=== 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: 26e59941a7734f05bfa3a24a29949956890fc20f ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, # Cray-FFTW 3.3.8.10, COSMA 2.6.6, ELPA 2023.05.001, # HDF5 1.14.2, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.2.2, LIBVORI 220621, LIBXSMM 1.17, # PLUMED 2.9.0, SPGLIB 1.16.2, LIBGRPP 20231215 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Last update: 25.01.2024 # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-hdf5 --with-libvdwxc --with-pexsi --with-plumed --with-sirius=no; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2023.05.001 USE_HDF5 := 1.14.2 USE_LIBGRPP := 20231215 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.9.0 #USE_QUIP := 0.9.10 #USE_DEEPMD := 2.2.7 #USE_SIRIUS := 7.5.2 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS #LIBVDWXC_VER := 0.4.0 #SPFFT_VER := 1.0.6 #SPLA_VER := 1.5.5 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_DEEPMD),) USE_DEEPMD := $(strip $(USE_DEEPMD)) DEEPMD_INC := $(INSTALL_PATH)/libdeepmd_c-$(USE_DEEPMD)/include DEEPMD_LIB := $(INSTALL_PATH)/libdeepmd_c-$(USE_DEEPMD)/lib CFLAGS += -I$(DEEPMD_INC) DFLAGS += -D__DEEPMD LIBS += $(DEEPMD_LIB)/libdeepmd.so LIBS += $(DEEPMD_LIB)/libdeepmd_c.so LIBS += $(DEEPMD_LIB)/libdeepmd_cc.so LIBS += $(DEEPMD_LIB)/libdeepmd_dyn_cudart.so LIBS += $(DEEPMD_LIB)/libdeepmd_op.so LIBS += $(DEEPMD_LIB)/libdeepmd_op_cuda.so LIBS += $(DEEPMD_LIB)/libtensorflow_cc.so.2 LIBS += $(DEEPMD_LIB)/libtensorflow_framework.so.2 endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBGRPP),) USE_LIBGRPP := $(strip $(USE_LIBGRPP)) LIBGRPP_INC := $(INSTALL_PATH)/libgrpp-main-$(USE_LIBGRPP)/include LIBGRPP_LIB := $(INSTALL_PATH)/libgrpp-main-$(USE_LIBGRPP)/lib CFLAGS += -I$(LIBGRPP_INC) DFLAGS += -D__LIBGRPP LIBS += $(LIBGRPP_LIB)/liblibgrpp.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a endif ifneq ($(USE_HDF5),) USE_HDF5 := $(strip $(USE_HDF5)) HDF5_INC := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/include HDF5_LIB := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/lib CFLAGS += -I$(HDF5_INC) DFLAGS += -D__HDF5 LIBS += $(HDF5_LIB)/libhdf5_fortran.a LIBS += $(HDF5_LIB)/libhdf5_hl.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/26e59941a7734f05bfa3a24a29949956890fc20f_performance_tests/01 job id: 51615451 --- 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/26e59941a7734f05bfa3a24a29949956890fc20f_performance_tests/02 job id: 51615452 --- Point --- name: 20 plot: h2o_32_ri_rpa_mp2 regex: Total MP2 Time= label: RI-MP2 (8n/6r/2t) --- Point --- name: 21 plot: h2o_32_ri_rpa_mp2_mem regex: Estimated peak process memory label: RI-MP2 (8n/6r/2t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/26e59941a7734f05bfa3a24a29949956890fc20f_performance_tests/03 job id: 51615453 --- Point --- name: 100 plot: h2o_64_md regex: CP2K label: (8n/12r/1t) --- Point --- name: 101 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/12r/1t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 6 # threads/rank = 2 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/26e59941a7734f05bfa3a24a29949956890fc20f_performance_tests/04 job id: 51615454 --- Point --- name: 102 plot: h2o_64_md regex: CP2K label: (8n/6r/2t) --- Point --- name: 103 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/6r/2t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 4 # threads/rank = 3 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/26e59941a7734f05bfa3a24a29949956890fc20f_performance_tests/05 job id: 51615455 --- Point --- name: 104 plot: h2o_64_md regex: CP2K label: (8n/4r/3t) --- Point --- name: 105 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/4r/3t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 3 # threads/rank = 4 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/26e59941a7734f05bfa3a24a29949956890fc20f_performance_tests/06 job id: 51615456 --- Point --- name: 106 plot: h2o_64_md regex: CP2K label: (8n/3r/4t) --- Point --- name: 107 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/3r/4t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/26e59941a7734f05bfa3a24a29949956890fc20f_performance_tests/07 job id: 51615458 --- Point --- name: 108 plot: h2o_64_md regex: CP2K label: (8n/2r/6t) --- Point --- name: 109 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/2r/6t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 1 # threads/rank = 12 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/26e59941a7734f05bfa3a24a29949956890fc20f_performance_tests/08 job id: 51615459 --- 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/26e59941a7734f05bfa3a24a29949956890fc20f_performance_tests/09 job id: 51615460 --- 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/26e59941a7734f05bfa3a24a29949956890fc20f_performance_tests/10 job id: 51615461 --- 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/26e59941a7734f05bfa3a24a29949956890fc20f_performance_tests/11 job id: 51615462 --- 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/26e59941a7734f05bfa3a24a29949956890fc20f_performance_tests/12 job id: 51615463 --- 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/26e59941a7734f05bfa3a24a29949956890fc20f_performance_tests/13 job id: 51615464 --- 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/26e59941a7734f05bfa3a24a29949956890fc20f_performance_tests/14 job id: 51615465 --- 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/26e59941a7734f05bfa3a24a29949956890fc20f_performance_tests/15 job id: 51615466 --- 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/26e59941a7734f05bfa3a24a29949956890fc20f_performance_tests/16 job id: 51615467 --- 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/26e59941a7734f05bfa3a24a29949956890fc20f_performance_tests/17 job id: 51615468 --- 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/26e59941a7734f05bfa3a24a29949956890fc20f_performance_tests/18 job id: 51615469 --- 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/26e59941a7734f05bfa3a24a29949956890fc20f_performance_tests/19 job id: 51615470 --- 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/26e59941a7734f05bfa3a24a29949956890fc20f_performance_tests/20 job id: 51615472 --- Point --- name: 410 plot: h2o_256_md regex: CP2K label: (8n/1r/12t) --- Point --- name: 411 plot: h2o_256_md_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/26e59941a7734f05bfa3a24a29949956890fc20f_performance_tests/21 job id: 51615474 --- Point --- name: 500 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/12r/1t) --- Point --- name: 501 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/12r/1t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 6 # threads/rank = 2 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/26e59941a7734f05bfa3a24a29949956890fc20f_performance_tests/22 job id: 51615475 --- Point --- name: 502 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/6r/2t) --- Point --- name: 503 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/6r/2t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 4 # threads/rank = 3 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/26e59941a7734f05bfa3a24a29949956890fc20f_performance_tests/23 job id: 51615476 --- Point --- name: 504 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/4r/3t) --- Point --- name: 505 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/4r/3t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 3 # threads/rank = 4 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/26e59941a7734f05bfa3a24a29949956890fc20f_performance_tests/24 job id: 51615477 --- Point --- name: 506 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/3r/4t) --- Point --- name: 507 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/3r/4t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/26e59941a7734f05bfa3a24a29949956890fc20f_performance_tests/25 job id: 51615478 --- Point --- name: 508 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/2r/6t) --- Point --- name: 509 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/2r/6t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 1 # threads/rank = 12 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/26e59941a7734f05bfa3a24a29949956890fc20f_performance_tests/26 job id: 51615479 --- Point --- name: 510 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/1r/12t) --- Point --- name: 511 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: 512 H2O (4 NVE MD steps on 64 nodes) input file: benchmarks/QS/00512_H2O/H2O-512_md.inp required files: [] output file: result.log # nodes = 64 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 20 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/26e59941a7734f05bfa3a24a29949956890fc20f_performance_tests/27 job id: 51615480 --- 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/26e59941a7734f05bfa3a24a29949956890fc20f_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 0.068 0.202 133.931 133.932 farming_run 1 2.0 132.917 132.918 133.692 133.764 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.482211E+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 0.025 114.248 114.249 qs_energies 1 2.0 0.001 0.001 114.055 114.059 mp2_main 1 3.0 0.000 0.000 111.990 111.994 mp2_gpw_main 1 4.0 0.027 0.034 111.011 111.016 mp2_ri_gpw_compute_in 1 5.0 0.197 0.566 92.132 92.480 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 54.547 54.896 mp2_eri_3c_integrate_gpw 272 7.0 0.152 0.165 40.959 46.037 get_2c_integrals 1 6.0 0.008 0.009 37.079 37.412 integrate_v_rspace 273 8.0 0.438 0.454 24.593 29.535 pw_transfer 6555 10.6 0.382 0.401 26.660 27.150 grid_integrate_task_list 273 9.0 20.513 25.915 20.513 25.915 fft_wrap_pw1pw2 5465 11.4 0.044 0.047 25.281 25.865 fft_wrap_pw1pw2_100 2178 12.4 0.028 0.029 22.811 23.386 rpa_ri_compute_en 1 5.0 0.020 0.041 18.774 18.951 compute_2c_integrals 1 7.0 0.002 0.002 18.904 18.904 compute_2c_integrals_loop_lm 1 8.0 0.002 0.004 18.494 18.631 mp2_eri_2c_integrate_gpw 1 9.0 2.383 2.439 18.492 18.630 cp_fm_cholesky_decompose 12 8.2 18.105 18.445 18.105 18.445 cholesky_decomp 1 7.0 0.000 0.000 17.023 17.357 fft3d_s 5443 13.4 16.053 16.394 16.075 16.416 ao_to_mo_and_store_B_mult_1 272 7.0 10.766 15.338 10.766 15.338 calculate_wavefunction 272 8.0 5.392 5.553 12.282 12.987 rpa_num_int 1 6.0 0.002 0.032 10.617 10.627 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.583 10.592 calc_mat_Q 8 8.0 0.000 0.000 9.454 9.542 contract_S_to_Q 8 9.0 0.000 0.000 8.878 8.962 calc_potential_gpw 544 9.5 0.004 0.005 8.202 8.661 parallel_gemm_fm 14 9.1 0.000 0.000 8.458 8.571 parallel_gemm_fm_cosma 14 10.1 8.458 8.571 8.458 8.571 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 8.019 8.250 potential_pw2rs 545 10.0 0.106 0.108 7.439 8.117 create_integ_mat 1 6.0 0.021 0.035 7.703 7.712 collocate_single_gaussian 272 10.0 0.039 0.042 7.288 7.523 array2fm 1 7.0 0.000 0.000 6.557 7.153 pw_scatter_s 2720 13.7 4.330 4.440 4.330 4.440 pw_gather_s 2722 13.2 3.380 3.761 3.380 3.761 array2fm_buffer_send 1 8.0 2.903 3.104 2.903 3.104 pw_poisson_solve 545 10.5 1.123 1.186 2.203 2.381 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.015693, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2798.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/26e59941a7734f05bfa3a24a29949956890fc20f_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 0.042 0.045 523.952 523.955 farming_run 1 2.0 522.974 522.983 523.892 523.902 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827128576 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788821 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.228067E+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 0.033 319.753 319.755 qs_energies 1 2.0 0.000 0.000 319.546 319.548 mp2_main 1 3.0 0.000 0.000 203.777 203.778 mp2_gpw_main 1 4.0 0.048 0.070 202.912 202.919 mp2_ri_gpw_compute_en 1 5.0 0.038 0.041 128.683 130.805 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.776 1.924 126.992 126.997 scf_env_do_scf 1 3.0 0.000 0.000 115.475 115.475 qs_ks_update_qs_env 5 5.0 0.000 0.000 114.584 114.590 rebuild_ks_matrix 4 6.0 0.000 0.000 114.583 114.589 qs_ks_build_kohn_sham_matrix 4 7.0 0.054 0.060 114.583 114.589 hfx_ks_matrix 4 8.0 0.001 0.001 114.237 114.243 integrate_four_center 4 9.0 0.153 0.456 114.236 114.242 mp2_ri_gpw_compute_en_expansio 172 7.0 0.469 0.508 111.261 112.093 local_gemm 172 8.0 110.793 111.615 110.793 111.615 integrate_four_center_main 4 10.0 0.131 0.465 101.921 106.161 integrate_four_center_bin 263 11.0 101.790 105.974 101.790 105.974 init_scf_loop 1 4.0 0.000 0.000 97.723 97.723 mp2_ri_gpw_compute_in 1 5.0 0.064 0.064 74.082 75.142 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 53.834 54.893 mp2_eri_3c_integrate_gpw 91 7.0 0.143 0.157 41.304 46.517 integrate_v_rspace 95 8.0 0.397 0.572 27.773 32.824 pw_transfer 2240 10.6 0.144 0.167 29.551 29.969 fft_wrap_pw1pw2 1868 11.4 0.017 0.020 28.549 28.954 ao_to_mo_and_store_B_mult_1 91 7.0 10.856 28.941 10.856 28.941 grid_integrate_task_list 95 9.0 23.193 28.445 23.193 28.445 fft_wrap_pw1pw2_100 730 12.4 0.013 0.014 26.288 26.705 get_2c_integrals 1 6.0 0.000 0.000 20.151 20.185 compute_2c_integrals 1 7.0 0.002 0.003 19.136 19.141 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 18.716 19.014 mp2_eri_2c_integrate_gpw 1 9.0 1.729 1.827 18.715 19.013 fft3d_s 1823 13.4 18.493 18.780 18.507 18.793 scf_env_do_scf_inner_loop 4 4.0 0.000 0.001 17.750 17.750 calculate_wavefunction 91 8.0 2.016 2.046 9.621 9.848 mp2_ri_gpw_compute_en_comm 22 7.0 0.495 0.517 7.865 9.178 potential_pw2rs 186 10.0 0.033 0.034 8.450 9.025 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.212 8.594 mp_sync 37 10.5 4.293 8.418 4.293 8.418 calc_potential_gpw 182 9.5 0.002 0.002 7.932 8.247 collocate_single_gaussian 91 10.0 0.017 0.023 7.785 8.019 mp_sendrecv_dm3 2068 8.0 5.909 7.250 5.909 7.250 integrate_four_center_load 4 10.0 0.000 0.000 6.770 6.774 hfx_load_balance 1 11.0 0.000 0.000 6.770 6.774 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=202.915842, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1488.