=== 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: 77ac92570e643352348b65cb9881e203655a06a9 ################# 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 2.6.2, ELPA 2022.11.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.1.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.1, # SIRIUS 7.4.3, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Author: Matthias Krack (03.02.2023) # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.1.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.1 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib 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__HDF5 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 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_prefixed_scalapack.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/77ac92570e643352348b65cb9881e203655a06a9_performance_tests/01 job id: 45795269 --- 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/77ac92570e643352348b65cb9881e203655a06a9_performance_tests/02 job id: 45795270 --- Point --- name: 20 plot: h2o_32_ri_rpa_mp2 regex: Total MP2 Time= label: RI-MP2 (8n/6r/2t) --- Point --- name: 21 plot: h2o_32_ri_rpa_mp2_mem regex: Estimated peak process memory label: RI-MP2 (8n/6r/2t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/77ac92570e643352348b65cb9881e203655a06a9_performance_tests/03 job id: 45795271 --- Point --- name: 100 plot: h2o_64_md regex: CP2K label: (8n/12r/1t) --- Point --- name: 101 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/12r/1t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 6 # threads/rank = 2 nrepeat = 1 time[min] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/77ac92570e643352348b65cb9881e203655a06a9_performance_tests/04 job id: 45795272 --- Point --- name: 102 plot: h2o_64_md regex: CP2K label: (8n/6r/2t) --- Point --- name: 103 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/6r/2t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 4 # threads/rank = 3 nrepeat = 1 time[min] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/77ac92570e643352348b65cb9881e203655a06a9_performance_tests/05 job id: 45795273 --- Point --- name: 104 plot: h2o_64_md regex: CP2K label: (8n/4r/3t) --- Point --- name: 105 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/4r/3t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 3 # threads/rank = 4 nrepeat = 1 time[min] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/77ac92570e643352348b65cb9881e203655a06a9_performance_tests/06 job id: 45795274 --- Point --- name: 106 plot: h2o_64_md regex: CP2K label: (8n/3r/4t) --- Point --- name: 107 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/3r/4t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/77ac92570e643352348b65cb9881e203655a06a9_performance_tests/07 job id: 45795275 --- Point --- name: 108 plot: h2o_64_md regex: CP2K label: (8n/2r/6t) --- Point --- name: 109 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/2r/6t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 1 # threads/rank = 12 nrepeat = 1 time[min] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/77ac92570e643352348b65cb9881e203655a06a9_performance_tests/08 job id: 45795276 --- 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/77ac92570e643352348b65cb9881e203655a06a9_performance_tests/09 job id: 45795277 --- 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/77ac92570e643352348b65cb9881e203655a06a9_performance_tests/10 job id: 45795278 --- 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/77ac92570e643352348b65cb9881e203655a06a9_performance_tests/11 