1 6369 11.0 22.704 29.899 22.704 29.899 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.853 28.865 multiply_cannon_multrec 1332 9.7 14.141 17.630 22.027 25.272 make_m2s 222 7.7 0.007 0.008 21.178 22.675 make_images 222 8.7 3.151 3.626 21.127 22.626 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.150 17.644 make_images_data 222 9.7 0.004 0.004 11.791 13.539 hybrid_alltoall_any 227 10.6 0.797 3.767 11.076 12.862 dbcsr_mm_accdrv_process 3641 10.4 0.315 0.486 7.529 9.038 dbcsr_mm_accdrv_process_sort 3641 11.4 7.085 8.546 7.085 8.546 mp_sum_l 887 5.1 4.215 7.464 4.215 7.464 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.091 6.043 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.594 5.996 mp_irecv_dv 3229 10.9 2.065 5.954 2.065 5.954 multiply_cannon_sync_h2d 1332 9.7 5.390 5.902 5.390 5.902 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.196 5.699 arnoldi_extremal 4 6.8 0.000 0.000 5.193 5.205 arnoldi_normal_ev 4 7.8 0.001 0.005 5.193 5.205 build_subspace 16 8.4 0.014 0.021 4.868 4.875 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.577 4.746 calculate_norms 2376 9.8 4.203 4.632 4.203 4.632 mp_allgather_i34 111 8.7 2.155 4.502 2.155 4.502 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.577 3.866 dbcsr_matrix_vector_mult_local 304 10.0 3.191 3.688 3.193 3.690 dbcsr_sort_data 658 11.4 3.011 3.379 3.011 3.379 ls_scf_post 1 4.0 0.000 0.000 3.238 3.242 dbcsr_special_finalize 555 9.7 0.006 0.007 2.750 3.223 dbcsr_merge_single_wm 555 10.7 0.529 0.661 2.741 3.214 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.032 3.083 ls_scf_store_result 1 5.0 0.000 0.000 2.981 3.040 dbcsr_data_release 10477 10.7 1.580 2.422 1.580 2.422 dbcsr_finalize 304 7.8 0.049 0.061 1.798 1.972 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.001000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2693.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1aa3ad7d27726d25ce9c9edce445974c62615587_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 4.724277E+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 0.035 0.054 92.122 92.123 qs_energies 1 2.0 0.000 0.000 91.376 91.379 ls_scf 1 3.0 0.000 0.000 89.353 89.355 dbcsr_multiply_generic 111 6.7 0.017 0.018 71.015 71.240 ls_scf_main 1 4.0 0.000 0.000 56.807 56.807 multiply_cannon 111 7.7 0.105 0.174 53.093 56.589 multiply_cannon_loop 111 8.7 0.088 0.094 50.507 51.705 density_matrix_trs4 2 5.0 0.002 0.003 49.808 49.874 ls_scf_init_scf 1 4.0 0.000 0.000 29.165 29.168 mp_waitall_1 5436 11.0 24.861 28.819 24.861 28.819 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.919 27.953 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.828 25.841 multiply_cannon_multrec 444 9.7 13.611 16.663 20.675 22.210 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.984 15.945 make_m2s 222 7.7 0.005 0.005 13.390 14.326 make_images 222 8.7 2.039 2.487 13.323 14.259 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.537 13.187 make_images_data 222 9.7 0.003 0.004 8.161 9.646 hybrid_alltoall_any 227 10.6 0.803 3.809 8.075 9.559 multiply_cannon_sync_h2d 444 9.7 6.804 8.029 6.804 8.029 dbcsr_mm_accdrv_process 3003 10.4 0.355 0.432 6.759 7.886 dbcsr_mm_accdrv_process_sort 3003 11.4 6.391 7.488 6.391 7.488 arnoldi_extremal 4 6.8 0.000 0.000 5.774 5.785 arnoldi_normal_ev 4 7.8 0.002 0.005 5.774 5.785 build_subspace 16 8.4 0.015 0.020 5.375 5.383 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.428 4.678 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.155 4.354 dbcsr_matrix_vector_mult_local 304 10.0 3.697 4.165 3.699 4.168 mp_sum_l 887 5.1 2.517 3.906 2.517 3.906 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.553 3.878 mp_irecv_dv 1241 11.2 1.537 3.847 1.537 3.847 mp_allgather_i34 111 8.7 1.149 3.704 1.149 3.704 calculate_norms 792 9.8 3.618 3.694 3.618 3.694 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.546 3.628 ls_scf_post 1 4.0 0.000 0.000 3.380 3.383 ls_scf_store_result 1 5.0 0.000 