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/26e59941a7734f05bfa3a24a29949956890fc20f_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 444.231680E+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 0.025 0.039 56.581 56.583 qs_mol_dyn_low 1 2.0 0.009 0.035 56.305 56.314 qs_forces 11 3.9 0.002 0.003 56.082 56.085 qs_energies 11 4.9 0.002 0.003 54.577 54.597 scf_env_do_scf 11 5.9 0.000 0.001 48.130 48.131 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 46.063 46.064 qs_scf_new_mos 108 7.5 0.000 0.000 35.834 36.149 qs_scf_loop_do_ot 108 8.5 0.000 0.001 35.833 36.149 dbcsr_multiply_generic 2286 12.5 0.094 0.098 35.187 35.669 ot_scf_mini 108 9.5 0.002 0.002 34.106 34.300 multiply_cannon 2286 13.5 0.187 0.194 26.869 28.527 multiply_cannon_loop 2286 14.5 1.805 1.899 26.158 27.825 velocity_verlet 10 3.0 0.012 0.048 26.469 26.476 ot_mini 108 10.5 0.001 0.001 20.764 21.015 qs_ot_get_derivative 108 11.5 0.001 0.001 17.708 17.904 mp_waitall_1 245248 16.5 9.021 15.554 9.021 15.554 multiply_cannon_metrocomm3 54864 15.5 0.072 0.077 6.356 13.289 multiply_cannon_multrec 54864 15.5 3.639 5.641 7.797 11.139 qs_ot_get_p 119 10.4 0.001 0.001 8.555 8.851 rebuild_ks_matrix 119 8.3 0.000 0.000 7.937 8.095 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 7.937 8.095 mp_sum_l 7287 12.8 6.208 7.994 6.208 7.994 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.978 7.122 multiply_cannon_sync_h2d 54864 15.5 5.132 6.761 5.132 6.761 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 6.338 6.477 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.723 6.166 dbcsr_mm_accdrv_process 76910 16.1 1.821 2.863 4.073 5.744 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.575 5.619 init_scf_run 11 5.9 0.000 0.001 5.134 5.134 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.134 5.134 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.785 4.786 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.552 4.552 cp_fm_redistribute_end 50 14.0 2.317 4.517 2.326 4.521 sum_up_and_integrate 119 10.3 0.001 0.002 4.511 4.518 integrate_v_rspace 119 11.3 0.002 0.003 4.500 4.509 cp_fm_diag_elpa_base 50 14.0 2.188 4.396 2.194 4.406 qs_rho_update_rho_low 119 7.7 0.000 0.000 4.237 4.355 calculate_rho_elec 119 8.7 0.011 0.016 4.236 4.355 multiply_cannon_metrocomm1 54864 15.5 0.055 0.060 1.941 3.480 calculate_dm_sparse 119 9.5 0.000 0.000 3.166 3.320 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.927 3.127 apply_single 119 13.6 0.000 0.000 2.927 3.127 calculate_first_density_matrix 1 7.0 0.000 0.001 2.851 2.855 jit_kernel_multiply 13 15.8 2.190 2.791 2.190 2.791 acc_transpose_blocks 54864 15.5 0.220 0.244 2.220 2.744 ot_diis_step 108 11.5 0.006 0.006 2.739 2.740 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.541 2.594 density_rs2pw 119 9.7 0.004 0.004 2.218 2.333 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.286 2.288 wfi_extrapolate 11 7.9 0.001 0.001 2.212 2.212 grid_integrate_task_list 119 12.3 2.000 2.099 2.000 2.099 init_scf_loop 11 6.9 0.000 0.000 2.045 2.045 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.978 2.029 mp_sum_d 4137 12.0 1.459 2.026 1.459 2.026 potential_pw2rs 119 12.3 0.004 0.004 1.878 1.900 pw_transfer 1439 11.6 0.051 0.057 1.786 1.864 make_m2s 4572 13.5 0.053 0.056 1.799 1.847 fft_wrap_pw1pw2 1201 12.6 0.006 0.007 1.710 1.789 make_images 4572 14.5 0.132 0.138 1.718 1.766 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.605 1.624 mp_alltoall_d11v 2130 13.8 1.421 1.583 1.421 1.583 fft3d_ps 1201 14.6 0.370 0.474 1.423 1.496 transfer_rs2pw 487 10.6 0.005 0.006 1.370 1.466 mp_waitany 12084 13.8 1.245 1.438 1.245 1.438 grid_collocate_task_list 119 9.7 1.350 1.417 1.350 1.417 acc_transpose_blocks_sync 164592 16.5 1.187 1.416 1.187 1.416 fft_wrap_pw1pw2_140 487 13.2 0.108 0.114 1.320 1.399 transfer_pw2rs 487 13.2 0.006 0.006 1.351 1.369 dbcsr_dot_sd 1205 11.9 0.048 0.059 0.796 1.189 parallel_gemm_fm 81 9.0 0.000 0.000 1.160 1.164 parallel_gemm_fm_cosma 81 10.0 1.160 1.164 1.160 1.164 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=56.583000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=423.272727, yerr=1.052349 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/26e59941a7734f05bfa3a24a29949956890fc20f_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 480.235520E+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 0.008 0.030 39.510 39.511 qs_mol_dyn_low 1 2.0 0.004 0.015 39.295 39.339 qs_forces 11 3.9 0.002 0.003 38.747 38.749 qs_energies 11 4.9 0.009 0.020 37.081 37.087 scf_env_do_scf 11 5.9 0.000 0.001 31.807 31.807 scf_env_do_scf_inner_loop 108 6.5 0.043 0.089 29.147 29.147 dbcsr_multiply_generic 2286 12.5 0.101 0.104 22.190 22.594 qs_scf_new_mos 108 7.5 0.001 0.001 20.551 20.816 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.551 20.815 ot_scf_mini 108 9.5 0.002 0.003 19.629 19.830 velocity_verlet 10 3.0 0.053 0.418 18.644 18.656 multiply_cannon 2286 13.5 0.207 0.216 17.029 18.390 multiply_cannon_loop 2286 14.5 1.179 1.241 15.874 17.342 ot_mini 108 10.5 0.001 0.001 12.135 12.402 mp_waitall_1 200699 16.5 5.561 10.739 5.561 10.739 qs_ot_get_derivative 108 11.5 0.001 0.001 9.760 9.962 multiply_cannon_metrocomm3 27432 15.5 0.072 0.074 4.028 9.416 multiply_cannon_multrec 27432 15.5 1.831 4.177 6.184 9.042 rebuild_ks_matrix 119 8.3 0.000 0.000 6.722 6.869 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.722 6.868 dbcsr_mm_accdrv_process 47894 16.0 3.565 5.897 4.273 6.447 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.910 6.042 qs_ot_get_p 119 10.4 0.001 0.001 4.861 5.083 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.662 4.540 mp_sum_l 7287 12.8 2.130 4.187 2.130 4.187 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.083 4.053 apply_single 119 13.6 0.000 0.000 3.082 4.053 init_scf_run 11 5.9 0.000 0.001 4.036 4.036 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.035 4.036 sum_up_and_integrate 119 10.3 0.001 0.001 3.664 3.671 integrate_v_rspace 119 11.3 0.002 0.003 3.650 3.657 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.580 3.602 calculate_rho_elec 119 8.7 0.021 0.024 3.579 3.601 qs_ot_p2m_diag 50 11.0 0.009 0.013 2.949 2.968 make_m2s 4572 13.5 0.051 0.053 2.577 2.917 make_images 4572 14.5 0.205 0.244 2.487 2.831 init_scf_loop 11 6.9 0.000 0.000 2.640 2.640 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.556 2.556 calculate_first_density_matrix 1 7.0 0.000 0.000 2.531 2.534 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.302 2.389 ot_diis_step 108 11.5 0.011 0.011 2.328 2.329 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.244 2.245 cp_fm_redistribute_end 50 14.0 1.137 2.220 1.140 2.222 multiply_cannon_sync_h2d 27432 15.5 1.702 2.222 1.702 2.222 calculate_dm_sparse 119 9.5 0.000 0.001 2.133 2.209 cp_fm_diag_elpa_base 50 14.0 1.050 2.129 1.078 2.171 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.035 2.037 acc_transpose_blocks 27432 15.5 0.113 0.119 1.685 2.020 density_rs2pw 119 9.7 0.004 0.004 1.873 1.942 grid_integrate_task_list 119 12.3 1.840 1.923 1.840 1.923 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.840 1.887 pw_transfer 1439 11.6 0.065 0.069 1.828 1.861 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.737 1.772 jit_kernel_multiply 8 16.4 0.652 1.610 0.652 1.610 make_images_data 4572 15.5 0.047 0.054 1.221 1.544 prepare_preconditioner 11 7.9 0.000 0.000 1.497 1.522 make_preconditioner 11 8.9 0.000 0.000 1.497 1.522 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.398 1.453 wfi_extrapolate 11 7.9 0.001 0.001 1.446 1.447 hybrid_alltoall_any 4725 16.4 0.053 0.115 1.075 1.423 potential_pw2rs 119 12.3 0.006 0.006 1.416 1.422 fft_wrap_pw1pw2_140 487 13.2 0.127 0.131 1.384 1.419 fft3d_ps 1201 14.6 0.522 0.580 1.368 1.396 grid_collocate_task_list 119 9.7 1.288 1.345 1.288 1.345 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.319 1.328 mp_alltoall_d11v 2130 13.8 1.183 1.314 1.183 1.314 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.194 1.242 mp_sum_d 4137 12.0 0.669 1.201 0.669 1.201 transfer_rs2pw 487 10.6 0.005 0.005 1.056 1.126 mp_allgather_i34 2286 14.5 0.581 0.968 0.581 0.968 acc_transpose_blocks_kernels 27432 16.5 0.189 0.278 0.737 0.947 qs_energies_init_hamiltonians 11 5.9 0.000 0.002 0.939 0.941 acc_transpose_blocks_sync 82296 16.5 0.809 0.930 0.809 0.930 transfer_pw2rs 487 13.2 0.004 0.005 0.882 0.889 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.868 0.884 arnoldi_extremal 119 11.4 0.002 0.002 0.700 0.824 arnoldi_normal_ev 119 12.4 0.139 0.277 0.699 0.822 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=39.511000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=456.727273, yerr=1.863082 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/26e59941a7734f05bfa3a24a29949956890fc20f_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 511.295488E+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 0.017 0.047 33.005 33.006 qs_mol_dyn_low 1 2.0 0.003 0.003 32.649 32.665 qs_forces 11 3.9 0.003 0.003 32.572 32.573 qs_energies 11 4.9 0.002 0.003 31.036 31.039 scf_env_do_scf 11 5.9 0.000 0.001 25.379 25.380 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 22.982 22.983 dbcsr_multiply_generic 2286 12.5 0.096 0.099 17.674 17.742 qs_scf_new_mos 108 7.5 0.001 0.001 15.348 15.367 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.347 15.367 multiply_cannon 2286 13.5 0.194 0.198 14.429 15.166 velocity_verlet 10 3.0 0.004 0.006 15.067 15.069 ot_scf_mini 108 9.5 0.002 0.002 14.590 14.603 multiply_cannon_loop 2286 14.5 0.855 0.898 13.659 14.418 ot_mini 108 10.5 0.001 0.001 9.025 9.038 multiply_cannon_multrec 18288 15.5 1.867 2.845 7.338 7.913 qs_ot_get_derivative 108 11.5 0.001 0.001 7.535 7.548 dbcsr_mm_accdrv_process 38222 16.0 5.320 6.000 5.378 6.064 rebuild_ks_matrix 119 8.3 0.000 0.000 5.960 5.975 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.960 5.975 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.246 5.259 init_scf_run 11 5.9 0.000 0.001 4.402 4.402 scf_env_initial_rho_setup 11 6.9 0.018 0.022 4.402 4.402 mp_waitall_1 158411 16.6 2.894 3.873 2.894 3.873 sum_up_and_integrate 119 10.3 0.001 0.001 3.526 3.531 integrate_v_rspace 119 11.3 0.003 0.003 3.513 3.520 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.841 3.445 qs_ot_get_p 119 10.4 0.001 0.001 3.324 3.343 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.243 3.248 calculate_rho_elec 119 8.7 0.031 0.031 3.243 3.248 calculate_first_density_matrix 1 7.0 0.000 0.000 3.165 3.184 calculate_dm_sparse 119 9.5 0.000 0.001 2.448 2.459 init_scf_loop 11 6.9 0.000 0.000 2.380 2.382 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.012 2.349 apply_single 119 13.6 0.000 0.000 2.012 2.349 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.220 2.227 multiply_cannon_metrocomm3 18288 15.5 0.049 0.050 1.518 2.205 make_m2s 4572 13.5 0.044 0.045 1.809 1.957 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.925 1.925 grid_integrate_task_list 119 12.3 1.813 1.920 1.813 1.920 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.900 1.901 make_images 4572 14.5 0.191 0.204 1.724 1.870 density_rs2pw 119 9.7 0.004 0.004 1.811 1.867 pw_transfer 1439 11.6 0.065 0.068 1.840 1.848 acc_transpose_blocks 18288 15.5 0.078 0.081 1.718 1.790 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.785 1.788 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.747 1.758 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.743 1.749 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.695 1.704 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.682 1.692 cp_fm_diag_elpa_base 50 14.0 1.658 1.673 1.680 1.690 prepare_preconditioner 11 7.9 0.000 0.000 1.622 1.625 make_preconditioner 11 8.9 0.000 0.000 1.622 1.625 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.483 1.565 ot_diis_step 108 11.5 0.011 0.011 1.472 1.472 mp_sum_l 7287 12.8 1.050 1.428 1.050 1.428 fft_wrap_pw1pw2_140 487 13.2 0.179 0.183 1.413 1.423 potential_pw2rs 119 12.3 0.007 0.008 1.326 1.329 fft3d_ps 1201 14.6 0.546 0.565 1.302 1.311 grid_collocate_task_list 119 9.7 1.234 1.300 1.234 1.300 multiply_cannon_sync_h2d 18288 15.5 1.058 1.216 1.058 1.216 wfi_extrapolate 11 7.9 0.001 0.001 1.168 1.168 transfer_rs2pw 487 10.6 0.005 0.005 0.971 1.046 multiply_cannon_metrocomm1 18288 15.5 0.030 0.032 0.649 1.029 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.940 0.957 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.950 0.952 make_images_data 4572 15.5 0.047 0.051 0.772 0.932 acc_transpose_blocks_kernels 18288 16.5 0.219 0.225 0.888 0.904 hybrid_alltoall_any 4725 16.4 0.058 0.116 0.669 0.861 transfer_pw2rs 487 13.2 0.004 0.004 0.804 0.807 acc_transpose_blocks_sync 54864 16.5 0.732 0.803 0.732 0.803 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.785 0.787 mp_alltoall_d11v 2130 13.8 0.646 0.760 0.646 0.760 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.669 0.729 jit_kernel_transpose 5 15.6 0.669 0.684 0.669 0.684 cp_fm_cholesky_invert 11 10.9 0.679 0.683 0.679 0.683 mp_alltoall_z22v 1201 16.6 0.608 0.672 0.608 0.672 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=33.006000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=487.454545, yerr=0.497930 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/26e59941a7734f05bfa3a24a29949956890fc20f_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 552.206336E+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 0.068 0.120 36.777 36.778 qs_mol_dyn_low 1 2.0 0.003 0.003 36.335 36.352 qs_forces 11 3.9 0.002 0.002 36.070 36.070 qs_energies 11 4.9 0.002 0.002 34.366 34.373 scf_env_do_scf 11 5.9 0.000 0.001 29.327 29.328 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 25.970 25.971 dbcsr_multiply_generic 2286 12.5 0.102 0.105 20.052 20.167 velocity_verlet 10 3.0 0.002 0.002 18.531 18.542 qs_scf_new_mos 108 7.5 0.001 0.001 17.962 18.017 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.962 18.016 ot_scf_mini 108 9.5 0.002 0.003 16.932 16.979 multiply_cannon 2286 13.5 0.219 0.226 16.414 16.840 multiply_cannon_loop 2286 14.5 1.538 1.615 15.470 15.775 ot_mini 108 10.5 0.001 0.001 10.514 10.573 multiply_cannon_multrec 27432 15.5 2.649 3.421 9.284 9.590 qs_ot_get_derivative 108 11.5 0.001 0.001 8.646 8.693 dbcsr_mm_accdrv_process 47916 15.9 6.191 7.926 6.532 8.089 rebuild_ks_matrix 119 8.3 0.000 0.000 6.272 6.327 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 6.272 6.327 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.556 5.604 init_scf_run 11 5.9 0.000 0.001 3.700 3.701 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.700 3.701 qs_ot_get_p 119 10.4 0.001 0.001 3.574 3.642 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.106 3.494 sum_up_and_integrate 119 10.3 0.001 0.001 3.448 3.456 integrate_v_rspace 119 11.3 0.003 0.003 3.437 3.445 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.333 3.365 calculate_rho_elec 119 8.7 0.040 0.046 3.332 3.365 init_scf_loop 11 6.9 0.000 0.000 3.338 3.338 acc_transpose_blocks 27432 15.5 0.118 0.120 2.460 2.554 prepare_preconditioner 11 7.9 0.000 0.000 2.512 2.520 make_preconditioner 11 8.9 0.000 0.000 2.512 2.520 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.121 2.445 make_m2s 4572 13.5 0.054 0.056 2.241 2.351 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.319 2.342 calculate_first_density_matrix 1 7.0 0.000 0.000 2.320 2.322 make_images 4572 14.5 0.272 0.333 2.133 2.241 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.106 2.233 apply_single 119 13.6 0.000 0.000 2.106 2.233 mp_waitall_1 137007 16.6 1.708 2.217 1.708 2.217 calculate_dm_sparse 119 9.5 0.000 0.000 2.134 2.187 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.162 2.173 pw_transfer 