job id: 45795279 --- 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/77ac92570e643352348b65cb9881e203655a06a9_performance_tests/12 job id: 45795280 --- 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/77ac92570e643352348b65cb9881e203655a06a9_performance_tests/13 job id: 45795281 --- 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/77ac92570e643352348b65cb9881e203655a06a9_performance_tests/14 job id: 45795282 --- 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/77ac92570e643352348b65cb9881e203655a06a9_performance_tests/15 job id: 45795283 --- 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/77ac92570e643352348b65cb9881e203655a06a9_performance_tests/16 job id: 45795284 --- 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/77ac92570e643352348b65cb9881e203655a06a9_performance_tests/17 job id: 45795285 --- 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/77ac92570e643352348b65cb9881e203655a06a9_performance_tests/18 job id: 45795286 --- 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/77ac92570e643352348b65cb9881e203655a06a9_performance_tests/19 job id: 45795287 --- 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/77ac92570e643352348b65cb9881e203655a06a9_performance_tests/20 job id: 45795289 --- Point --- name: 410 plot: h2o_256_md regex: CP2K label: (8n/1r/12t) --- Point --- name: 411 plot: h2o_256_md_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/77ac92570e643352348b65cb9881e203655a06a9_performance_tests/21 job id: 45795291 --- Point --- name: 500 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/12r/1t) --- Point --- name: 501 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/12r/1t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 6 # threads/rank = 2 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/77ac92570e643352348b65cb9881e203655a06a9_performance_tests/22 job id: 45795293 --- Point --- name: 502 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/6r/2t) --- Point --- name: 503 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/6r/2t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 4 # threads/rank = 3 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/77ac92570e643352348b65cb9881e203655a06a9_performance_tests/23 job id: 45795295 --- Point --- name: 504 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/4r/3t) --- Point --- name: 505 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/4r/3t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 3 # threads/rank = 4 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/77ac92570e643352348b65cb9881e203655a06a9_performance_tests/24 job id: 45795298 --- Point --- name: 506 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/3r/4t) --- Point --- name: 507 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/3r/4t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/77ac92570e643352348b65cb9881e203655a06a9_performance_tests/25 job id: 45795299 --- Point --- name: 508 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/2r/6t) --- Point --- name: 509 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/2r/6t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 1 # threads/rank = 12 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/77ac92570e643352348b65cb9881e203655a06a9_performance_tests/26 job id: 45795301 --- Point --- name: 510 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/1r/12t) --- Point --- name: 511 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: 512 H2O (4 NVE MD steps on 64 nodes) input file: benchmarks/QS/00512_H2O/H2O-512_md.inp required files: [] output file: result.log # nodes = 64 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/77ac92570e643352348b65cb9881e203655a06a9_performance_tests/27 job id: 45795302 --- 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/77ac92570e643352348b65cb9881e203655a06a9_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 344 9. 