0.000 3.172 3.205 make_images_sizes 222 9.7 0.000 0.000 0.846 3.108 mp_alltoall_i44 222 10.7 0.846 3.108 0.846 3.108 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.763 2.922 dbcsr_data_new 4608 9.7 1.779 2.305 1.779 2.305 dbcsr_finalize 304 7.8 0.062 0.077 2.204 2.290 dbcsr_merge_all 275 8.9 0.482 0.527 2.063 2.153 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.041 2.042 rebuild_ks_matrix 3 7.3 0.000 0.000 2.008 2.009 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.008 2.009 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.007 2.007 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=92.123000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3730.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1aa3ad7d27726d25ce9c9edce445974c62615587_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 8.867426E+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 0.075 0.095 106.110 106.110 qs_energies 1 2.0 0.000 0.000 104.784 104.791 ls_scf 1 3.0 0.000 0.000 101.815 101.821 dbcsr_multiply_generic 111 6.7 0.024 0.028 75.498 75.625 ls_scf_main 1 4.0 0.000 0.000 63.402 63.403 density_matrix_trs4 2 5.0 0.002 0.003 54.382 54.444 multiply_cannon 111 7.7 0.110 0.165 49.153 51.129 multiply_cannon_loop 111 8.7 0.098 0.101 46.194 46.909 ls_scf_init_scf 1 4.0 0.000 0.000 34.621 34.622 ls_scf_init_matrix_S 1 5.0 0.000 0.000 33.058 33.085 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 30.350 30.356 mp_waitall_1 4527 11.1 21.701 25.303 21.701 25.303 make_m2s 222 7.7 0.005 0.005 22.651 23.726 make_images 222 8.7 3.575 3.879 22.544 23.618 multiply_cannon_multrec 444 9.7 17.837 18.430 22.481 23.055 hybrid_alltoall_any 227 10.6 1.654 3.611 12.789 15.791 make_images_data 222 9.7 0.003 0.004 13.049 15.173 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.047 11.529 multiply_cannon_sync_h2d 444 9.7 8.789 8.821 8.789 8.821 arnoldi_extremal 4 6.8 0.000 0.000 7.345 7.352 arnoldi_normal_ev 4 7.8 0.003 0.009 7.345 7.352 build_subspace 16 8.4 0.026 0.036 6.795 6.806 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.436 5.587 dbcsr_matrix_vector_mult_local 304 10.0 5.030 5.334 5.032 5.337 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.165 5.260 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.860 5.139 dbcsr_mm_accdrv_process 1814 10.4 0.274 0.385 4.462 4.622 dbcsr_mm_accdrv_process_sort 1814 11.4 4.108 4.237 4.108 4.237 ls_scf_post 1 4.0 0.000 0.000 3.791 3.799 make_images_sizes 222 9.7 0.000 0.000 1.467 3.610 mp_alltoall_i44 222 10.7 1.467 3.609 1.467 3.609 ls_scf_store_result 1 5.0 0.000 0.000 3.511 3.550 mp_allgather_i34 111 8.7 1.082 3.473 1.082 3.473 calculate_norms 792 9.8 3.231 3.271 3.231 3.271 dbcsr_finalize 304 7.8 0.082 0.089 3.071 3.165 dbcsr_merge_all 275 8.9 0.884 0.916 2.856 2.945 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.939 2.939 dbcsr_complete_redistribute 5 7.6 1.439 1.506 2.746 2.876 matrix_ls_to_qs 2 6.0 0.000 0.000 2.402 2.516 dbcsr_sort_data 325 11.1 2.437 2.498 2.437 2.498 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.461 2.463 rebuild_ks_matrix 3 7.3 0.000 0.000 2.394 2.396 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.394 2.396 dbcsr_data_new 6591 9.6 1.895 2.355 1.895 2.355 dbcsr_new_transposed 4 7.5 0.242 0.253 2.268 2.280 dbcsr_frobenius_norm 74 6.6 2.056 2.127 2.200 2.238 dbcsr_add_d 103 6.2 0.000 0.000 2.124 2.199 dbcsr_add_anytype 103 7.2 0.858 0.890 2.123 2.198 dbcsr_data_release 12724 10.6 1.976 2.195 1.976 2.195 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=106.110000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=7018.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1aa3ad7d27726d25ce9c9edce445974c62615587_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 580.390912E+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 3723339. 