1439 11.6 0.065 0.069 1.961 2.000 grid_integrate_task_list 119 12.3 1.809 1.913 1.809 1.913 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.869 1.911 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.907 1.908 density_rs2pw 119 9.7 0.004 0.004 1.776 1.896 ot_diis_step 108 11.5 0.012 0.012 1.827 1.828 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.763 1.764 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.706 1.720 fft_wrap_pw1pw2_140 487 13.2 0.227 0.234 1.556 1.598 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.528 1.537 cp_fm_diag_elpa_base 50 14.0 1.494 1.511 1.526 1.535 acc_transpose_blocks_sync 82296 16.5 1.429 1.519 1.429 1.519 fft3d_ps 1201 14.6 0.592 0.653 1.357 1.385 grid_collocate_task_list 119 9.7 1.247 1.339 1.247 1.339 wfi_extrapolate 11 7.9 0.001 0.001 1.328 1.328 multiply_cannon_metrocomm3 27432 15.5 0.040 0.041 0.741 1.313 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.289 1.304 mp_sum_l 7287 12.8 1.033 1.276 1.033 1.276 potential_pw2rs 119 12.3 0.008 0.009 1.259 1.261 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.230 1.241 cp_fm_upper_to_full 72 14.2 0.816 1.160 0.816 1.160 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.084 1.084 dbcsr_complete_redistribute 329 12.2 0.134 0.182 0.803 1.074 mp_alltoall_d11v 2130 13.8 0.756 0.986 0.756 0.986 transfer_rs2pw 487 10.6 0.004 0.005 0.867 0.981 make_images_data 4572 15.5 0.048 0.053 0.830 0.943 acc_transpose_blocks_kernels 27432 16.5 0.274 0.283 0.883 0.907 jit_kernel_multiply 5 16.8 0.277 0.901 0.277 0.901 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.801 0.879 hybrid_alltoall_any 4725 16.4 0.066 0.156 0.716 0.877 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.590 0.855 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.804 0.810 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=36.778000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=524.363636, yerr=3.391774 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/26e59941a7734f05bfa3a24a29949956890fc20f_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 617.799680E+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 0.028 0.043 30.828 30.829 qs_mol_dyn_low 1 2.0 0.003 0.003 30.571 30.580 qs_forces 11 3.9 0.002 0.002 30.485 30.488 qs_energies 11 4.9 0.002 0.002 28.720 28.724 scf_env_do_scf 11 5.9 0.000 0.001 23.741 23.741 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 20.919 20.919 velocity_verlet 10 3.0 0.002 0.002 15.783 15.787 dbcsr_multiply_generic 2286 12.5 0.095 0.098 13.749 13.846 qs_scf_new_mos 108 7.5 0.001 0.001 12.873 12.893 qs_scf_loop_do_ot 108 8.5 0.001 0.001 12.872 12.893 ot_scf_mini 108 9.5 0.002 0.002 12.122 12.149 multiply_cannon 2286 13.5 0.225 0.229 10.798 11.335 multiply_cannon_loop 2286 14.5 0.643 0.664 9.825 10.047 ot_mini 108 10.5 0.001 0.001 6.824 6.856 multiply_cannon_multrec 9144 15.5 1.674 1.968 5.991 6.278 rebuild_ks_matrix 119 8.3 0.000 0.000 6.032 6.052 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 6.031 6.051 qs_ot_get_derivative 108 11.5 0.001 0.001 5.488 5.516 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.364 5.381 dbcsr_mm_accdrv_process 12550 15.8 3.671 4.231 4.209 4.283 sum_up_and_integrate 119 10.3 0.001 0.001 3.627 3.632 integrate_v_rspace 119 11.3 0.003 0.003 3.617 3.622 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.583 3.591 calculate_rho_elec 119 8.7 0.060 0.061 3.582 3.591 init_scf_run 11 5.9 0.000 0.001 3.487 3.487 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.486 3.487 qs_ot_get_p 119 10.4 0.001 0.001 3.337 3.372 init_scf_loop 11 6.9 0.000 0.000 2.793 2.794 pw_transfer 1439 11.6 0.065 0.068 2.285 2.297 mp_waitall_1 115863 16.7 1.726 2.255 1.726 2.255 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.192 2.205 qs_ot_p2m_diag 50 11.0 0.022 0.023 2.201 2.202 calculate_first_density_matrix 1 7.0 0.000 0.000 2.197 2.198 make_m2s 4572 13.5 0.034 0.035 1.969 2.145 density_rs2pw 119 9.7 0.003 0.003 1.968 2.071 make_images 4572 14.5 0.268 0.299 1.879 2.055 prepare_preconditioner 11 7.9 0.000 0.000 2.002 2.007 make_preconditioner 11 8.9 0.000 0.000 2.002 2.007 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.944 1.945 grid_integrate_task_list 119 12.3 1.863 1.935 1.863 1.935 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.881 1.911 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.825 1.848 calculate_dm_sparse 119 9.5 0.000 0.000 1.806 1.823 fft_wrap_pw1pw2_140 487 13.2 0.325 0.333 1.764 1.780 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.698 1.699 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.609 1.617 cp_fm_diag_elpa_base 50 14.0 1.580 1.598 1.606 1.615 acc_transpose_blocks 9144 15.5 0.042 0.044 1.529 1.562 fft3d_ps 1201 14.6 0.653 0.665 1.544 1.551 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.499 1.509 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.470 1.482 potential_pw2rs 119 12.3 0.010 0.011 1.409 1.414 jit_kernel_multiply 8 16.2 0.499 1.399 0.499 1.399 grid_collocate_task_list 119 9.7 1.300 1.377 1.300 1.377 ot_diis_step 108 11.5 0.013 0.013 1.318 1.319 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.246 1.283 apply_single 119 13.6 0.000 0.000 1.246 1.282 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.243 1.245 wfi_extrapolate 11 7.9 0.001 0.001 1.226 1.226 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.160 1.166 hybrid_alltoall_any 4725 16.4 0.065 0.177 0.840 1.078 cp_fm_cholesky_invert 11 10.9 1.060 1.064 1.060 1.064 make_images_data 4572 15.5 0.042 0.046 0.841 1.053 mp_alltoall_d11v 2130 13.8 0.899 1.024 0.899 1.024 transfer_rs2pw 487 10.6 0.004 0.005 0.893 0.996 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.876 0.933 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.878 0.884 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 0.434 0.846 mp_alltoall_z22v 1201 16.6 0.764 0.799 0.764 0.799 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.791 0.794 acc_transpose_blocks_sync 27432 16.5 0.721 0.753 0.721 0.753 acc_transpose_blocks_kernels 9144 16.5 0.118 0.120 0.750 0.752 transfer_pw2rs 487 13.2 0.003 0.004 0.744 0.749 mp_allgather_i34 2286 14.5 0.289 0.737 0.289 0.737 qs_env_update_s_mstruct 11 6.9 0.000 0.001 0.687 0.736 mp_sum_l 7287 12.8 0.528 0.679 0.528 0.679 dbcsr_complete_redistribute 329 12.2 0.159 0.168 0.619 0.648 jit_kernel_transpose 5 15.6 0.631 0.633 0.631 0.633 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=30.829000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=588.818182, yerr=3.737237 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/26e59941a7734f05bfa3a24a29949956890fc20f_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 756.834304E+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 0.019 0.048 43.166 43.166 qs_mol_dyn_low 1 2.0 0.003 0.003 42.812 42.819 qs_forces 11 3.9 0.002 0.002 42.754 42.755 qs_energies 11 4.9 0.002 0.002 40.746 40.749 scf_env_do_scf 11 5.9 0.001 0.001 34.605 34.605 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 26.812 26.813 velocity_verlet 10 3.0 0.002 0.002 24.068 24.073 dbcsr_multiply_generic 2286 12.5 0.103 0.106 19.231 19.359 qs_scf_new_mos 108 7.5 0.001 0.001 17.146 17.248 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.145 17.248 multiply_cannon 2286 13.5 0.300 0.306 15.278 16.176 ot_scf_mini 108 9.5 0.002 0.002 15.991 16.091 multiply_cannon_loop 2286 14.5 0.868 0.889 14.037 14.943 ot_mini 108 10.5 0.001 0.001 9.714 9.832 multiply_cannon_multrec 9144 15.5 3.421 4.755 8.890 9.018 init_scf_loop 11 6.9 0.000 0.000 7.766 7.768 qs_ot_get_derivative 108 11.5 0.001 0.001 7.660 7.761 rebuild_ks_matrix 119 8.3 0.000 0.000 7.259 7.400 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 7.258 7.400 prepare_preconditioner 11 7.9 0.000 0.000 6.809 6.822 make_preconditioner 11 8.9 0.000 0.000 6.809 6.822 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.358 6.702 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.540 6.668 dbcsr_mm_accdrv_process 12550 15.8 4.542 6.509 5.336 6.593 cp_fm_upper_to_full 72 14.2 3.218 4.613 3.218 4.613 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.184 4.190 calculate_rho_elec 119 8.7 0.117 0.120 4.183 4.189 init_scf_run 11 5.9 0.000 0.001 4.081 4.081 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.081 4.081 sum_up_and_integrate 119 10.3 0.001 0.001 3.841 3.846 integrate_v_rspace 119 11.3 0.004 0.004 3.830 3.835 qs_ot_get_p 119 10.4 0.001 0.001 3.349 3.487 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.764 3.212 mp_waitall_1 94719 16.7 2.265 3.192 2.265 3.192 pw_transfer 1439 11.6 0.068 0.069 3.008 3.013 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.910 2.915 dbcsr_complete_redistribute 329 12.2 0.285 0.289 1.977 2.798 calculate_first_density_matrix 1 7.0 0.000 0.000 2.618 2.620 make_m2s 4572 13.5 0.037 0.037 2.375 2.518 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.678 2.494 fft_wrap_pw1pw2_140 487 13.2 0.650 0.652 2.487 2.494 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.251 2.478 apply_single 119 13.6 0.000 0.000 2.251 2.478 calculate_dm_sparse 119 9.5 0.000 0.000 2.391 2.407 make_images 4572 14.5 0.351 0.384 2.256 2.398 mp_alltoall_i22 627 13.8 1.521 2.350 1.521 2.350 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.445 2.254 density_rs2pw 119 9.7 0.003 0.003 2.220 2.231 multiply_cannon_metrocomm3 9144 15.5 0.021 0.021 1.276 2.175 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.152 2.153 grid_integrate_task_list 119 12.3 2.083 2.120 2.083 2.120 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.055 2.105 acc_transpose_blocks 9144 15.5 0.044 0.045 2.017 2.054 ot_diis_step 108 11.5 0.014 0.014 2.029 2.030 qs_ot_p2m_diag 50 11.0 0.043 0.043 1.970 1.972 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.806 1.871 fft3d_ps 1201 14.6 0.867 0.885 1.831 1.834 qs_energies_init_hamiltonians 11 5.9 0.000 0.000 1.778 1.779 mp_sum_l 7287 12.8 1.030 1.695 1.030 1.695 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.618 1.618 grid_collocate_task_list 119 9.7 1.524 1.548 1.524 1.548 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.402 1.408 cp_fm_cholesky_invert 11 10.9 1.402 1.405 1.402 1.405 wfi_extrapolate 11 7.9 0.001 0.001 1.399 1.399 potential_pw2rs 119 12.3 0.014 0.014 1.349 1.350 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.340 1.340 cp_fm_diag_elpa_base 50 14.0 1.196 1.246 1.338 1.339 hybrid_alltoall_any 4725 16.4 0.091 0.150 1.112 1.333 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.304 1.324 make_images_data 4572 15.5 0.046 0.050 1.053 1.231 mp_alltoall_d11v 2130 13.8 1.137 1.178 1.137 1.178 acc_transpose_blocks_sync 27432 16.5 1.114 1.146 1.114 1.146 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.106 1.124 jit_kernel_multiply 5 15.4 0.765 1.104 0.765 1.104 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.995 1.035 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.929 0.942 qs_create_task_list 11 7.9 0.000 0.000 0.925 0.937 generate_qs_task_list 11 8.9 0.369 0.389 0.924 0.937 acc_transpose_blocks_kernels 9144 16.5 0.121 0.122 0.842 0.902 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=43.166000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=714.090909, yerr=11.712111 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/26e59941a7734f05bfa3a24a29949956890fc20f_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 491.974656E+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 0.045 0.133 82.083 82.086 qs_mol_dyn_low 1 2.0 0.011 0.058 81.545 81.604 qs_forces 11 3.9 0.003 0.005 81.366 81.369 qs_energies 11 4.9 0.022 0.027 78.516 78.538 scf_env_do_scf 11 5.9 0.000 0.001 69.322 69.324 scf_env_do_scf_inner_loop 99 6.5 0.002 0.008 63.962 63.962 dbcsr_multiply_generic 2055 12.4 0.104 0.108 51.014 51.331 qs_scf_new_mos 99 7.5 0.000 0.001 47.116 47.265 qs_scf_loop_do_ot 99 8.5 0.001 0.001 47.115 47.264 ot_scf_mini 99 9.5 0.002 0.002 44.744 44.848 multiply_cannon 2055 13.4 0.182 0.189 42.643 43.723 velocity_verlet 10 3.0 0.023 0.064 43.024 43.025 multiply_cannon_loop 2055 14.4 1.767 1.809 41.717 42.809 ot_mini 99 10.5 0.001 0.001 26.711 26.793 qs_ot_get_derivative 99 11.5 0.001 0.001 19.928 20.023 multiply_cannon_multrec 49320 15.4 11.425 12.101 17.445 18.216 rebuild_ks_matrix 110 8.3 0.000 0.000 14.345 14.507 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.345 14.507 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.584 12.731 mp_waitall_1 220248 16.4 10.803 11.745 10.803 11.745 multiply_cannon_sync_h2d 49320 15.4 9.636 10.165 9.636 10.165 qs_ot_get_p 110 10.4 0.001 0.001 9.602 9.706 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.778 8.334 multiply_cannon_metrocomm3 49320 15.4 0.081 0.084 6.655 7.936 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.234 7.731 apply_single 110 13.6 0.000 0.000 7.234 7.731 init_scf_run 11 5.9 0.000 0.001 7.096 7.096 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.096 7.096 sum_up_and_integrate 110 10.3 0.002 0.003 6.885 6.896 integrate_v_rspace 110 11.3 0.003 0.003 6.859 6.877 ot_diis_step 99 11.5 0.005 0.006 6.592 6.592 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.397 6.550 calculate_rho_elec 110 8.6 0.044 0.214 6.396 6.549 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.433 6.468 dbcsr_mm_accdrv_process 87628 16.1 3.028 3.118 5.889 6.152 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.578 5.578 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.348 5.391 init_scf_loop 11 6.9 0.000 0.000 5.334 5.334 mp_sum_l 6594 12.7 4.204 5.104 4.204 5.104 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.019 5.044 cp_fm_diag_elpa_base 48 14.0 5.007 5.032 5.018 5.042 wfi_extrapolate 11 7.9 0.001 0.001 3.981 3.981 make_m2s 4110 13.4 0.060 0.064 3.837 3.939 make_images 4110 14.4 0.177 0.189 3.742 3.850 calculate_dm_sparse 110 9.5 0.001 0.001 3.729 3.832 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.657 3.661 multiply_cannon_metrocomm1 49320 15.4 0.064 0.066 2.387 3.484 density_rs2pw 110 9.6 0.004 0.004 3.247 3.457 grid_integrate_task_list 110 12.3 3.243 3.426 3.243 3.426 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.244 3.303 prepare_preconditioner 11 7.9 0.000 0.000 3.187 3.205 make_preconditioner 11 8.9 0.000 0.000 3.187 3.205 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.146 3.197 pw_transfer 1331 11.6 0.055 0.065 2.995 3.072 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.972 3.016 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 2.907 2.987 calculate_first_density_matrix 1 7.0 0.000 0.002 2.832 2.834 mp_alltoall_d11v 2046 13.8 2.182 2.677 2.182 2.677 jit_kernel_multiply 13 15.9 2.586 2.606 2.586 2.606 fft_wrap_pw1pw2_140 451 13.1 0.309 0.332 2.502 2.588 acc_transpose_blocks 49320 15.4 0.205 0.212 2.529 2.580 potential_pw2rs 110 12.3 0.006 0.006 2.503 2.518 grid_collocate_task_list 110 9.6 2.159 2.284 2.159 2.284 fft3d_ps 1111 14.6 0.797 0.885 2.200 2.253 transfer_rs2pw 451 10.6 0.005 0.006 1.846 2.060 mp_waitany 14300 13.8 1.817 2.002 1.817 2.002 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.890 1.908 make_images_data 4110 15.4 0.043 0.047 1.739 1.877 mp_sum_d 3891 11.9 1.321 1.844 1.321 1.844 hybrid_alltoall_any 4261 16.3 0.083 0.482 1.498 1.761 cp_fm_cholesky_invert 11 10.9 1.736 1.740 1.736 1.740 transfer_pw2rs 451 13.1 0.006 0.007 1.692 1.701 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.636 1.660 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=82.086000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=467.636364, yerr=2.307014 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/26e59941a7734f05bfa3a24a29949956890fc20f_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 585.064448E+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 3280327. 