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 0.024 0.035 141.919 141.920 farming_run 1 2.0 141.268 141.269 141.881 141.883 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.459261E+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 1103589. MP_Allreduce 491 2254389. 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 0.008 0.027 118.645 118.645 qs_energies 1 2.0 0.000 0.000 118.393 118.395 mp2_main 1 3.0 0.000 0.000 115.699 115.701 mp2_gpw_main 1 4.0 0.025 0.029 114.132 114.134 mp2_ri_gpw_compute_in 1 5.0 0.180 0.212 94.065 94.443 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 55.504 55.886 mp2_eri_3c_integrate_gpw 272 7.0 0.152 0.166 41.802 47.122 get_2c_integrals 1 6.0 0.000 0.000 37.420 38.379 integrate_v_rspace 273 8.0 0.441 0.456 25.186 30.353 pw_transfer 6555 10.6 0.374 0.393 27.387 27.765 grid_integrate_task_list 273 9.0 20.987 26.647 20.987 26.647 fft_wrap_pw1pw2 5465 11.4 0.046 0.048 26.039 26.421 fft_wrap_pw1pw2_100 2178 12.4 1.196 1.384 23.558 23.935 rpa_ri_compute_en 1 5.0 0.007 0.011 19.941 20.155 compute_2c_integrals 1 7.0 0.012 0.014 19.663 19.686 compute_2c_integrals_loop_lm 1 8.0 0.006 0.007 18.877 19.209 mp2_eri_2c_integrate_gpw 1 9.0 2.379 2.459 18.871 19.203 cp_fm_cholesky_decompose 12 8.2 17.993 18.941 17.993 18.941 cholesky_decomp 1 7.0 0.000 0.000 16.607 17.576 fft3d_s 5443 13.4 16.168 16.544 16.189 16.566 ao_to_mo_and_store_B_mult_1 272 7.0 10.866 15.591 10.866 15.591 calculate_wavefunction 272 8.0 5.422 5.584 12.543 13.157 rpa_num_int 1 6.0 0.001 0.008 11.102 11.111 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.771 11.110 calc_mat_Q 8 8.0 0.000 0.000 9.364 9.510 contract_S_to_Q 8 9.0 0.000 0.000 8.789 8.932 calc_potential_gpw 544 9.5 0.005 0.006 8.226 8.634 parallel_gemm_fm 14 9.1 0.000 0.000 8.380 8.466 parallel_gemm_fm_cosma 14 10.1 8.380 8.466 8.380 8.466 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.204 8.449 create_integ_mat 1 6.0 0.015 0.026 8.354 8.363 potential_pw2rs 545 10.0 0.107 0.109 7.714 8.320 collocate_single_gaussian 272 10.0 0.040 0.042 7.468 7.708 array2fm 1 7.0 0.000 0.000 6.826 7.368 pw_scatter_s 2720 13.7 4.410 4.633 4.410 4.633 pw_gather_s 2722 13.2 3.853 4.167 3.853 4.167 array2fm_buffer_send 1 8.0 3.042 3.169 3.042 3.169 scf_env_do_scf 1 3.0 0.000 0.000 2.408 2.408 pw_poisson_solve 545 10.5 1.093 1.134 2.148 2.392 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=114.132208, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2731.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/77ac92570e643352348b65cb9881e203655a06a9_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 205321. MP_Allreduce 344 10. 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 0.046 0.055 406.112 406.113 farming_run 1 2.0 404.524 404.528 406.046 406.050 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.230246E+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 743 386399. MP_Allreduce 1941 22272. MP_Sync 37 MP_Alltoall 77 9818196. 