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 0.019 0.066 224.276 224.280 qs_mol_dyn_low 1 2.0 0.004 0.014 223.160 223.183 qs_forces 5 3.8 0.004 0.005 223.077 223.081 qs_energies 5 4.8 0.003 0.005 220.210 220.233 scf_env_do_scf 5 5.8 0.000 0.001 206.311 206.315 scf_env_do_scf_inner_loop 105 6.6 0.002 0.006 180.243 180.246 qs_scf_new_mos 105 7.6 0.000 0.001 141.469 141.658 qs_scf_loop_do_ot 105 8.6 0.001 0.001 141.469 141.658 dbcsr_multiply_generic 1445 12.2 0.129 0.138 132.681 133.460 ot_scf_mini 105 9.6 0.003 0.003 131.579 131.731 multiply_cannon 1445 13.2 0.273 0.285 113.882 116.114 multiply_cannon_loop 1445 14.2 2.840 2.983 112.264 114.662 velocity_verlet 4 3.0 0.002 0.020 105.650 105.663 ot_mini 105 10.6 0.001 0.001 60.193 60.291 multiply_cannon_multrec 69360 15.2 29.766 34.781 39.655 44.671 mp_waitall_1 488190 16.1 34.871 42.464 34.871 42.464 qs_ot_get_p 112 10.4 0.001 0.001 41.404 41.736 qs_ot_get_derivative 55 11.6 0.001 0.001 38.542 38.658 multiply_cannon_sync_h2d 69360 15.2 29.108 34.184 29.108 34.184 multiply_cannon_metrocomm3 69360 15.2 0.198 0.210 25.900 33.644 qs_ot_p2m_diag 40 11.0 0.020 0.030 30.399 30.541 rebuild_ks_matrix 110 8.4 0.000 0.000 28.280 28.435 qs_ks_build_kohn_sham_matrix 110 9.4 0.011 0.013 28.280 28.435 cp_dbcsr_syevd 40 12.0 0.002 0.002 27.227 27.227 qs_ks_update_qs_env 112 7.6 0.001 0.001 25.984 26.131 init_scf_loop 7 6.6 0.000 0.000 26.039 26.040 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 22.955 23.243 apply_single 62 13.6 0.000 0.000 22.955 23.243 cp_fm_syevd 40 13.0 0.000 0.001 22.220 22.350 prepare_preconditioner 7 7.6 0.000 0.000 21.311 21.342 make_preconditioner 7 8.6 0.000 0.000 21.311 21.342 ot_new_cg_direction 55 11.6 0.001 0.001 20.951 20.951 cp_fm_redistribute_end 40 14.0 8.783 17.527 8.787 17.528 cp_fm_syevd_base 40 14.0 8.734 17.481 8.734 17.481 qs_rho_update_rho_low 110 7.6 0.000 0.001 15.945 16.280 calculate_rho_elec 110 8.6 0.029 0.032 15.944 16.280 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.858 15.034 make_full_inverse_cholesky 7 9.6 0.000 0.000 14.396 14.465 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 13.960 14.061 mp_sum_l 4764 12.2 12.273 13.195 12.273 13.195 calculate_dm_sparse 110 9.5 0.000 0.000 11.496 11.665 pw_transfer 1645 12.4 0.079 0.097 11.262 11.424 init_scf_run 5 5.8 0.000 0.000 11.312 11.313 scf_env_initial_rho_setup 5 6.8 0.002 0.003 11.312 11.313 fft_wrap_pw1pw2 1425 13.5 0.012 0.014 11.123 11.292 density_rs2pw 110 9.6 0.005 0.006 10.398 10.796 dbcsr_mm_accdrv_process 154766 15.8 6.216 6.438 9.757 10.575 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 10.270 10.343 qs_vxc_create 110 10.4 0.002 0.003 10.188 10.219 cp_fm_cholesky_invert 7 10.6 10.160 10.168 10.160 10.168 fft_wrap_pw1pw2_240 915 15.0 0.837 0.921 9.778 9.966 multiply_cannon_metrocomm1 69360 15.2 0.095 0.101 4.912 9.656 check_diag 80 13.5 8.684 8.934 9.378 9.508 sum_up_and_integrate 60 10.3 0.001 0.002 8.157 8.167 integrate_v_rspace 60 11.3 0.001 0.002 8.140 8.151 fft3d_pb 915 16.0 2.395 2.591 7.955 8.117 transfer_rs2pw 445 10.6 0.007 0.008 7.227 7.654 acc_transpose_blocks 69360 15.2 0.349 0.369 6.764 7.568 xc_rho_set_and_dset_create 110 12.4 0.075 0.095 7.218 7.459 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.001 6.949 6.965 xc_vxc_pw_create 60 11.3 0.038 0.049 6.800 6.831 make_full_single_inverse 7 9.6 0.001 0.001 6.651 6.683 make_m2s 2890 13.2 0.078 0.088 6.046 6.630 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 6.528 6.608 calculate_first_density_matrix 1 7.0 0.000 0.001 6.544 6.566 make_images 2890 14.2 0.237 0.258 5.938 6.522 xc_pw_derive 510 13.4 0.005 0.006 5.996 6.065 mp_alltoall_z22v 2340 17.7 4.943 5.228 4.943 5.228 mp_waitany 7680 13.5 4.304 4.775 4.304 4.775 acc_transpose_blocks_kernels 69360 16.2 0.850 0.888 4.042 4.740 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=224.280000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=552.800000, yerr=1.939072 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 1aa3ad7d27726d25ce9c9edce445974c62615587 Summary: empty Status: OK