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 0.042 0.071 68.852 68.853 qs_mol_dyn_low 1 2.0 0.003 0.003 68.578 68.608 qs_forces 11 3.9 0.003 0.003 68.513 68.514 qs_energies 11 4.9 0.003 0.004 65.198 65.201 scf_env_do_scf 11 5.9 0.000 0.001 56.114 56.117 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 48.684 48.685 dbcsr_multiply_generic 2055 12.4 0.115 0.119 38.063 38.257 velocity_verlet 10 3.0 0.001 0.002 35.741 35.742 qs_scf_new_mos 99 7.5 0.001 0.001 32.964 33.143 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.963 33.143 multiply_cannon 2055 13.4 0.224 0.245 31.560 32.619 ot_scf_mini 99 9.5 0.003 0.003 31.299 31.483 multiply_cannon_loop 2055 14.4 1.157 1.183 30.283 31.025 ot_mini 99 10.5 0.001 0.001 18.309 18.500 multiply_cannon_multrec 24660 15.4 6.964 8.794 14.290 15.800 rebuild_ks_matrix 110 8.3 0.000 0.001 13.447 13.550 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 13.447 13.550 qs_ot_get_derivative 99 11.5 0.001 0.001 12.497 12.684 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.819 11.909 mp_waitall_1 176588 16.5 7.457 10.129 7.457 10.129 multiply_cannon_metrocomm3 24660 15.4 0.074 0.076 5.125 8.071 multiply_cannon_sync_h2d 24660 15.4 6.353 7.505 6.353 7.505 dbcsr_mm_accdrv_process 52282 16.1 5.777 6.780 7.155 7.501 init_scf_loop 11 6.9 0.000 0.000 7.396 7.396 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.479 7.143 apply_single 110 13.6 0.000 0.001 6.478 7.143 init_scf_run 11 5.9 0.000 0.001 6.719 6.720 scf_env_initial_rho_setup 11 6.9 0.001 0.002 6.719 6.720 qs_ot_get_p 110 10.4 0.001 0.001 6.349 6.558 sum_up_and_integrate 110 10.3 0.001 0.003 6.313 6.324 integrate_v_rspace 110 11.3 0.002 0.003 6.286 6.299 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.830 5.840 calculate_rho_elec 110 8.6 0.039 0.047 5.830 5.839 ot_diis_step 99 11.5 0.010 0.010 5.764 5.764 prepare_preconditioner 11 7.9 0.000 0.000 5.404 5.428 make_preconditioner 11 8.9 0.000 0.000 5.404 5.428 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.702 5.408 make_full_inverse_cholesky 11 9.9 0.000 0.000 4.980 5.131 make_m2s 4110 13.4 0.057 0.061 4.153 4.610 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.485 4.506 make_images 4110 14.4 0.405 0.464 4.041 4.498 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.026 4.026 pw_transfer 1331 11.6 0.066 0.071 3.509 3.646 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.402 3.543 wfi_extrapolate 11 7.9 0.001 0.001 3.492 3.492 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.468 3.483 cp_fm_diag_elpa_base 48 14.0 3.422 3.440 3.465 3.480 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.264 3.367 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.326 3.328 grid_integrate_task_list 110 12.3 3.161 3.287 3.161 3.287 density_rs2pw 110 9.6 0.004 0.004 3.085 3.239 calculate_dm_sparse 110 9.5 0.001 0.001 3.186 3.218 calculate_first_density_matrix 1 7.0 0.000 0.000 3.133 3.135 fft_wrap_pw1pw2_140 451 13.1 0.359 0.378 2.928 3.067 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.973 3.030 make_images_data 4110 15.4 0.049 0.053 2.264 2.784 hybrid_alltoall_any 4261 16.3 0.104 0.454 1.981 2.727 fft3d_ps 1111 14.6 1.116 1.344 2.493 2.645 mp_sum_l 6594 12.7 1.726 2.535 1.726 2.535 cp_fm_cholesky_invert 11 10.9 2.524 2.531 2.524 2.531 grid_collocate_task_list 110 9.6 2.166 2.343 2.166 2.343 potential_pw2rs 110 12.3 0.008 0.009 2.196 2.211 acc_transpose_blocks 24660 15.4 0.115 0.118 1.992 2.054 jit_kernel_multiply 9 16.3 1.015 1.988 1.015 1.988 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.969 1.988 mp_alltoall_d11v 2046 13.8 1.739 1.890 1.739 1.890 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.858 1.859 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.783 1.798 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.756 1.766 multiply_cannon_metrocomm4 22605 15.4 0.074 0.077 0.782 1.681 mp_allgather_i34 2055 14.4 0.585 1.595 0.585 1.595 transfer_rs2pw 451 10.6 0.006 0.007 1.400 1.547 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.388 1.490 mp_irecv_dv 57340 16.2 0.655 1.461 0.655 1.461 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=68.853000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=552.727273, yerr=7.324813 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/26e59941a7734f05bfa3a24a29949956890fc20f_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 659.681280E+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 0.033 0.058 61.818 61.819 qs_mol_dyn_low 1 2.0 0.008 0.040 61.462 61.472 qs_forces 11 3.9 0.003 0.003 61.359 61.362 qs_energies 11 4.9 0.002 0.008 58.180 58.184 scf_env_do_scf 11 5.9 0.001 0.001 49.512 49.513 scf_env_do_scf_inner_loop 99 6.5 0.003 0.006 40.822 40.822 velocity_verlet 10 3.0 0.003 0.007 32.832 32.834 dbcsr_multiply_generic 2055 12.4 0.109 0.113 30.080 30.349 qs_scf_new_mos 99 7.5 0.001 0.001 26.136 26.231 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.135 26.230 ot_scf_mini 99 9.5 0.002 0.003 24.883 25.008 multiply_cannon 2055 13.4 0.213 0.222 23.615 24.760 multiply_cannon_loop 2055 14.4 0.813 0.857 22.512 23.544 ot_mini 99 10.5 0.001 0.001 14.291 14.419 rebuild_ks_matrix 110 8.3 0.000 0.000 12.154 12.294 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.153 12.294 multiply_cannon_multrec 16440 15.4 3.552 4.347 10.904 11.627 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.696 10.819 mp_waitall_1 139946 16.5 7.103 10.119 7.103 10.119 qs_ot_get_derivative 99 11.5 0.001 0.001 9.822 9.949 init_scf_loop 11 6.9 0.001 0.004 8.654 8.655 multiply_cannon_metrocomm3 16440 15.4 0.046 0.047 4.509 7.386 dbcsr_mm_accdrv_process 34862 16.1 6.543 6.904 7.197 7.374 prepare_preconditioner 11 7.9 0.000 0.000 6.909 6.924 make_preconditioner 11 8.9 0.000 0.001 6.909 6.924 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.240 6.584 sum_up_and_integrate 110 10.3 0.001 0.002 6.286 6.301 integrate_v_rspace 110 11.3 0.003 0.003 6.260 6.275 init_scf_run 11 5.9 0.000 0.001 6.245 6.245 scf_env_initial_rho_setup 11 6.9 0.001 0.002 6.244 6.245 qs_ot_get_p 110 10.4 0.001 0.001 5.731 5.887 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.777 5.788 calculate_rho_elec 110 8.6 0.059 0.059 5.776 5.788 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.990 5.432 apply_single 110 13.6 0.000 0.000 4.989 5.431 make_m2s 4110 13.4 0.049 0.051 4.240 4.531 ot_diis_step 99 11.5 0.010 0.011 4.437 4.437 make_images 4110 14.4 0.397 0.520 4.123 4.413 qs_ot_p2m_diag 48 11.0 0.042 0.044 4.024 4.029 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.217 3.890 multiply_cannon_sync_h2d 16440 15.4 3.196 3.721 3.196 3.721 pw_transfer 1331 11.6 0.065 0.072 3.641 3.650 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.642 3.643 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.534 3.547 grid_integrate_task_list 110 12.3 3.202 3.407 3.202 3.407 calculate_dm_sparse 110 9.5 0.001 0.001 3.323 3.355 density_rs2pw 110 9.6 0.004 0.004 3.011 3.193 calculate_first_density_matrix 1 7.0 0.001 0.005 3.174 3.175 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.092 3.102 cp_fm_diag_elpa_base 48 14.0 3.026 3.051 3.091 3.100 fft_wrap_pw1pw2_140 451 13.1 0.460 0.470 3.033 3.049 wfi_extrapolate 11 7.9 0.001 0.001 2.978 2.978 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.771 2.771 cp_fm_cholesky_invert 11 10.9 2.755 2.761 2.755 2.761 make_images_data 4110 15.4 0.046 0.051 2.374 2.725 hybrid_alltoall_any 4261 16.3 0.107 0.377 2.114 2.711 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.499 2.570 multiply_cannon_metrocomm4 14385 15.4 0.047 0.051 0.923 2.514 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.436 2.491 fft3d_ps 1111 14.6 1.120 1.133 2.451 2.466 mp_irecv_dv 48980 15.7 0.847 2.377 0.847 2.377 grid_collocate_task_list 110 9.6 2.222 2.352 2.222 2.352 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 2.258 2.272 mp_sum_l 6594 12.7 1.622 2.223 1.622 2.223 potential_pw2rs 110 12.3 0.011 0.011 2.144 2.151 mp_alltoall_d11v 2046 13.8 1.822 2.145 1.822 2.145 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.975 1.978 dbcsr_complete_redistribute 325 12.2 0.322 0.347 1.430 1.883 cp_fm_upper_to_full 70 14.2 1.393 1.769 1.393 1.769 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.652 1.670 acc_transpose_blocks 16440 15.4 0.078 0.082 1.579 1.664 cp_fm_cholesky_decompose 22 10.9 1.597 1.615 1.597 1.615 mp_allgather_i34 2055 14.4 0.418 1.487 0.418 1.487 transfer_rs2pw 451 10.6 0.005 0.006 1.271 1.464 build_core_hamiltonian_matrix_ 11 4.9 0.003 0.020 1.372 1.450 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.972 1.416 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.306 1.314 mp_waitany 17072 13.8 1.125 1.306 1.125 1.306 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=61.819000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=623.272727, yerr=9.086362 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/26e59941a7734f05bfa3a24a29949956890fc20f_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 717.193216E+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 0.022 0.041 66.801 66.802 qs_mol_dyn_low 1 2.0 0.009 0.029 66.471 66.481 qs_forces 11 3.9 0.004 0.005 65.960 65.967 qs_energies 11 4.9 0.008 0.030 62.568 62.575 scf_env_do_scf 11 5.9 0.001 0.002 53.308 53.311 scf_env_do_scf_inner_loop 99 6.5 0.003 0.006 41.828 41.829 velocity_verlet 10 3.0 0.021 0.081 37.407 37.429 dbcsr_multiply_generic 2055 12.4 0.117 0.123 31.341 31.505 qs_scf_new_mos 99 7.5 0.001 0.001 27.236 27.347 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.236 27.347 multiply_cannon 2055 13.4 0.239 0.261 24.285 25.657 ot_scf_mini 99 9.5 0.003 0.003 25.551 25.653 multiply_cannon_loop 2055 14.4 1.421 1.480 22.848 23.337 multiply_cannon_multrec 24660 15.4 4.091 6.490 13.794 15.036 ot_mini 99 10.5 0.001 0.001 14.665 14.790 rebuild_ks_matrix 110 8.3 0.000 0.000 11.973 12.077 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.015 11.973 12.077 init_scf_loop 11 6.9 0.001 0.004 11.441 11.442 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.577 10.667 qs_ot_get_derivative 99 11.5 0.001 0.001 10.504 10.608 dbcsr_mm_accdrv_process 52304 16.0 8.463 10.164 9.543 10.458 prepare_preconditioner 11 7.9 0.000 0.000 9.702 9.716 make_preconditioner 11 8.9 0.000 0.001 9.702 9.716 make_full_inverse_cholesky 11 9.9 0.000 0.001 7.957 9.417 init_scf_run 11 5.9 0.000 0.001 6.327 6.327 scf_env_initial_rho_setup 11 6.9 0.001 0.002 6.326 6.327 sum_up_and_integrate 110 10.3 0.001 0.002 6.198 6.211 integrate_v_rspace 110 11.3 0.003 0.003 6.172 6.184 qs_ot_get_p 110 10.4 0.001 0.001 5.730 5.883 make_m2s 4110 13.4 0.059 0.061 5.451 5.856 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.822 5.831 calculate_rho_elec 110 8.6 0.078 0.082 5.822 5.831 mp_waitall_1 121746 16.5 4.110 5.821 4.110 5.821 make_images 4110 14.4 0.572 0.691 5.309 5.710 cp_fm_upper_to_full 70 14.2 3.359 4.861 3.359 4.861 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.046 4.141 apply_single 110 13.6 0.000 0.000 4.046 4.141 ot_diis_step 99 11.5 0.011 0.011 4.124 4.124 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.881 3.897 pw_transfer 1331 11.6 0.065 0.073 3.736 3.768 dbcsr_complete_redistribute 325 12.2 0.415 0.465 2.620 3.763 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.629 3.664 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.468 3.516 grid_integrate_task_list 110 12.3 3.271 3.462 3.271 3.462 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.328 3.386 calculate_dm_sparse 110 9.5 0.001 0.001 3.335 3.376 cp_dbcsr_syevd 48 12.0 0.003 0.005 3.323 3.324 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.255 3.258 calculate_first_density_matrix 1 7.0 0.001 0.003 3.248 3.251 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.118 3.241 fft_wrap_pw1pw2_140 451 13.1 0.531 0.545 3.134 3.170 hybrid_alltoall_any 4261 16.3 0.122 0.459 2.264 3.143 make_images_data 4110 15.4 0.049 0.053 2.720 3.120 density_rs2pw 110 9.6 0.004 0.004 2.938 3.049 multiply_cannon_metrocomm3 24660 15.4 0.038 0.039 1.295 3.038 wfi_extrapolate 11 7.9 0.001 0.001 2.981 2.982 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.002 2.873 2.911 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.736 2.845 mp_alltoall_i22 605 13.7 1.647 2.816 1.647 2.816 cp_fm_diag_elpa 48 13.0 0.000 0.001 2.800 2.811 cp_fm_diag_elpa_base 48 14.0 2.653 2.712 2.797 2.808 cp_fm_cholesky_invert 11 10.9 2.573 2.581 2.573 2.581 acc_transpose_blocks 24660 15.4 0.110 0.114 2.428 2.557 multiply_cannon_sync_h2d 24660 15.4 2.388 2.513 2.388 2.513 fft3d_ps 1111 14.6 1.152 1.196 2.455 2.473 qs_energies_init_hamiltonians 11 5.9 0.016 0.031 2.407 2.416 grid_collocate_task_list 110 9.6 2.262 2.393 2.262 2.393 potential_pw2rs 110 12.3 0.012 0.013 2.008 2.015 mp_alltoall_d11v 2046 13.8 1.757 1.988 1.757 1.988 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.002 1.890 1.905 jit_kernel_multiply 8 16.1 0.740 1.851 0.740 1.851 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.792 1.824 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.615 1.709 cp_fm_cholesky_decompose 22 10.9 1.609 1.652 1.609 1.652 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.635 1.649 mp_allgather_i34 2055 14.4 0.522 1.639 0.522 1.639 acc_transpose_blocks_sync 73980 16.4 1.405 1.520 1.405 1.520 mp_sum_l 6594 12.7 0.920 1.519 0.920 1.519 multiply_cannon_metrocomm4 20550 15.4 0.059 0.062 0.843 1.490 qs_env_update_s_mstruct 11 6.9 0.000 0.001 1.318 1.421 mp_irecv_dv 62702 16.1 0.739 1.408 0.739 1.408 transfer_rs2pw 451 10.6 0.005 0.006 1.237 1.375 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=66.802000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=681.363636, yerr=6.064393 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/26e59941a7734f05bfa3a24a29949956890fc20f_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 853.725184E+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 0.032 0.083 56.473 56.474 qs_mol_dyn_low 1 2.0 0.003 0.003 56.097 56.130 qs_forces 11 3.9 0.003 0.003 56.024 56.025 qs_energies 11 4.9 0.001 