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 0.013 0.039 211.159 211.159 qs_energies 1 2.0 0.002 0.011 210.535 210.555 scf_env_do_scf 1 3.0 0.000 0.000 106.848 106.848 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.746 105.754 rebuild_ks_matrix 4 6.0 0.000 0.000 105.745 105.752 qs_ks_build_kohn_sham_matrix 4 7.0 0.059 0.066 105.745 105.752 hfx_ks_matrix 4 8.0 0.001 0.001 105.293 105.298 integrate_four_center 4 9.0 0.144 0.451 105.292 105.297 mp2_main 1 3.0 0.002 0.025 103.351 103.372 mp2_gpw_main 1 4.0 0.038 0.092 101.811 101.836 integrate_four_center_main 4 10.0 0.103 0.460 96.504 99.470 integrate_four_center_bin 266 11.0 96.401 99.311 96.401 99.311 init_scf_loop 1 4.0 0.000 0.000 92.403 92.403 mp2_ri_gpw_compute_in 1 5.0 0.066 0.080 75.036 76.092 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.513 55.591 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.160 42.083 47.349 integrate_v_rspace 95 8.0 0.400 0.575 28.524 33.638 pw_transfer 2240 10.6 0.146 0.159 29.888 30.350 fft_wrap_pw1pw2 1868 11.4 0.018 0.020 28.922 29.375 grid_integrate_task_list 95 9.0 23.750 29.048 23.750 29.048 mp2_ri_gpw_compute_en 1 5.0 0.057 0.082 26.590 28.303 fft_wrap_pw1pw2_100 730 12.4 1.254 1.469 26.608 27.062 ao_to_mo_and_store_B_mult_1 91 7.0 10.739 26.714 10.739 26.714 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.829 1.894 24.854 24.863 get_2c_integrals 1 6.0 0.000 0.000 20.364 20.465 compute_2c_integrals 1 7.0 0.004 0.014 19.334 19.347 compute_2c_integrals_loop_lm 1 8.0 0.001 0.002 18.773 19.175 mp2_eri_2c_integrate_gpw 1 9.0 1.742 1.819 18.772 19.174 fft3d_s 1823 13.4 18.395 18.749 18.408 18.762 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.443 14.443 calculate_wavefunction 91 8.0 2.019 2.043 9.743 9.996 mp2_ri_gpw_compute_en_expansio 172 7.0 0.558 0.592 8.716 9.571 potential_pw2rs 186 10.0 0.033 0.035 8.659 9.283 local_gemm 172 8.0 8.158 8.998 8.158 8.998 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.192 8.670 mp2_ri_gpw_compute_en_comm 22 7.0 0.506 0.527 7.911 8.279 calc_potential_gpw 182 9.5 0.002 0.002 7.895 8.131 collocate_single_gaussian 91 10.0 0.017 0.023 7.837 8.062 mp2_ri_gpw_compute_en_ener 172 7.0 6.353 6.419 6.353 6.419 mp_sendrecv_dm3 2068 8.0 5.930 6.293 5.930 6.293 mp_sync 37 10.5 3.478 6.218 3.478 6.218 pw_gather_s 912 13.2 4.915 5.387 4.915 5.387 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.797749, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1517.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/77ac92570e643352348b65cb9881e203655a06a9_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 452.190208E+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 62385. MP_Allreduce 10249 272. MP_Sync 530 MP_Alltoall 2083 MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.029 0.071 60.987 61.012 qs_mol_dyn_low 1 2.0 0.003 0.003 60.617 60.623 qs_forces 11 3.9 0.003 0.003 60.552 60.553 qs_energies 11 4.9 0.001 0.001 58.915 58.929 scf_env_do_scf 11 5.9 0.000 0.001 51.756 51.756 scf_env_do_scf_inner_loop 108 6.5 0.002 0.023 49.133 49.133 qs_scf_new_mos 108 7.5 0.000 0.001 36.884 37.180 qs_scf_loop_do_ot 108 8.5 0.000 0.001 36.883 37.179 dbcsr_multiply_generic 2286 12.5 0.101 0.104 35.686 36.112 ot_scf_mini 108 9.5 0.002 0.002 35.208 35.410 multiply_cannon 2286 13.5 0.190 0.197 27.170 28.805 velocity_verlet 10 3.0 0.001 0.001 28.501 28.502 multiply_cannon_loop 2286 14.5 1.460 1.559 26.131 27.821 ot_mini 108 10.5 0.001 0.001 20.586 20.823 qs_ot_get_derivative 108 11.5 0.001 0.001 17.571 17.755 mp_waitall_1 245248 16.5 9.616 16.113 9.616 16.113 multiply_cannon_metrocomm3 54864 15.5 0.069 0.074 6.250 13.432 multiply_cannon_multrec 54864 15.5 4.244 6.626 7.987 11.730 qs_ot_get_p 119 10.4 0.001 0.001 9.726 10.034 rebuild_ks_matrix 119 8.3 0.000 0.000 9.187 9.363 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 9.187 9.363 