0.002 52.357 52.361 scf_env_do_scf 11 5.9 0.000 0.001 42.956 42.957 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 35.313 35.314 velocity_verlet 10 3.0 0.008 0.009 31.050 31.054 dbcsr_multiply_generic 2055 12.4 0.106 0.109 24.458 24.589 qs_scf_new_mos 99 7.5 0.001 0.001 20.715 20.765 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.715 20.764 multiply_cannon 2055 13.4 0.239 0.252 18.913 20.286 ot_scf_mini 99 9.5 0.002 0.002 19.464 19.494 multiply_cannon_loop 2055 14.4 0.602 0.620 17.624 17.935 rebuild_ks_matrix 110 8.3 0.000 0.000 11.656 11.679 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 11.656 11.679 ot_mini 99 10.5 0.001 0.001 10.750 10.776 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.371 10.390 multiply_cannon_multrec 8220 15.4 3.228 4.318 8.500 9.311 mp_waitall_1 103326 16.6 5.808 7.611 5.808 7.611 init_scf_loop 11 6.9 0.000 0.000 7.598 7.599 qs_ot_get_derivative 99 11.5 0.001 0.001 7.067 7.096 init_scf_run 11 5.9 0.000 0.001 6.282 6.283 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.282 6.282 sum_up_and_integrate 110 10.3 0.001 0.002 6.205 6.219 integrate_v_rspace 110 11.3 0.003 0.003 6.179 6.192 dbcsr_mm_accdrv_process 17442 15.9 3.521 4.611 5.132 6.103 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.000 6.012 calculate_rho_elec 110 8.6 0.112 0.113 6.000 6.012 prepare_preconditioner 11 7.9 0.000 0.000 5.955 5.959 make_preconditioner 11 8.9 0.000 0.000 5.955 5.959 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.545 5.620 qs_ot_get_p 110 10.4 0.001 0.001 4.753 4.772 make_m2s 4110 13.4 0.037 0.039 4.199 4.485 make_images 4110 14.4 0.644 0.706 4.069 4.357 multiply_cannon_metrocomm3 8220 15.4 0.018 0.019 2.965 4.317 pw_transfer 1331 11.6 0.066 0.071 4.063 4.074 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.956 3.969 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.719 3.768 apply_single 110 13.6 0.000 0.000 3.718 3.767 ot_diis_step 99 11.5 0.012 0.012 3.659 3.659 grid_integrate_task_list 110 12.3 3.385 3.506 3.385 3.506 calculate_first_density_matrix 1 7.0 0.000 0.000 3.498 3.499 fft_wrap_pw1pw2_140 451 13.1 0.718 0.733 3.435 3.452 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.386 3.390 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.309 3.311 density_rs2pw 110 9.6 0.004 0.004 3.002 3.150 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.050 3.050 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.927 2.941 calculate_dm_sparse 110 9.5 0.001 0.001 2.891 2.929 cp_fm_cholesky_invert 11 10.9 2.847 2.850 2.847 2.850 hybrid_alltoall_any 4261 16.3 0.198 0.844 2.280 2.730 wfi_extrapolate 11 7.9 0.001 0.001 2.688 2.688 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.668 2.669 make_images_data 4110 15.4 0.040 0.046 2.253 2.626 fft3d_ps 1111 14.6 1.290 1.298 2.500 2.512 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.496 2.504 cp_fm_diag_elpa_base 48 14.0 2.440 2.462 2.495 2.502 multiply_cannon_sync_h2d 8220 15.4 2.380 2.472 2.380 2.472 grid_collocate_task_list 110 9.6 2.367 2.463 2.367 2.463 jit_kernel_multiply 8 15.4 1.297 2.273 1.297 2.273 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.034 2.065 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.786 2.023 potential_pw2rs 110 12.3 0.015 0.015 2.009 2.011 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.900 1.918 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.842 1.853 mp_alltoall_d11v 2046 13.8 1.682 1.833 1.682 1.833 acc_transpose_blocks 8220 15.4 0.040 0.042 1.726 1.766 cp_fm_cholesky_decompose 22 10.9 1.657 1.676 1.657 1.676 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.613 1.616 qs_env_update_s_mstruct 11 6.9 0.001 0.003 1.495 1.612 mp_allgather_i34 2055 14.4 0.453 1.551 0.453 1.551 dbcsr_complete_redistribute 325 12.2 0.576 0.602 1.432 1.513 qs_create_task_list 11 7.9 0.001 0.002 1.210 1.309 generate_qs_task_list 11 8.9 0.374 0.442 1.209 1.308 transfer_rs2pw 451 10.6 0.005 0.005 1.097 1.260 mp_waitany 9240 13.8 1.048 1.230 1.048 1.230 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.189 1.212 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=56.474000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=805.818182, yerr=11.884430 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/26e59941a7734f05bfa3a24a29949956890fc20f_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 1.364128E+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 0.047 0.099 89.941 89.941 qs_mol_dyn_low 1 2.0 0.003 0.003 89.499 89.532 qs_forces 11 3.9 0.013 0.014 89.048 89.049 qs_energies 11 4.9 0.005 0.005 84.903 84.904 scf_env_do_scf 11 5.9 0.000 0.001 73.850 73.850 velocity_verlet 10 3.0 0.002 0.002 56.275 56.329 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 45.642 45.644 dbcsr_multiply_generic 2055 12.4 0.121 0.125 31.432 31.588 init_scf_loop 11 6.9 0.000 0.000 28.138 28.140 qs_scf_new_mos 99 7.5 0.001 0.001 27.723 27.818 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.722 27.817 prepare_preconditioner 11 7.9 0.000 0.000 26.116 26.130 make_preconditioner 11 8.9 0.000 0.000 26.116 26.130 ot_scf_mini 99 9.5 0.002 0.002 25.935 26.024 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.366 25.595 multiply_cannon 2055 13.4 0.332 0.359 24.134 24.597 multiply_cannon_loop 2055 14.4 0.828 0.849 22.348 23.005 cp_fm_upper_to_full 70 14.2 12.778 18.257 12.778 18.257 ot_mini 99 10.5 0.001 0.001 15.123 15.208 rebuild_ks_matrix 110 8.3 0.000 0.001 13.984 14.084 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 13.984 14.083 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.684 12.774 multiply_cannon_multrec 8220 15.4 4.538 4.835 11.041 11.511 dbcsr_complete_redistribute 325 12.2 1.007 1.036 7.292 10.568 qs_ot_get_derivative 99 11.5 0.001 0.001 10.412 10.501 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.282 9.548 mp_waitall_1 84994 16.7 7.709 8.992 7.709 8.992 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.734 8.964 mp_alltoall_i22 605 13.7 5.372 8.669 5.372 8.669 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.444 7.482 calculate_rho_elec 110 8.6 0.222 0.222 7.443 7.481 sum_up_and_integrate 110 10.3 0.002 0.002 6.930 6.945 integrate_v_rspace 110 11.3 0.004 0.004 6.902 6.917 init_scf_run 11 5.9 0.000 0.001 6.757 6.757 scf_env_initial_rho_setup 11 6.9 0.002 0.003 6.757 6.757 dbcsr_mm_accdrv_process 11614 15.7 3.919 4.123 6.354 6.558 pw_transfer 1331 11.6 0.075 0.075 5.789 5.797 make_m2s 4110 13.4 0.043 0.043 5.256 5.756 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 5.115 5.735 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 5.671 5.679 qs_ot_get_p 110 10.4 0.001 0.001 5.543 5.647 make_images 4110 14.4 0.878 0.932 5.068 5.566 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.805 5.327 apply_single 110 13.6 0.000 0.000 4.805 5.327 cp_fm_cholesky_invert 11 10.9 5.123 5.127 5.123 5.127 fft_wrap_pw1pw2_140 451 13.1 1.368 1.371 4.990 4.997 ot_diis_step 99 11.5 0.015 0.016 4.695 4.695 qs_ot_p2m_diag 48 11.0 0.151 0.156 3.931 3.937 density_rs2pw 110 9.6 0.004 0.004 3.912 3.928 grid_integrate_task_list 110 12.3 3.687 3.774 3.687 3.774 qs_energies_init_hamiltonians 11 5.9 0.008 0.014 3.690 3.691 calculate_dm_sparse 110 9.5 0.001 0.001 3.530 3.558 hybrid_alltoall_any 4261 16.3 0.263 0.561 2.764 3.515 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.495 3.499 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.005 3.477 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.419 3.419 fft3d_ps 1111 14.6 1.884 1.891 3.375 3.385 make_images_data 4110 15.4 0.045 0.048 2.760 3.374 wfi_extrapolate 11 7.9 0.001 0.001 3.339 3.339 calculate_first_density_matrix 1 7.0 0.000 0.000 3.300 3.301 multiply_cannon_sync_h2d 8220 15.4 3.117 3.151 3.117 3.151 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.989 3.022 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.836 2.836 cp_fm_diag_elpa_base 48 14.0 2.304 2.503 2.834 2.834 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.651 2.699 grid_collocate_task_list 110 9.6 2.675 2.693 2.675 2.693 potential_pw2rs 110 12.3 0.021 0.021 2.443 2.450 qs_env_update_s_mstruct 11 6.9 0.001 0.002 2.195 2.269 jit_kernel_multiply 8 15.0 2.241 2.259 2.241 2.259 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.100 2.194 mp_alltoall_d11v 2046 13.8 2.004 2.066 2.004 2.066 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 2.037 2.052 cp_fm_cholesky_decompose 22 10.9 1.935 1.958 1.935 1.958 acc_transpose_blocks 8220 15.4 0.041 0.042 1.886 1.931 qs_create_task_list 11 7.9 0.001 0.001 1.875 1.919 generate_qs_task_list 11 8.9 0.733 0.785 1.874 1.919 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.899 1.908 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.747 1.802 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=89.941000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1230.000000, yerr=57.172784 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/26e59941a7734f05bfa3a24a29949956890fc20f_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 622.759936E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175955383376 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 57761. MP_Allreduce 11084 796. MP_Sync 87 MP_Alltoall 2226 1816722. MP_SendRecv 24320 18752. MP_ISendRecv 24320 18752. MP_Wait 42476 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.048 211.827 211.829 qs_mol_dyn_low 1 2.0 0.008 0.026 211.296 211.309 qs_forces 11 3.9 0.005 0.005 211.139 211.143 qs_energies 11 4.9 0.003 0.005 205.112 205.129 scf_env_do_scf 11 5.9 0.001 0.001 188.247 188.251 scf_env_do_scf_inner_loop 117 6.6 0.021 0.084 167.228 167.231 qs_scf_new_mos 117 7.6 0.001 0.001 127.237 127.626 qs_scf_loop_do_ot 117 8.6 0.001 0.001 127.236 127.625 dbcsr_multiply_generic 2507 12.6 0.179 0.184 126.372 127.097 velocity_verlet 10 3.0 0.011 0.041 125.561 125.569 ot_scf_mini 117 9.6 0.003 0.003 120.526 120.889 multiply_cannon 2507 13.6 0.234 0.240 101.975 103.956 multiply_cannon_loop 2507 14.6 2.390 2.465 99.733 101.632 ot_mini 117 10.6 0.001 0.001 66.772 67.157 multiply_cannon_multrec 60168 15.6 31.907 34.205 41.831 43.931 qs_ot_get_derivative 117 11.6 0.001 0.002 41.814 42.164 rebuild_ks_matrix 128 8.3 0.001 0.001 33.917 34.229 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.020 33.917 34.228 mp_waitall_1 267128 16.5 29.105 32.021 29.105 32.021 qs_ot_get_p 128 10.4 0.001 0.001 31.344 31.833 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.462 30.737 multiply_cannon_sync_h2d 60168 15.6 26.436 28.830 26.436 28.830 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.398 25.091 apply_single 128 13.6 0.001 0.001 24.397 25.091 qs_ot_p2m_diag 83 11.4 0.079 0.091 24.638 24.729 ot_diis_step 117 11.6 0.007 0.008 24.685 24.686 cp_dbcsr_syevd 83 12.4 0.005 0.005 21.837 21.839 init_scf_loop 11 6.9 0.000 0.001 20.942 20.944 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 20.178 20.430 cp_fm_diag_elpa 83 13.4 0.000 0.000 18.712 18.748 cp_fm_diag_elpa_base 83 14.4 18.630 18.680 18.707 18.742 multiply_cannon_metrocomm3 60168 15.6 0.116 0.122 15.968 18.108 prepare_preconditioner 11 7.9 0.000 0.000 16.325 16.364 make_preconditioner 11 8.9 0.000 0.000 16.325 16.364 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.543 15.751 make_m2s 5014 13.6 0.104 0.113 14.133 14.554 make_images 5014 14.6 0.397 0.416 13.948 14.379 sum_up_and_integrate 128 10.3 0.002 0.005 14.238 14.255 integrate_v_rspace 128 11.3 0.003 0.004 14.179 14.197 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.455 13.603 calculate_rho_elec 128 8.7 0.045 0.064 13.454 13.603 init_scf_run 11 5.9 0.000 0.001 12.662 12.663 scf_env_initial_rho_setup 11 6.9 0.001 0.002 12.662 12.663 mp_sum_l 7950 12.9 9.013 10.189 9.013 10.189 dbcsr_mm_accdrv_process 124484 16.2 4.726 4.936 9.486 10.121 cp_fm_cholesky_invert 11 10.9 9.374 9.382 9.374 9.382 wfi_extrapolate 11 7.9 0.002 0.011 9.245 9.245 calculate_dm_sparse 128 9.5 0.001 0.001 8.716 8.799 qs_ot_get_orbitals 117 10.6 0.001 0.001 8.177 8.268 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.005 8.167 pw_transfer 1547 11.6 0.074 0.087 7.854 8.032 multiply_cannon_metrocomm1 60168 15.6 0.093 0.097 6.424 7.926 density_rs2pw 128 9.7 0.006 0.006 7.025 7.923 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.652 7.834 make_images_data 5014 15.6 0.067 0.074 6.880 7.777 grid_integrate_task_list 128 12.3 7.070 7.480 7.070 7.480 hybrid_alltoall_any 5200 16.5 0.294 2.277 6.040 7.257 fft_wrap_pw1pw2_140 523 13.2 0.858 0.900 6.625 6.809 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.684 6.695 fft3d_ps 1291 14.7 2.197 2.854 5.656 6.004 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.776 5.925 mp_alltoall_d11v 2415 14.1 4.525 5.515 4.525 5.515 grid_collocate_task_list 128 9.7 4.832 5.290 4.832 5.290 mp_sum_d 4472 12.1 3.801 4.770 3.801 4.770 cp_fm_cholesky_decompose 22 10.9 4.744 4.758 4.744 4.758 potential_pw2rs 128 12.3 0.009 0.010 4.702 4.724 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=211.829000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=589.727273, yerr=5.878944 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/26e59941a7734f05bfa3a24a29949956890fc20f_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 835.727360E+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 5102735. 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 0.011 0.036 188.734 188.742 qs_mol_dyn_low 1 2.0 0.006 0.023 188.253 188.276 qs_forces 11 3.9 0.004 0.004 188.000 188.003 qs_energies 11 4.9 0.035 0.084 181.280 181.295 scf_env_do_scf 11 5.9 0.001 0.001 163.566 163.576 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 131.103 131.106 velocity_verlet 10 3.0 0.003 0.010 117.275 117.280 dbcsr_multiply_generic 2485 12.5 0.186 0.192 97.551 98.836 qs_scf_new_mos 116 7.6 0.001 0.001 93.104 93.613 qs_scf_loop_do_ot 116 8.6 0.001 0.001 93.103 93.613 ot_scf_mini 116 9.6 0.004 0.004 88.346 88.930 multiply_cannon 2485 13.5 0.496 0.554 77.855 82.172 multiply_cannon_loop 2485 14.5 1.547 1.613 74.980 77.598 ot_mini 116 10.6 0.001 0.001 49.936 50.471 multiply_cannon_multrec 29820 15.5 20.830 25.538 32.683 37.970 mp_waitall_1 212858 16.6 23.672 37.288 23.672 37.288 rebuild_ks_matrix 127 8.3 0.001 0.001 31.745 32.433 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.019 31.745 32.433 init_scf_loop 11 6.9 0.000 0.000 32.377 32.378 qs_ks_update_qs_env 127 7.6 0.001 0.001 28.517 29.159 qs_ot_get_derivative 116 11.6 0.001 0.002 28.126 28.703 multiply_cannon_metrocomm3 29820 15.5 0.098 0.105 15.513 28.300 prepare_preconditioner 11 7.9 0.000 0.000 28.128 28.221 make_preconditioner 11 8.9 0.000 0.000 28.128 28.221 make_full_inverse_cholesky 11 9.9 0.000 0.000 26.812 27.356 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 21.934 23.119 apply_single 127 13.6 0.001 0.001 21.933 23.118 ot_diis_step 116 11.6 0.014 0.015 21.637 21.639 qs_ot_get_p 127 10.4 0.001 0.001 20.747 21.460 multiply_cannon_sync_h2d 29820 15.5 17.914 19.901 17.914 19.901 cp_fm_cholesky_invert 11 10.9 16.426 16.438 16.426 16.438 qs_ot_p2m_diag 82 11.4 0.185 0.213 15.962 15.995 make_m2s 4970 13.5 0.085 0.092 13.871 15.055 make_images 4970 14.5 1.150 1.340 13.657 14.839 cp_dbcsr_syevd 82 12.4 0.005 0.006 14.784 14.785 sum_up_and_integrate 127 10.3 0.002 0.004 13.783 13.811 integrate_v_rspace 127 11.3 0.003 0.003 13.723 13.754 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.307 13.345 calculate_rho_elec 127 8.7 0.086 0.103 13.306 13.344 init_scf_run 11 5.9 0.000 0.001 12.841 12.843 scf_env_initial_rho_setup 11 6.9 0.025 0.032 12.841 12.843 dbcsr_mm_accdrv_process 61726 16.2 6.241 7.430 11.302 11.941 cp_fm_diag_elpa 82 13.4 0.000 0.000 11.554 11.581 cp_fm_diag_elpa_base 82 14.4 11.307 11.385 11.548 11.577 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 10.999 11.439 multiply_cannon_metrocomm4 27335 15.5 0.098 0.112 3.836 10.653 mp_irecv_dv 68888 16.3 3.633 10.246 3.633 10.246 make_images_data 4970 15.5 0.066 0.074 8.168 9.697 hybrid_alltoall_any 5155 16.4 0.347 1.544 7.175 9.418 pw_transfer 1535 11.6 0.084 0.096 8.437 8.495 wfi_extrapolate 11 7.9 0.001 0.001 8.287 8.288 fft_wrap_pw1pw2 1281 12.7 0.010 0.010 8.216 8.278 density_rs2pw 127 9.7 0.006 0.006 6.918 7.490 grid_integrate_task_list 127 12.3 7.139 7.454 7.139 7.454 fft_wrap_pw1pw2_140 519 13.2 0.926 0.951 7.294 7.382 calculate_dm_sparse 127 9.5 0.001 0.001 7.225 7.373 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.249 7.017 cp_fm_cholesky_decompose 22 10.9 6.819 6.891 6.819 6.891 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.654 6.664 mp_sum_l 7884 12.9 4.091 6.168 4.091 6.168 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.750 5.932 fft3d_ps 1281 14.7 2.804 2.979 5.793 5.851 qs_ot_get_orbitals 116 10.6 0.001 0.001 5.324 5.393 grid_collocate_task_list 127 9.7 4.963 5.370 4.963 5.370 mp_alltoall_d11v 2401 14.1 4.559 5.063 4.559 5.063 mp_allgather_i34 2485 14.5 1.412 4.379 1.412 4.379 potential_pw2rs 127 12.3 0.015 0.017 4.356 4.378 calculate_first_density_matrix 1 7.0 0.000 0.000 4.347 4.373 mp_sum_d 4454 12.1 2.608 4.004 2.608 4.004 dbcsr_complete_redistribute 393 12.7 0.760 0.841 3.052 3.895 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=188.742000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=797.181818, yerr=1.748671 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/26e59941a7734f05bfa3a24a29949956890fc20f_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 2.928533E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 3984192 0.0% 0.0% 100.0% average stack size 0.0 0.0 1695.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 917.102592E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1042912 MPI messages size (bytes): total size 2.716210E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.604448E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 281856 36943429632 131072 < size <= 4194304 660064 996105256960 4194304 < size <= 16777216 65632 931530938576 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58351. MP_Allreduce 11057 1000. MP_Sync 87 MP_Alltoall 1712 9388896. MP_SendRecv 7936 75008. MP_ISendRecv 7936 75008. MP_Wait 21820 MP_ISend 11748 275205. MP_IRecv 11748 275205. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.020 0.054 175.965 175.968 qs_mol_dyn_low 1 2.0 0.010 0.056 175.428 175.442 qs_forces 11 3.9 0.008 0.035 175.276 175.285 qs_energies 11 4.9 0.016 0.073 168.719 168.733 scf_env_do_scf 11 5.9 0.001 0.002 151.049 151.051 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 116.685 116.686 velocity_verlet 10 3.0 0.017 0.051 111.319 111.324 dbcsr_multiply_generic 2507 12.6 0.184 0.188 83.421 84.759 qs_scf_new_mos 117 7.6 0.001 0.001 79.625 79.952 qs_scf_loop_do_ot 117 8.6 0.001 0.001 79.624 79.951 ot_scf_mini 117 9.6 0.004 0.004 75.379 75.780 multiply_cannon 2507 13.6 0.506 0.521 63.626 69.370 multiply_cannon_loop 2507 14.6 1.125 1.194 60.717 63.761 ot_mini 117 10.6 0.001 0.001 42.478 42.884 mp_waitall_1 170520 16.6 24.468 34.774 24.468 34.774 init_scf_loop 11 6.9 0.001 0.005 34.268 34.269 rebuild_ks_matrix 128 8.3 0.001 0.001 30.125 30.612 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.023 30.125 30.612 prepare_preconditioner 11 7.9 0.000 0.000 30.261 30.311 make_preconditioner 11 8.9 0.000 0.001 30.261 30.311 make_full_inverse_cholesky 11 9.9 0.000 0.001 27.971 29.318 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.159 27.601 multiply_cannon_multrec 20056 15.6 13.207 16.674 23.522 27.057 multiply_cannon_metrocomm3 20056 15.6 0.062 0.065 15.079 25.059 qs_ot_get_derivative 117 11.6 0.001 0.002 23.033 23.436 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.519 20.622 apply_single 128 13.6 0.001 0.001 19.519 20.621 ot_diis_step 117 11.6 0.018 0.018 19.339 19.340 qs_ot_get_p 128 10.4 0.001 0.001 18.287 18.743 make_m2s 5014 13.6 0.078 0.086 14.651 15.824 make_images 5014 14.6 1.141 1.232 14.414 15.580 multiply_cannon_sync_h2d 20056 15.6 13.640 15.345 13.640 15.345 cp_fm_cholesky_invert 11 10.9 14.040 14.048 14.040 14.048 qs_ot_p2m_diag 83 11.4 0.265 0.273 13.937 13.944 sum_up_and_integrate 128 10.3 0.002 0.004 13.838 13.863 integrate_v_rspace 128 11.3 0.003 0.004 13.778 13.803 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.574 13.613 calculate_rho_elec 128 8.7 0.129 0.143 13.573 13.612 cp_dbcsr_syevd 83 12.4 0.006 0.008 12.890 12.891 init_scf_run 11 5.9 0.000 0.001 12.439 12.440 scf_env_initial_rho_setup 11 6.9 0.001 0.003 12.439 12.440 make_images_data 5014 15.6 0.064 0.074 8.813 10.419 hybrid_alltoall_any 5200 16.5 0.446 2.028 7.651 10.040 dbcsr_mm_accdrv_process 41502 16.2 6.665 6.973 9.767 9.894 cp_fm_diag_elpa 83 13.4 0.001 0.002 9.771 9.788 cp_fm_diag_elpa_base 83 14.4 9.365 9.525 9.766 9.781 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.103 9.404 multiply_cannon_metrocomm4 17549 15.6 0.065 0.074 3.451 9.110 mp_irecv_dv 50230 16.2 3.321 8.851 3.321 8.851 pw_transfer 1547 11.6 0.084 0.103 8.706 8.806 fft_wrap_pw1pw2 1291 12.7 0.009 0.011 8.483 8.589 fft_wrap_pw1pw2_140 523 13.2 1.061 1.096 7.558 7.669 grid_integrate_task_list 128 12.3 7.321 7.640 7.321 7.640 wfi_extrapolate 11 7.9 0.002 0.012 7.435 7.435 density_rs2pw 128 9.7 0.006 0.006 6.817 7.320 cp_fm_cholesky_decompose 22 10.9 7.163 7.178 7.163 7.178 cp_fm_upper_to_full 105 14.8 5.712 7.163 5.712 7.163 calculate_dm_sparse 128 9.5 0.001 0.001 6.813 6.934 dbcsr_complete_redistribute 395 12.7 1.158 1.186 4.450 6.212 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.037 6.043 fft3d_ps 1291 14.7 2.818 3.059 5.797 5.885 grid_collocate_task_list 128 9.7 5.162 5.541 5.162 5.541 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.601 5.306 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 5.149 5.279 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.325 5.085 mp_alltoall_d11v 2415 14.1 4.348 4.933 4.348 4.933 calculate_first_density_matrix 1 7.0 0.001 0.004 4.818 4.821 mp_sum_l 7950 12.9 3.315 4.795 3.315 4.795 mp_allgather_i34 2507 14.6 1.417 4.666 1.417 4.666 potential_pw2rs 128 12.3 0.020 0.022 4.258 4.265 transfer_fm_to_dbcsr 11 9.9 0.021 0.024 2.270 4.012 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.897 3.921 mp_alltoall_i22 716 14.1 1.933 3.868 1.933 3.868 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.852 3.859 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=175.968000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=869.818182, yerr=4.508944 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/26e59941a7734f05bfa3a24a29949956890fc20f_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 4.387043E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 6024576 0.0% 0.0% 100.0% average stack size 0.0 0.0 1129.8 marketing flops 145.647559E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 1.136685E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1152312 MPI messages size (bytes): total size 2.039354E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.769794E+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 321600 36333158400 131072 < size <= 4194304 721560 792041881600 4194304 < size <= 16777216 70800 669921260160 16777216 < size 30960 541065216000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4024 57894. MP_Allreduce 11137 1080. MP_Sync 87 MP_Alltoall 1724 12509439. MP_SendRecv 5934 75008. MP_ISendRecv 5934 75008. MP_Wait 22612 MP_ISend 15064 244788. MP_IRecv 15064 244788. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.055 0.091 190.485 190.486 qs_mol_dyn_low 1 2.0 0.003 0.003 189.696 189.709 qs_forces 11 3.9 0.004 0.004 189.588 189.596 qs_energies 11 4.9 0.002 0.002 182.455 182.462 scf_env_do_scf 11 5.9 0.001 0.001 164.293 164.305 velocity_verlet 10 3.0 0.002 0.002 125.289 125.292 scf_env_do_scf_inner_loop 118 6.6 0.003 0.009 118.000 118.002 dbcsr_multiply_generic 2527 12.6 0.194 0.200 82.884 83.659 qs_scf_new_mos 118 7.6 0.001 0.001 81.431 81.741 qs_scf_loop_do_ot 118 8.6 0.001 0.001 81.431 81.740 ot_scf_mini 118 9.6 0.003 0.004 76.686 77.000 multiply_cannon 2527 13.6 0.556 0.586 58.012 61.455 multiply_cannon_loop 2527 14.6 1.858 1.922 54.137 55.984 init_scf_loop 11 6.9 0.000 0.000 46.169 46.171 ot_mini 118 10.6 0.001 0.001 43.478 43.783 prepare_preconditioner 11 7.9 0.000 0.000 42.108 42.140 make_preconditioner 11 8.9 0.000 0.000 42.108 42.140 make_full_inverse_cholesky 11 9.9 0.011 0.023 35.670 40.774 multiply_cannon_multrec 30324 15.6 13.701 19.075 27.935 32.987 rebuild_ks_matrix 129 8.3 0.001 0.001 29.426 29.773 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.018 29.425 29.772 mp_waitall_1 149060 16.7 17.211 26.895 17.211 26.895 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.546 26.865 qs_ot_get_derivative 118 11.6 0.001 0.002 23.613 23.932 make_m2s 5054 13.6 0.092 0.096 20.427 21.522 make_images 5054 14.6 1.952 2.299 20.115 21.204 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 19.235 19.800 apply_single 129 13.6 0.001 0.001 19.234 19.800 ot_diis_step 118 11.6 0.018 0.020 19.735 19.736 qs_ot_get_p 129 10.4 0.001 0.001 19.212 19.567 cp_fm_upper_to_full 105 14.8 11.519 17.028 11.519 17.028 cp_fm_cholesky_invert 11 10.9 16.002 16.011 16.002 16.011 multiply_cannon_metrocomm3 30324 15.6 0.050 0.053 6.450 15.050 qs_ot_p2m_diag 83 11.4 0.342 0.389 14.928 14.978 dbcsr_mm_accdrv_process 62756 16.2 9.534 10.461 13.800 14.305 sum_up_and_integrate 129 10.3 0.002 0.003 14.041 14.068 integrate_v_rspace 129 11.3 0.003 0.004 13.981 14.008 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.888 13.914 calculate_rho_elec 129 8.7 0.172 0.187 13.887 13.913 cp_dbcsr_syevd 83 12.4 0.005 0.006 13.526 13.527 dbcsr_complete_redistribute 395 12.7 1.486 1.597 9.129 12.915 make_images_data 5054 15.6 0.066 0.073 10.946 12.792 init_scf_run 11 5.9 0.000 0.001 12.230 12.231 scf_env_initial_rho_setup 11 6.9 0.019 0.023 12.230 12.231 hybrid_alltoall_any 5240 16.5 0.534 2.229 9.782 11.662 copy_fm_to_dbcsr 209 11.7 0.001 0.002 7.766 11.540 multiply_cannon_sync_h2d 30324 15.6 10.486 11.465 10.486 11.465 cp_fm_diag_elpa 83 13.4 0.000 0.000 10.327 10.348 cp_fm_diag_elpa_base 83 14.4 9.348 9.660 10.321 10.340 transfer_fm_to_dbcsr 11 9.9 0.001 0.008 6.415 10.113 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.845 10.082 mp_alltoall_i22 718 14.1 5.665 9.525 5.665 9.525 pw_transfer 1559 11.6 0.087 0.104 9.214 9.284 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 8.987 9.062 fft_wrap_pw1pw2_140 527 13.2 1.257 1.291 8.016 8.105 grid_integrate_task_list 129 12.3 7.566 7.892 7.566 7.892 wfi_extrapolate 11 7.9 0.001 0.001 7.589 7.589 cp_fm_cholesky_decompose 22 10.9 7.394 7.489 7.394 7.489 density_rs2pw 129 9.7 0.005 0.006 6.846 7.487 calculate_dm_sparse 129 9.5 0.001 0.001 7.299 7.410 multiply_cannon_metrocomm4 25270 15.6 0.079 0.089 2.848 7.184 mp_irecv_dv 76685 16.2 2.692 6.899 2.692 6.899 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.081 6.123 fft3d_ps 1301 14.7 3.020 3.086 6.014 6.091 mp_alltoall_d11v 2423 14.1 5.089 5.616 5.089 5.616 grid_collocate_task_list 129 9.7 5.326 5.613 5.326 5.613 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.981 5.118 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 4.589 4.697 calculate_first_density_matrix 1 7.0 0.000 0.000 4.426 4.445 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 4.425 4.426 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.343 4.405 potential_pw2rs 129 12.3 0.023 0.024 4.267 4.273 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=190.486000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1077.181818, yerr=17.087040 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/26e59941a7734f05bfa3a24a29949956890fc20f_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 5.909858E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 1976192 0.0% 0.0% 100.0% average stack size 0.0 0.0 3444.2 marketing flops 145.646636E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 1.539781E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 242592 MPI messages size (bytes): total size 1.341705E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.530704E+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 114816 60196651008 4194304 < size <= 16777216 105840 554906419200 16777216 < size 20352 726591885696 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 9003 51. MP_Alltoall 9708 795382. MP_ISend 40388 2101966. MP_IRecv 40388 2101070. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4024 57892. MP_Allreduce 11137 1164. MP_Sync 87 MP_Alltoall 1724 18848449. 