qs_ks_update_qs_env 119 7.6 0.001 0.001 8.125 8.286 mp_sum_l 7207 12.9 5.755 7.487 5.755 7.487 multiply_cannon_sync_h2d 54864 15.5 5.911 6.852 5.911 6.852 qs_ot_p2m_diag 50 11.0 0.004 0.006 6.464 6.530 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.906 6.310 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 6.188 6.306 init_scf_run 11 5.9 0.000 0.001 5.818 5.819 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.818 5.819 qs_rho_update_rho_low 119 7.7 0.001 0.001 5.486 5.631 calculate_rho_elec 119 8.7 0.012 0.017 5.485 5.630 cp_dbcsr_syevd 50 12.0 0.002 0.003 5.541 5.541 sum_up_and_integrate 119 10.3 0.012 0.014 5.511 5.519 integrate_v_rspace 119 11.3 0.003 0.003 5.499 5.508 dbcsr_mm_accdrv_process 76910 16.1 1.134 1.808 3.664 5.171 cp_fm_diag_elpa 50 13.0 0.000 0.001 5.133 5.134 cp_fm_redistribute_end 50 14.0 2.597 5.066 2.614 5.071 cp_fm_diag_elpa_base 50 14.0 2.446 4.916 2.454 4.928 rs_pw_transfer 974 11.9 0.011 0.013 3.923 4.082 density_rs2pw 119 9.7 0.004 0.004 3.325 3.439 calculate_first_density_matrix 1 7.0 0.000 0.000 3.352 3.358 jit_kernel_multiply 13 15.8 2.469 3.310 2.469 3.310 calculate_dm_sparse 119 9.5 0.000 0.001 3.095 3.256 multiply_cannon_metrocomm1 54864 15.5 0.053 0.057 1.993 3.209 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.870 3.136 apply_single 119 13.6 0.000 0.000 2.870 3.136 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.992 2.993 pw_transfer 1439 11.6 0.052 0.058 2.713 2.823 potential_pw2rs 119 12.3 0.004 0.004 2.760 2.819 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 2.638 2.749 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.628 2.713 ot_diis_step 108 11.5 0.006 0.006 2.712 2.713 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.551 2.608 init_scf_loop 11 6.9 0.000 0.000 2.583 2.584 make_m2s 4572 13.5 0.053 0.056 2.461 2.554 make_images 4572 14.5 0.133 0.139 2.379 2.471 wfi_extrapolate 11 7.9 0.001 0.001 2.390 2.390 fft3d_ps 1201 14.6 0.378 0.490 2.287 2.384 acc_transpose_blocks 54864 15.5 0.228 0.257 1.819 2.376 mp_alltoall_d11v 2130 13.8 1.846 2.138 1.846 2.138 fft_wrap_pw1pw2_140 487 13.2 0.179 0.198 2.014 2.124 grid_integrate_task_list 119 12.3 2.005 2.112 2.005 2.112 mp_sum_d 4135 12.0 1.407 2.001 1.407 2.001 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.601 1.626 mp_waitany 12084 13.8 1.315 1.494 1.315 1.494 mp_alltoall_z22v 1201 16.6 1.360 1.431 1.360 1.431 grid_collocate_task_list 119 9.7 1.293 1.387 1.293 1.387 make_images_sizes 4572 15.5 0.004 0.005 1.111 1.370 mp_alltoall_i44 4572 16.5 1.107 1.366 1.107 1.366 prepare_preconditioner 11 7.9 0.000 0.000 1.272 1.301 make_preconditioner 11 8.9 0.000 0.000 1.272 1.301 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.204 1.243 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=61.012000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=431.090909, yerr=1.378705 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/77ac92570e643352348b65cb9881e203655a06a9_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 487.837696E+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 62664. MP_Allreduce 10226 305. MP_Sync 54 MP_Alltoall 2060 1221641. MP_SendRecv 16779 37093. MP_ISendRecv 16779 37093. MP_Wait 23539 MP_comm_split 50 MP_ISend 5720 128509. MP_IRecv 5720 128509. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.028 41.026 41.026 qs_mol_dyn_low 1 2.0 0.003 0.003 40.778 40.782 qs_forces 11 3.9 0.003 0.003 40.714 40.715 qs_energies 11 4.9 0.001 0.001 39.016 39.020 scf_env_do_scf 11 5.9 0.000 0.001 32.726 32.726 scf_env_do_scf_inner_loop 108 6.5 0.004 0.012 30.135 30.136 dbcsr_multiply_generic 2286 12.5 0.101 0.103 22.531 22.880 qs_scf_new_mos 108 7.5 0.001 0.001 20.630 20.869 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.630 20.868 ot_scf_mini 108 9.5 0.002 0.003 19.738 19.910 velocity_verlet 10 3.0 0.001 0.001 18.594 18.595 multiply_cannon 2286 13.5 0.207 0.216 17.004 18.559 multiply_cannon_loop 2286 14.5 0.893 0.976 15.870 17.610 ot_mini 108 10.5 0.001 0.001 12.345 12.578 mp_waitall_1 200699 16.5 6.006 11.501 6.006 11.501 qs_ot_get_derivative 108 11.5 0.001 0.001 9.890 10.067 multiply_cannon_metrocomm3 27432 15.5 0.067 0.070 4.324 9.793 multiply_cannon_multrec 27432 15.5 1.982 4.528 6.258 9.262 rebuild_ks_matrix 119 8.3 0.000 0.000 7.447 7.587 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.446 7.587 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.600 6.727 dbcsr_mm_accdrv_process 47894 16.0 3.141 5.300 4.206 6.238 init_scf_run 11 5.9 0.000 0.001 5.085 5.086 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.085 5.085 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.720 5.008 qs_ot_get_p 119 10.4 0.001 0.001 4.520 4.748 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.481 4.616 apply_single 119 13.6 0.000 0.000 3.481 4.616 mp_sum_l 7207 12.9 2.497 4.509 2.497 4.509 sum_up_and_integrate 119 10.3 0.024 0.027 4.392 4.421 integrate_v_rspace 119 11.3 0.002 0.003 4.367 4.396 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.967 4.061 calculate_rho_elec 119 8.7 0.021 0.024 3.966 4.061 calculate_first_density_matrix 1 7.0 0.000 0.000 3.527 3.531 rs_pw_transfer 974 11.9 0.010 0.011 2.861 3.247 qs_ot_p2m_diag 50 11.0 0.009 0.013 3.018 3.037 multiply_cannon_sync_h2d 27432 15.5 2.203 2.926 2.203 2.926 make_m2s 4572 13.5 0.052 0.054 2.584 2.816 make_images 4572 14.5 0.201 0.239 2.496 2.726 density_rs2pw 119 9.7 0.004 0.004 2.286 2.695 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.582 2.582 init_scf_loop 11 6.9 0.000 0.000 2.570 2.570 calculate_dm_sparse 119 9.5 0.000 0.001 2.443 2.521 ot_diis_step 108 11.5 0.011 0.011 2.407 2.407 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.140 2.230 cp_fm_diag_elpa 50 13.0 0.001 0.001 2.220 2.221 cp_fm_redistribute_end 50 14.0 1.125 2.193 1.128 2.195 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.178 2.180 potential_pw2rs 119 12.3 0.006 0.006 2.111 2.147 cp_fm_diag_elpa_base 50 14.0 1.035 2.102 1.063 2.139 jit_kernel_multiply 8 16.2 1.012 2.072 1.012 2.072 pw_transfer 1439 11.6 0.065 0.070 2.018 2.063 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.965 2.007 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.926 1.973 grid_integrate_task_list 119 12.3 1.832 1.921 1.832 1.921 acc_transpose_blocks 27432 15.5 0.109 0.114 1.282 1.850 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.658 1.668 make_images_data 4572 15.5 0.044 0.051 1.177 1.605 prepare_preconditioner 11 7.9 0.000 0.000 1.553 1.578 make_preconditioner 11 8.9 0.000 0.000 1.553 1.578 fft_wrap_pw1pw2_140 487 13.2 0.201 0.211 1.505 1.551 fft3d_ps 1201 14.6 0.522 0.578 1.479 1.521 hybrid_alltoall_any 4725 16.4 0.051 0.112 1.019 1.518 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.448 1.505 wfi_extrapolate 11 7.9 0.001 0.001 1.504 1.504 mp_alltoall_d11v 2130 13.8 1.235 1.384 1.235 1.384 grid_collocate_task_list 119 9.7 1.228 1.366 1.228 1.366 mp_allgather_i34 2286 14.5 0.559 1.344 0.559 1.344 acc_transpose_blocks_kernels 27432 16.5 0.181 0.270 0.753 1.236 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.174 1.221 mp_sum_d 4135 12.0 0.583 1.039 0.583 1.039 parallel_gemm_fm 81 9.0 0.000 0.000 1.001 1.026 parallel_gemm_fm_cosma 