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 0.045 0.070 170.810 170.810 qs_mol_dyn_low 1 2.0 0.003 0.003 169.970 170.008 qs_forces 11 3.9 0.004 0.004 169.851 169.854 qs_energies 11 4.9 0.009 0.020 162.454 162.460 scf_env_do_scf 11 5.9 0.001 0.001 143.761 143.771 velocity_verlet 10 3.0 0.030 0.038 110.853 110.857 scf_env_do_scf_inner_loop 118 6.6 0.003 0.009 108.783 108.784 dbcsr_multiply_generic 2527 12.6 0.185 0.190 73.871 74.263 qs_scf_new_mos 118 7.6 0.001 0.001 71.449 71.535 qs_scf_loop_do_ot 118 8.6 0.001 0.001 71.448 71.535 ot_scf_mini 118 9.6 0.003 0.004 67.023 67.083 multiply_cannon 2527 13.6 0.567 0.606 55.411 58.592 multiply_cannon_loop 2527 14.6 0.811 0.849 52.505 53.357 ot_mini 118 10.6 0.001 0.001 37.911 37.968 init_scf_loop 11 6.9 0.000 0.000 34.832 34.832 prepare_preconditioner 11 7.9 0.000 0.000 31.006 31.029 make_preconditioner 11 8.9 0.000 0.000 31.006 31.029 mp_waitall_1 126780 16.7 24.606 30.736 24.606 30.736 rebuild_ks_matrix 129 8.3 0.001 0.001 29.381 29.491 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.018 29.381 29.490 make_full_inverse_cholesky 11 9.9 0.015 0.025 28.978 29.269 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.826 26.925 multiply_cannon_multrec 10108 15.6 10.459 14.557 19.671 22.959 qs_ot_get_derivative 118 11.6 0.002 0.002 20.809 20.872 multiply_cannon_metrocomm3 10108 15.6 0.024 0.026 12.646 20.104 cp_fm_cholesky_invert 11 10.9 17.634 17.639 17.634 17.639 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 17.090 17.245 apply_single 129 13.6 0.001 0.001 17.090 17.244 ot_diis_step 118 11.6 0.020 0.021 17.031 17.031 qs_ot_get_p 129 10.4 0.001 0.001 16.373 16.467 make_m2s 5054 13.6 0.065 0.068 14.656 15.455 make_images 5054 14.6 2.168 2.602 14.345 15.140 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.282 14.319 calculate_rho_elec 129 8.7 0.255 0.266 14.282 14.319 sum_up_and_integrate 129 10.3 0.002 0.002 14.257 14.299 integrate_v_rspace 129 11.3 0.004 0.004 14.196 14.241 qs_ot_p2m_diag 83 11.4 0.495 0.501 12.661 12.679 init_scf_run 11 5.9 0.000 0.001 11.849 11.849 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.849 11.849 multiply_cannon_sync_h2d 10108 15.6 10.980 11.591 10.980 11.591 cp_dbcsr_syevd 83 12.4 0.005 0.006 11.504 11.505 pw_transfer 1559 11.6 0.086 0.095 9.906 9.940 make_images_data 5054 15.6 0.055 0.065 8.446 9.726 fft_wrap_pw1pw2 1301 12.7 0.010 0.010 9.682 9.721 hybrid_alltoall_any 5240 16.5 0.846 3.809 8.317 9.695 dbcsr_mm_accdrv_process 20926 16.1 4.258 5.933 8.856 9.541 cp_fm_diag_elpa 83 13.4 0.000 0.000 8.678 8.690 cp_fm_diag_elpa_base 83 14.4 8.445 8.521 8.674 8.686 fft_wrap_pw1pw2_140 527 13.2 1.688 1.725 8.600 8.640 grid_integrate_task_list 129 12.3 7.838 8.110 7.838 8.110 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 7.938 7.989 cp_fm_cholesky_decompose 22 10.9 7.766 7.868 7.766 7.868 density_rs2pw 129 9.7 0.005 0.005 6.974 7.295 wfi_extrapolate 11 7.9 0.001 0.001 7.271 7.271 multiply_cannon_metrocomm1 10108 15.6 0.030 0.031 4.316 7.224 calculate_dm_sparse 129 9.5 0.001 0.001 7.010 7.089 fft3d_ps 1301 14.7 3.164 3.265 6.103 6.139 grid_collocate_task_list 129 9.7 5.609 5.887 5.609 5.887 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.679 5.693 dbcsr_complete_redistribute 395 12.7 2.091 2.169 5.078 5.465 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.269 5.269 mp_alltoall_d11v 2423 14.1 4.441 4.860 4.441 4.860 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.620 4.670 mp_allgather_i34 2527 14.6 1.091 4.548 1.091 4.548 potential_pw2rs 129 12.3 0.026 0.026 4.362 4.369 calculate_first_density_matrix 1 7.0 0.000 0.000 4.331 4.332 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 4.092 4.156 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.595 3.885 multiply_cannon_metrocomm4 7581 15.6 0.026 0.028 1.736 3.866 mp_irecv_dv 29086 15.9 1.695 3.788 1.695 3.788 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.577 3.604 copy_fm_to_dbcsr 209 11.7 0.001 0.002 3.262 3.590 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.388 3.535 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.465 3.478 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=170.810000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1439.909091, yerr=37.705689 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/26e59941a7734f05bfa3a24a29949956890fc20f_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456868864 0.0% 0.0% 100.0% flops 32 x 32 x 32 1967095021568 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986252263424 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992004743168 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753954734080 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 11613074706432 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239170400256 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239170400256 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911117062144 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.237285E+12 0.0% 0.0% 100.0% flops max/rank 11.786603E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806448352 0.0% 0.0% 100.0% number of processed stacks 1981024 0.0% 0.0% 100.0% average stack size 0.0 0.0 3435.8 marketing flops 145.655226E+12 ------------------------------------------------------------------------------- # multiplications 2531 max memory usage/rank 3.176133E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101240 MPI messages size (bytes): total size 1.145036E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.310119E+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 45728 35500589056 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592178544 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4080 58046. MP_Allreduce 11277 1498. MP_Sync 87 MP_Alltoall 1724 36993574. MP_SendRecv 1806 218624. MP_ISendRecv 1806 218624. MP_Wait 9876 MP_ISend 6456 1080169. MP_IRecv 6456 1080169. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.022 0.042 295.064 295.066 qs_mol_dyn_low 1 2.0 0.003 0.003 293.701 294.274 qs_forces 11 3.9 0.004 0.005 293.295 293.300 qs_energies 11 4.9 0.002 0.003 284.362 284.369 scf_env_do_scf 11 5.9 0.001 0.001 259.883 259.895 velocity_verlet 10 3.0 0.003 0.006 211.875 211.912 scf_env_do_scf_inner_loop 118 6.6 0.032 0.037 136.351 136.353 init_scf_loop 11 6.9 0.000 0.000 123.264 123.266 prepare_preconditioner 11 7.9 0.000 0.000 118.281 118.311 make_preconditioner 11 8.9 0.000 0.000 118.281 118.311 make_full_inverse_cholesky 11 9.9 0.038 0.039 94.482 115.412 qs_scf_new_mos 118 7.6 0.001 0.001 90.918 91.035 qs_scf_loop_do_ot 118 8.6 0.001 0.001 90.918 91.034 ot_scf_mini 118 9.6 0.004 0.004 85.971 86.056 dbcsr_multiply_generic 2531 12.6 0.217 0.227 85.067 85.856 cp_fm_upper_to_full 105 14.8 52.326 75.100 52.326 75.100 multiply_cannon 2531 13.6 0.691 0.723 61.414 62.468 multiply_cannon_loop 2531 14.6 1.044 1.057 57.150 59.227 ot_mini 118 10.6 0.001 0.001 45.310 45.418 dbcsr_complete_redistribute 395 12.7 3.973 4.063 29.722 42.252 copy_fm_to_dbcsr 209 11.7 0.001 0.002 26.255 38.765 transfer_fm_to_dbcsr 11 9.9 0.029 0.031 23.756 36.145 rebuild_ks_matrix 129 8.3 0.001 0.001 35.213 35.362 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.018 35.212 35.361 mp_waitall_1 104660 16.8 28.752 34.351 28.752 34.351 mp_alltoall_i22 718 14.1 21.551 34.202 21.551 34.202 cp_fm_cholesky_invert 11 10.9 33.174 33.180 33.174 33.180 qs_ks_update_qs_env 129 7.6 0.001 0.001 32.714 32.850 qs_ot_get_p 129 10.4 0.001 0.001 25.407 25.507 qs_ot_get_derivative 118 11.6 0.002 0.002 24.865 24.950 multiply_cannon_metrocomm3 10124 15.6 0.025 0.026 19.337 21.295 qs_ot_p2m_diag 83 11.4 0.879 0.884 21.178 21.207 make_m2s 5062 13.6 0.074 0.077 18.825 20.529 ot_diis_step 118 11.6 0.022 0.024 20.399 20.399 make_images 5062 14.6 3.071 3.289 18.339 20.042 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.556 19.909 apply_single 129 13.6 0.001 0.001 19.556 19.908 multiply_cannon_multrec 10124 15.6 10.353 12.256 19.349 19.486 cp_dbcsr_syevd 83 12.4 0.006 0.006 19.389 19.393 qs_rho_update_rho_low 129 7.7 0.001 0.001 17.465 17.475 calculate_rho_elec 129 8.7 0.482 0.482 17.464 17.475 sum_up_and_integrate 129 10.3 0.002 0.002 16.608 16.695 integrate_v_rspace 129 11.3 0.005 0.005 16.544 16.632 cp_fm_diag_elpa 83 13.4 0.000 0.000 16.229 16.229 cp_fm_diag_elpa_base 83 14.4 11.835 13.455 16.224 16.225 multiply_cannon_sync_h2d 10124 15.6 14.433 14.477 14.433 14.477 init_scf_run 11 5.9 0.000 0.001 13.759 13.760 scf_env_initial_rho_setup 11 6.9 0.001 0.001 13.759 13.760 hybrid_alltoall_any 5248 16.5 1.324 3.116 11.005 13.539 pw_transfer 1559 11.6 0.096 0.098 13.294 13.301 make_images_data 5062 15.6 0.064 0.069 10.715 13.119 fft_wrap_pw1pw2 1301 12.7 0.011 0.011 13.054 13.060 fft_wrap_pw1pw2_140 527 13.2 3.239 3.312 11.616 11.628 dbcsr_mm_accdrv_process 20942 16.1 4.346 6.279 8.743 10.690 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.696 9.759 wfi_extrapolate 11 7.9 0.001 0.001 9.041 9.041 grid_integrate_task_list 129 12.3 8.620 8.817 8.620 8.817 cp_fm_cholesky_decompose 22 10.9 8.767 8.785 8.767 8.785 qs_energies_init_hamiltonians 11 5.9 0.010 0.012 8.512 8.514 density_rs2pw 129 9.7 0.005 0.005 8.430 8.478 calculate_dm_sparse 129 9.5 0.001 0.001 8.040 8.090 fft3d_ps 1301 14.7 4.005 4.034 7.479 7.523 mp_alltoall_d11v 2423 14.1 6.973 7.108 6.973 7.108 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.736 6.797 grid_collocate_task_list 129 9.7 6.468 6.545 6.468 6.545 copy_dbcsr_to_fm 186 11.8 0.004 0.004 6.149 6.288 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=295.066000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2827.000000, yerr=180.334033 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/26e59941a7734f05bfa3a24a29949956890fc20f_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 1.251832E+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 41318200. 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 0.054 0.122 88.081 88.083 qs_energies 1 2.0 0.000 0.000 87.115 87.141 ls_scf 1 3.0 0.000 0.000 85.574 85.595 dbcsr_multiply_generic 111 6.7 0.014 0.015 74.189 74.357 multiply_cannon 111 7.7 0.017 0.020 57.637 58.935 multiply_cannon_loop 111 8.7 0.227 0.240 54.194 55.457 ls_scf_main 1 4.0 0.000 0.000 52.205 52.205 density_matrix_trs4 2 5.0 0.002 0.003 46.702 46.794 ls_scf_init_scf 1 4.0 0.000 0.000 30.326 30.327 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.202 29.264 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.605 26.622 mp_waitall_1 11031 10.9 22.592 24.843 22.592 24.843 multiply_cannon_multrec 2664 9.7 8.149 8.947 15.540 17.345 multiply_cannon_sync_h2d 2664 9.7 13.456 15.456 13.456 15.456 make_m2s 222 7.7 0.009 0.011 12.950 13.512 make_images 222 8.7 0.098 0.107 12.927 13.491 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.713 12.426 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.650 8.754 make_images_data 222 9.7 0.004 0.005 7.567 8.048 dbcsr_mm_accdrv_process 4760 10.4 0.598 0.697 7.009 7.946 hybrid_alltoall_any 227 10.6 0.215 1.834 6.525 7.744 dbcsr_mm_accdrv_process_sort 4760 11.4 6.212 7.058 6.212 7.058 calculate_norms 4752 9.8 5.490 6.067 5.490 6.067 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.030 5.124 mp_sum_l 887 5.1 3.080 4.659 3.080 4.659 multiply_cannon_metrocomm4 2442 9.7 0.012 0.016 2.052 4.158 mp_irecv_dv 6231 10.9 2.035 4.122 2.035 4.122 make_images_sizes 222 9.7 0.000 0.000 0.666 3.452 mp_alltoall_i44 222 10.7 0.665 3.451 0.665 3.451 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.307 3.322 arnoldi_extremal 4 6.8 0.000 0.000 3.198 3.228 arnoldi_normal_ev 4 7.8 0.001 0.003 3.198 3.228 build_subspace 16 8.4 0.009 0.012 3.106 3.108 ls_scf_post 1 4.0 0.000 0.000 3.044 3.064 ls_scf_store_result 1 5.0 0.000 0.000 2.843 2.876 dbcsr_special_finalize 555 9.7 0.005 0.006 2.364 2.776 dbcsr_merge_single_wm 555 10.7 0.460 0.592 2.355 2.768 compute_matrix_preconditioner 1 6.0 0.000 0.001 2.746 2.750 make_images_pack 222 9.7 2.207 2.670 2.209 2.671 dbcsr_matrix_vector_mult 304 9.0 0.006 0.014 2.313 2.563 dbcsr_sort_data 658 11.4 2.153 2.501 2.153 2.501 dbcsr_matrix_vector_mult_local 304 10.0 2.070 2.476 2.072 2.478 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.268 2.340 acc_transpose_blocks 2664 9.7 0.018 0.020 1.958 2.124 buffer_matrices_ensure_size 222 8.7 1.751 2.040 1.751 2.040 acc_transpose_blocks_kernels 2664 10.7 0.035 0.038 1.819 1.981 jit_kernel_transpose 1 13.0 1.783 1.944 1.783 1.944 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=88.083000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1133.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/26e59941a7734f05bfa3a24a29949956890fc20f_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 2.155405E+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 0.060 0.125 93.928 93.932 qs_energies 1 2.0 0.002 0.009 93.023 93.069 ls_scf 1 3.0 0.007 0.027 90.578 90.618 dbcsr_multiply_generic 111 6.7 0.015 0.016 75.749 76.294 multiply_cannon 111 7.7 0.028 0.043 54.302 58.539 multiply_cannon_loop 111 8.7 0.135 0.146 51.585 54.938 ls_scf_main 1 4.0 0.051 0.245 54.805 54.806 density_matrix_trs4 2 5.0 0.129 0.582 48.576 49.040 ls_scf_init_scf 1 4.0 0.000 0.002 32.030 32.032 ls_scf_init_matrix_S 1 5.0 0.000 0.000 30.312 30.410 mp_waitall_1 9105 10.9 20.807 29.649 20.807 29.649 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.004 27.964 27.974 multiply_cannon_multrec 1332 9.7 13.337 17.330 22.633 27.764 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.572 20.582 make_m2s 222 7.7 0.006 0.007 14.926 15.457 make_images 222 8.7 1.369 1.691 14.896 15.426 dbcsr_mm_accdrv_process 4041 10.4 0.341 0.521 8.891 10.406 dbcsr_mm_accdrv_process_sort 4041 11.4 8.406 9.884 8.406 9.884 make_images_data 222 9.7 0.004 0.005 8.634 9.480 hybrid_alltoall_any 227 10.6 0.542 2.546 8.039 9.249 mp_sum_l 887 5.1 5.069 8.273 5.069 8.273 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.223 7.792 mp_irecv_dv 3311 11.0 3.202 7.733 3.202 7.733 calculate_norms 2376 9.8 6.055 6.770 6.055 6.770 multiply_cannon_sync_h2d 1332 9.7 4.849 6.136 4.849 6.136 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.468 5.954 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.271 5.638 arnoldi_extremal 4 6.8 0.000 0.000 4.718 4.732 arnoldi_normal_ev 4 7.8 0.001 0.005 4.718 4.731 build_subspace 16 8.4 0.014 0.021 4.462 4.465 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.248 4.185 ls_scf_post 1 4.0 0.012 0.049 3.737 3.782 ls_scf_store_result 1 5.0 0.000 0.000 3.447 3.554 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.137 3.384 dbcsr_matrix_vector_mult_local 304 10.0 2.741 3.228 2.743 3.230 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.763 2.949 mp_allgather_i34 111 8.7 0.702 2.852 0.702 2.852 compute_matrix_preconditioner 1 6.0 0.000 0.001 2.449 2.456 qs_energies_init_hamiltonians 1 3.0 0.004 0.022 2.433 2.433 dbcsr_data_new 4174 10.1 2.117 2.410 2.117 2.410 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.367 2.370 rebuild_ks_matrix 3 7.3 0.000 0.001 2.352 2.355 qs_ks_build_kohn_sham_matrix 3 8.3 0.092 0.388 2.352 2.355 make_images_pack 222 9.7 1.820 2.142 1.822 2.144 dbcsr_sort_data 436 11.2 1.838 2.075 1.838 2.075 acc_transpose_blocks 1332 9.7 0.008 0.008 1.744 1.972 acc_transpose_blocks_kernels 1332 10.7 0.020 0.023 1.668 1.893 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=93.932000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1798.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/26e59941a7734f05bfa3a24a29949956890fc20f_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 2.892829E+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 0.214 0.835 97.802 97.805 qs_energies 1 2.0 0.000 0.001 95.619 96.353 ls_scf 1 3.0 0.001 0.007 94.154 94.889 dbcsr_multiply_generic 111 6.7 0.016 0.022 76.207 76.579 ls_scf_main 1 4.0 0.039 0.335 57.181 57.189 multiply_cannon 111 7.7 0.031 0.052 52.736 56.624 multiply_cannon_loop 111 8.7 0.116 0.130 49.964 54.491 density_matrix_trs4 2 5.0 0.052 0.467 50.799 51.251 mp_waitall_1 7281 11.0 23.627 34.020 23.627 34.020 ls_scf_init_scf 1 4.0 0.000 0.003 30.931 30.938 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.706 29.800 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.003 27.334 27.350 multiply_cannon_multrec 888 9.7 12.700 15.186 21.269 24.359 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.081 23.827 make_m2s 222 7.7 0.006 0.008 16.337 17.100 make_images 222 8.7 1.582 1.837 16.299 17.063 make_images_data 222 9.7 0.004 0.005 9.505 10.538 hybrid_alltoall_any 227 10.6 0.643 2.954 9.047 10.123 dbcsr_mm_accdrv_process 3754 10.4 0.312 0.487 8.102 9.331 mp_sum_l 887 5.1 5.453 8.866 5.453 8.866 dbcsr_mm_accdrv_process_sort 3754 11.4 7.666 8.844 7.666 8.844 multiply_cannon_sync_h2d 888 9.7 5.993 7.533 5.993 7.533 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.470 7.390 mp_irecv_dv 2335 11.1 2.454 7.342 2.454 7.342 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.690 7.266 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.014 7.217 ls_scf_post 1 4.0 0.088 0.700 6.040 6.775 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.007 5.267 arnoldi_extremal 4 6.8 0.000 0.000 5.067 5.086 arnoldi_normal_ev 4 7.8 0.001 0.005 5.067 5.086 calculate_norms 1584 9.8 4.417 4.919 4.417 4.919 build_subspace 16 8.4 0.014 0.020 4.752 4.759 mp_allgather_i34 111 8.7 0.888 3.825 0.888 3.825 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.419 3.757 dbcsr_matrix_vector_mult_local 304 10.0 3.017 3.599 3.019 3.601 ls_scf_store_result 1 5.0 0.000 0.000 3.415 3.489 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.041 3.171 write_mo_free_results 1 5.0 1.018 1.638 2.522 2.741 dbcsr_data_new 4116 9.9 2.113 2.470 2.113 2.470 make_images_sizes 222 9.7 0.000 0.000 0.979 2.273 mp_alltoall_i44 222 10.7 0.979 2.272 0.979 2.272 dbcsr_sort_data 325 11.1 1.888 2.169 1.888 2.169 mp_waitany 462 12.2 0.793 1.981 0.793 1.981 transfer_rs2pw 14 8.9 0.000 0.000 0.799 1.978 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=97.805000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2260.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/26e59941a7734f05bfa3a24a29949956890fc20f_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 3.321405E+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 0.109 0.637 99.203 99.203 qs_energies 1 2.0 0.000 0.000 97.799 98.026 ls_scf 1 3.0 0.000 0.000 96.128 96.356 dbcsr_multiply_generic 111 6.7 0.017 0.018 79.315 79.619 ls_scf_main 1 4.0 0.000 0.000 59.180 59.309 multiply_cannon 111 7.7 0.044 0.142 52.279 57.258 density_matrix_trs4 2 5.0 0.002 0.003 52.947 53.064 multiply_cannon_loop 111 8.7 0.154 0.168 47.266 50.508 ls_scf_init_scf 1 4.0 0.000 0.000 33.329 33.459 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.985 32.057 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.532 29.545 mp_waitall_1 6369 11.0 22.800 29.041 22.800 29.041 multiply_cannon_multrec 1332 9.7 14.181 17.476 22.107 25.322 make_m2s 222 7.7 0.007 0.008 21.125 22.512 make_images 222 8.7 3.145 3.615 21.075 22.463 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.286 18.187 make_images_data 222 9.7 0.004 0.004 11.783 13.453 hybrid_alltoall_any 227 10.6 0.798 3.766 11.187 12.861 dbcsr_mm_accdrv_process 3641 10.4 0.313 0.483 7.561 9.066 dbcsr_mm_accdrv_process_sort 3641 11.4 7.118 8.570 7.118 8.570 mp_sum_l 887 5.1 4.840 8.502 4.840 8.502 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.347 6.346 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.092 6.176 mp_irecv_dv 3229 10.9 2.066 6.098 2.066 6.098 multiply_cannon_sync_h2d 1332 9.7 5.372 5.840 5.372 5.840 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.565 5.696 arnoldi_extremal 4 6.8 0.000 0.000 5.195 5.209 arnoldi_normal_ev 4 7.8 0.001 0.005 5.194 5.208 build_subspace 16 8.4 0.014 0.021 4.855 4.862 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.507 4.721 calculate_norms 2376 9.8 4.218 4.623 4.218 4.623 mp_allgather_i34 111 8.7 2.136 4.407 2.136 4.407 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.571 3.872 ls_scf_post 1 4.0 0.000 0.000 3.619 3.846 dbcsr_matrix_vector_mult_local 304 10.0 3.183 3.685 3.185 3.687 dbcsr_sort_data 658 11.4 3.058 3.340 3.058 3.340 dbcsr_special_finalize 555 9.7 0.006 0.007 2.804 3.155 dbcsr_merge_single_wm 555 10.7 0.537 0.660 2.796 3.146 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.008 3.086 ls_scf_store_result 1 5.0 0.000 0.000 3.000 3.063 dbcsr_data_release 10477 10.7 1.594 2.422 1.594 2.422 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.046 2.175 rebuild_ks_matrix 3 7.3 0.000 0.000 2.023 2.152 qs_ks_build_kohn_sham_matrix 3 8.3 0.128 0.257 2.023 2.152 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=99.203000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2773.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/26e59941a7734f05bfa3a24a29949956890fc20f_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 4.734489E+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 0.037 0.076 93.938 93.938 qs_energies 1 2.0 0.000 0.000 92.855 92.860 ls_scf 1 3.0 0.000 0.000 90.845 90.850 dbcsr_multiply_generic 111 6.7 0.017 0.018 72.130 72.338 multiply_cannon 111 7.7 0.074 0.132 53.852 57.646 ls_scf_main 1 4.0 0.000 0.000 56.251 56.253 multiply_cannon_loop 111 8.7 0.088 0.095 51.303 52.699 density_matrix_trs4 2 5.0 0.002 0.003 49.260 49.341 ls_scf_init_scf 1 4.0 0.000 0.000 31.201 31.201 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.663 29.703 mp_waitall_1 5436 11.0 24.163 28.889 24.163 28.889 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.548 27.560 multiply_cannon_multrec 444 9.7 13.721 16.257 20.809 23.251 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.580 15.176 make_m2s 222 7.7 0.005 0.005 13.604 14.524 make_images 222 8.7 2.039 2.486 13.537 14.455 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.073 13.992 hybrid_alltoall_any 227 10.6 0.802 3.830 8.148 9.783 make_images_data 222 9.7 0.003 0.004 8.329 9.591 multiply_cannon_sync_h2d 444 9.7 6.705 8.470 6.705 8.470 dbcsr_mm_accdrv_process 3003 10.4 0.355 0.427 6.782 7.892 dbcsr_mm_accdrv_process_sort 3003 11.4 6.415 7.487 6.415 7.487 arnoldi_extremal 4 6.8 0.000 0.000 5.815 5.827 arnoldi_normal_ev 4 7.8 0.009 0.017 5.815 5.827 build_subspace 16 8.4 0.015 0.020 5.412 5.422 mp_sum_l 887 5.1 2.692 4.859 2.692 4.859 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.440 4.646 dbcsr_matrix_vector_mult 304 9.0 0.011 0.022 4.155 4.361 dbcsr_matrix_vector_mult_local 304 10.0 3.698 4.169 3.700 4.171 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.552 3.992 mp_irecv_dv 1241 11.2 1.537 3.963 1.537 3.963 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.877 3.873 calculate_norms 792 9.8 3.596 3.724 3.596 3.724 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.541 3.621 mp_allgather_i34 111 8.7 1.128 3.614 1.128 3.614 ls_scf_post 1 4.0 0.000 0.000 3.392 3.398 make_images_sizes 222 9.7 0.000 0.000 0.891 3.364 mp_alltoall_i44 222 10.7 0.891 3.364 0.891 3.364 ls_scf_store_result 1 5.0 0.000 0.000 3.188 3.220 dbcsr_finalize 304 7.8 0.062 0.077 2.192 2.286 compute_matrix_preconditioner 1 6.0 0.001 0.002 2.275 2.283 dbcsr_data_new 4608 9.7 1.807 2.213 1.807 2.213 dbcsr_merge_all 275 8.9 0.478 0.529 2.051 2.121 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.088 2.089 rebuild_ks_matrix 3 7.3 0.000 0.000 2.055 2.056 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 2.055 2.056 qs_energies_init_hamiltonians 1 3.0 0.020 0.045 1.994 1.994 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=93.938000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3689.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/26e59941a7734f05bfa3a24a29949956890fc20f_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 8.876839E+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 0.072 0.094 113.528 113.529 qs_energies 1 2.0 0.000 0.000 112.012 112.021 ls_scf 1 3.0 0.000 0.000 109.024 109.032 dbcsr_multiply_generic 111 6.7 0.024 0.026 82.048 82.176 ls_scf_main 1 4.0 0.000 0.000 63.820 63.820 multiply_cannon 111 7.7 0.180 0.327 52.590 54.868 density_matrix_trs4 2 5.0 0.002 0.003 54.771 54.826 multiply_cannon_loop 111 8.7 0.098 0.099 49.383 52.565 ls_scf_init_scf 1 4.0 0.000 0.000 41.447 41.448 ls_scf_init_matrix_S 1 5.0 0.000 0.000 39.478 39.504 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 36.321 36.327 mp_waitall_1 4527 11.1 22.352 26.194 22.352 26.194 make_m2s 222 7.7 0.005 0.005 22.902 24.204 make_images 222 8.7 3.572 3.869 22.794 24.096 multiply_cannon_multrec 444 9.7 17.949 18.573 22.665 23.289 hybrid_alltoall_any 227 10.6 1.655 3.645 12.917 16.002 make_images_data 222 9.7 0.003 0.004 13.095 15.652 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.638 13.785 multiply_cannon_sync_h2d 444 9.7 8.791 8.839 8.791 8.839 compute_matrix_preconditioner 1 6.0 0.002 0.002 7.627 7.632 arnoldi_extremal 4 6.8 0.000 0.000 7.505 7.513 arnoldi_normal_ev 4 7.8 0.096 0.106 7.505 7.513 build_subspace 16 8.4 0.026 0.035 6.870 6.884 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.467 5.625 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.260 5.514 dbcsr_matrix_vector_mult_local 304 10.0 5.042 5.354 5.045 5.357 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.192 5.283 mp_sum_l 887 5.1 4.143 5.142 4.143 5.142 dbcsr_mm_accdrv_process 1814 10.4 0.376 0.506 4.534 4.699 mp_allgather_i34 111 8.7 1.314 4.307 1.314 4.307 dbcsr_mm_accdrv_process_sort 1814 11.4 4.125 4.254 4.125 4.254 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.588 4.240 ls_scf_post 1 4.0 0.000 0.000 3.757 3.765 acc_transpose_blocks 444 9.7 0.003 0.003 3.551 3.669 acc_transpose_blocks_kernels 444 10.7 0.006 0.006 3.468 3.588 jit_kernel_transpose 1 13.0 3.462 3.582 3.462 3.582 make_images_sizes 222 9.7 0.000 0.000 1.476 3.562 mp_alltoall_i44 222 10.7 1.476 3.561 1.476 3.561 ls_scf_store_result 1 5.0 0.000 0.000 3.481 3.516 calculate_norms 792 9.8 3.229 3.282 3.229 3.282 dbcsr_finalize 304 7.8 0.082 0.089 3.080 3.168 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.958 2.958 dbcsr_merge_all 275 8.9 0.885 0.912 2.863 2.945 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.883 2.885 dbcsr_complete_redistribute 5 7.6 1.427 1.466 2.730 2.846 rebuild_ks_matrix 3 7.3 0.000 0.000 2.816 2.818 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.816 2.818 matrix_ls_to_qs 2 6.0 0.000 0.000 2.386 2.519 dbcsr_sort_data 325 11.1 2.443 2.500 2.443 2.500 dbcsr_data_new 6591 9.6 1.840 2.372 1.840 2.372 dbcsr_new_transposed 4 7.5 0.246 0.259 2.293 2.303 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=113.529000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6960.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/26e59941a7734f05bfa3a24a29949956890fc20f_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 582.475776E+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 4805974. 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 0.048 0.167 235.122 235.124 qs_mol_dyn_low 1 2.0 0.010 0.066 233.653 233.695 qs_forces 5 3.8 0.005 0.043 233.092 233.111 qs_energies 5 4.8 0.004 0.027 230.082 230.126 scf_env_do_scf 5 5.8 0.001 0.011 213.047 213.049 scf_env_do_scf_inner_loop 105 6.6 0.002 0.006 186.072 186.074 qs_scf_new_mos 105 7.6 0.000 0.000 145.953 146.111 qs_scf_loop_do_ot 105 8.6 0.001 0.001 145.952 146.111 dbcsr_multiply_generic 1445 12.2 0.128 0.135 137.491 137.914 ot_scf_mini 105 9.6 0.003 0.004 136.027 136.198 multiply_cannon 1445 13.2 0.272 0.283 117.838 119.868 multiply_cannon_loop 1445 14.2 2.836 2.973 116.189 117.265 velocity_verlet 4 3.0 0.070 0.455 109.397 109.402 ot_mini 105 10.6 0.001 0.001 62.151 62.265 multiply_cannon_multrec 69360 15.2 29.681 34.879 40.223 45.836 qs_ot_get_p 112 10.4 0.001 0.003 43.722 43.997 mp_waitall_1 488190 16.1 36.744 43.789 36.744 43.789 qs_ot_get_derivative 55 11.6 0.001 0.001 40.444 40.579 multiply_cannon_metrocomm3 69360 15.2 0.198 0.209 26.892 34.211 multiply_cannon_sync_h2d 69360 15.2 28.930 33.062 28.930 33.062 qs_ot_p2m_diag 40 11.0 0.020 0.030 32.581 32.682 cp_dbcsr_syevd 40 12.0 0.002 0.002 29.413 29.413 rebuild_ks_matrix 110 8.4 0.000 0.001 28.915 29.078 qs_ks_build_kohn_sham_matrix 110 9.4 0.012 0.026 28.914 29.078 init_scf_loop 7 6.6 0.000 0.005 26.944 26.946 qs_ks_update_qs_env 112 7.6 0.001 0.001 26.520 26.665 cp_fm_syevd 40 13.0 0.000 0.001 24.317 24.447 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 22.951 23.206 apply_single 62 13.6 0.000 0.000 22.951 23.206 prepare_preconditioner 7 7.6 0.000 0.000 22.133 22.171 make_preconditioner 7 8.6 0.004 0.112 22.132 22.171 ot_new_cg_direction 55 11.6 0.001 0.001 20.959 20.959 cp_fm_redistribute_end 40 14.0 9.799 19.557 9.804 19.558 cp_fm_syevd_base 40 14.0 9.747 19.506 9.747 19.506 qs_rho_update_rho_low 110 7.6 0.000 0.001 16.968 17.392 calculate_rho_elec 110 8.6 0.029 0.032 16.968 17.392 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.985 15.186 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 15.052 15.161 make_full_inverse_cholesky 7 9.6 0.000 0.001 14.690 14.762 init_scf_run 5 5.8 0.000 0.001 14.237 14.238 scf_env_initial_rho_setup 5 6.8 0.001 0.003 14.237 14.238 mp_sum_l 4764 12.2 13.015 13.808 13.015 13.808 pw_transfer 1645 12.4 0.079 0.095 11.576 11.755 calculate_dm_sparse 110 9.5 0.000 0.001 11.566 11.755 density_rs2pw 110 9.6 0.005 0.006 11.167 11.624 fft_wrap_pw1pw2 1425 13.5 0.012 0.014 11.438 11.619 dbcsr_mm_accdrv_process 154766 15.8 6.249 6.431 10.410 11.479 qs_vxc_create 110 10.4 0.002 0.003 10.353 10.388 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 10.324 10.387 cp_fm_cholesky_invert 7 10.6 10.305 10.312 10.305 10.312 fft_wrap_pw1pw2_240 915 15.0 0.835 0.911 10.074 10.262 acc_transpose_blocks 69360 15.2 0.354 0.371 8.901 9.809 check_diag 80 13.5 8.686 8.940 9.498 9.637 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.001 9.615 9.633 calculate_first_density_matrix 1 7.0 0.001 0.027 9.221 9.232 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 8.770 8.828 sum_up_and_integrate 60 10.3 0.001 0.003 8.570 8.582 integrate_v_rspace 60 11.3 0.002 0.003 8.553 8.566 transfer_rs2pw 445 10.6 0.007 0.008 7.984 8.451 fft3d_pb 915 16.0 2.397 2.598 8.255 8.437 multiply_cannon_metrocomm1 69360 15.2 0.094 0.100 5.314 8.213 xc_rho_set_and_dset_create 110 12.4 0.076 0.095 7.288 7.531 make_full_single_inverse 7 9.6 0.001 0.003 7.169 7.207 acc_transpose_blocks_kernels 69360 16.2 0.857 0.900 6.143 6.994 xc_vxc_pw_create 60 11.3 0.038 0.050 6.934 6.968 make_m2s 2890 13.2 0.078 0.087 6.230 6.828 make_images 2890 14.2 0.237 0.259 6.122 6.718 xc_pw_derive 510 13.4 0.005 0.006 6.142 6.204 jit_kernel_transpose 5 15.0 5.287 6.111 5.287 6.111 mp_alltoall_z22v 2340 17.7 5.257 5.545 5.257 5.545 mp_waitany 7680 13.5 4.450 5.148 4.450 5.148 potential_pw2rs 60 12.3 0.002 0.003 4.806 4.824 wfi_extrapolate 5 7.8 0.010 0.203 4.738 4.738 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=235.124000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=553.600000, yerr=2.332381 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 26e59941a7734f05bfa3a24a29949956890fc20f Summary: empty Status: OK