81 10.0 1.001 1.026 1.001 1.026 jit_kernel_transpose 5 15.5 0.572 0.970 0.572 0.970 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.930 0.931 rs_pw_transfer_RS2PW_140 130 11.5 0.138 0.146 0.532 0.923 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.903 0.916 make_basis_sm 11 9.8 0.000 0.000 0.909 0.911 mp_waitany 5720 13.7 0.518 0.906 0.518 0.906 make_images_sizes 4572 15.5 0.005 0.005 0.592 0.851 rs_pw_transfer_PW2RS_50 119 14.3 0.589 0.609 0.792 0.847 mp_alltoall_i44 4572 16.5 0.587 0.847 0.587 0.847 mp_alltoall_z22v 1201 16.6 0.749 0.831 0.749 0.831 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=41.026000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=464.181818, yerr=1.266217 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/77ac92570e643352348b65cb9881e203655a06a9_performance_tests/05/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 59.051995E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3143552 0.0% 0.0% 100.0% average stack size 0.0 0.0 46.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 522.690560E+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 63497. MP_Allreduce 10075 307. MP_Sync 54 MP_Alltoall 1821 860297. 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 0.110 0.213 42.177 42.194 qs_mol_dyn_low 1 2.0 0.003 0.004 41.381 41.389 qs_forces 11 3.9 0.002 0.003 41.273 41.274 qs_energies 11 4.9 0.012 0.054 39.471 39.480 scf_env_do_scf 11 5.9 0.004 0.037 32.965 32.967 scf_env_do_scf_inner_loop 108 6.5 0.010 0.037 29.150 29.153 dbcsr_multiply_generic 2286 12.5 0.096 0.098 19.699 20.013 velocity_verlet 10 3.0 0.001 0.001 18.796 18.799 qs_scf_new_mos 108 7.5 0.003 0.009 18.570 18.593 qs_scf_loop_do_ot 108 8.5 0.002 0.007 18.567 18.593 ot_scf_mini 108 9.5 0.009 0.030 17.703 17.726 multiply_cannon 2286 13.5 0.196 0.203 15.085 15.832 multiply_cannon_loop 2286 14.5 0.633 0.665 13.993 14.808 ot_mini 108 10.5 0.001 0.001 10.424 10.465 qs_ot_get_derivative 108 11.5 0.001 0.001 8.722 8.748 multiply_cannon_multrec 18288 15.5 1.913 2.784 7.655 8.119 rebuild_ks_matrix 119 8.3 0.000 0.000 7.865 7.963 qs_ks_build_kohn_sham_matrix 119 9.3 0.014 0.027 7.865 7.963 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.942 7.031 dbcsr_mm_accdrv_process 38222 16.0 4.588 6.463 5.660 6.691 mp_waitall_1 158411 16.6 4.530 6.093 4.530 6.093 sum_up_and_integrate 119 10.3 0.030 0.031 5.224 5.253 integrate_v_rspace 119 11.3 0.003 0.003 5.193 5.223 init_scf_run 11 5.9 0.000 0.001 5.129 5.129 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.129 5.129 qs_ot_get_p 119 10.4 0.003 0.012 5.072 5.114 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.887 4.910 calculate_rho_elec 119 8.7 0.031 0.032 4.886 4.910 rs_pw_transfer 974 11.9 0.009 0.010 4.164 4.368 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.416 4.060 init_scf_loop 11 6.9 0.010 0.043 3.761 3.764 multiply_cannon_metrocomm3 18288 15.5 0.044 0.045 2.038 3.620 density_rs2pw 119 9.7 0.006 0.018 3.299 3.549 calculate_first_density_matrix 1 7.0 0.002 0.008 3.447 3.451 jit_kernel_multiply 13 15.9 1.020 3.272 1.020 3.272 qs_ot_p2m_diag 50 11.0 0.012 0.013 3.228 3.234 potential_pw2rs 119 12.3 0.007 0.009 2.914 2.955 cp_dbcsr_syevd 50 12.0 0.007 0.034 2.922 2.923 calculate_dm_sparse 119 9.5 0.000 0.001 2.758 2.775 make_m2s 4572 13.5 0.044 0.045 2.498 2.758 make_images 4572 14.5 0.191 0.204 2.413 2.673 pw_transfer 1439 11.6 0.066 0.069 2.610 2.672 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.517 2.577 cp_fm_diag_elpa 50 13.0 0.012 0.019 2.531 2.543 cp_fm_diag_elpa_base 50 14.0 2.471 2.486 2.498 2.507 prepare_preconditioner 11 7.9 0.000 0.001 2.428 2.431 make_preconditioner 11 8.9 0.001 0.005 2.428 2.431 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.112 2.410 apply_single 119 13.6 0.000 0.000 2.112 2.410 make_full_inverse_cholesky 11 9.9 0.002 0.013 2.230 2.339 mp_sum_l 7207 12.9 1.762 2.200 1.762 2.200 fft3d_ps 1201 14.6 0.539 0.578 2.007 2.074 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.066 2.068 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.033 2.043 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 2.003 2.016 grid_integrate_task_list 119 12.3 1.786 1.891 1.786 1.891 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.829 1.864 fft_wrap_pw1pw2_140 487 13.2 0.250 0.260 1.717 1.777 ot_diis_step 108 11.5 0.011 0.011 1.618 1.619 multiply_cannon_sync_h2d 18288 15.5 1.317 1.516 1.317 1.516 make_images_data 4572 15.5 0.044 0.048 1.130 1.515 wfi_extrapolate 11 7.9 0.001 0.001 1.506 1.506 acc_transpose_blocks 18288 15.5 0.076 0.077 1.428 1.450 grid_collocate_task_list 119 9.7 1.213 1.356 1.213 1.356 mp_alltoall_z22v 1201 16.6 1.228 1.305 1.228 1.305 hybrid_alltoall_any 4725 16.4 0.056 0.114 0.950 1.246 mp_alltoall_d11v 2130 13.8 1.023 1.212 1.023 1.212 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.113 1.195 mp_sendrecv_dv 11067 12.7 1.158 1.187 1.158 1.187 cp_fm_cholesky_invert 11 10.9 1.126 1.136 1.126 1.136 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.007 1.019 mp_allgather_i34 2286 14.5 0.552 1.008 0.552 1.008 acc_transpose_blocks_kernels 18288 16.5 0.210 0.219 0.988 0.998 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 0.996 0.998 arnoldi_extremal 119 11.4 0.013 0.048 0.973 0.997 arnoldi_normal_ev 119 12.4 0.014 0.075 0.960 0.995 rs_pw_transfer_PW2RS_50 119 14.3 0.401 0.412 0.908 0.992 make_images_sizes 4572 15.5 0.005 0.005 0.685 0.955 mp_alltoall_i44 4572 16.5 0.681 0.950 0.681 0.950 mp_sum_d 4135 12.0 0.660 0.931 0.660 0.931 mp_waitany 9880 13.7 0.650 0.909 0.650 0.909 rs_pw_transfer_RS2PW_50 119 11.7 0.244 0.266 0.833 0.899 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=42.194000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=497.000000, yerr=1.906925 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/77ac92570e643352348b65cb9881e203655a06a9_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 560.386048E+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 63496. MP_Allreduce 10074 349. 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 0.053 0.143 42.882 42.884 qs_mol_dyn_low 1 2.0 0.003 0.003 42.216 42.224 qs_forces 11 3.9 0.002 0.003 42.154 42.155 qs_energies 11 4.9 0.003 0.012 40.316 40.320 scf_env_do_scf 11 5.9 0.000 0.001 33.682 33.683 scf_env_do_scf_inner_loop 108 6.5 0.032 0.038 29.502 29.503 dbcsr_multiply_generic 2286 12.5 0.099 0.102 21.085 21.198 velocity_verlet 10 3.0 0.001 0.001 20.171 20.172 qs_scf_new_mos 108 7.5 0.001 0.001 19.418 19.474 qs_scf_loop_do_ot 108 8.5 0.001 0.001 19.418 19.473 ot_scf_mini 108 9.5 0.002 0.003 18.424 18.471 multiply_cannon 2286 13.5 0.226 0.270 16.555 17.370 multiply_cannon_loop 2286 14.5 0.933 0.961 15.379 16.018 ot_mini 108 10.5 0.001 0.001 10.882 10.944 multiply_cannon_multrec 27432 15.5 2.443 3.210 9.895 10.583 qs_ot_get_derivative 108 11.5 0.001 0.001 9.010 9.059 dbcsr_mm_accdrv_process 47916 15.9 6.052 8.081 7.356 9.007 rebuild_ks_matrix 119 8.3 0.000 0.000 7.626 7.674 qs_ks_build_kohn_sham_matrix 119 9.3 0.026 0.077 7.626 7.674 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.789 6.832 init_scf_run 11 5.9 0.000 0.001 5.170 5.170 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.169 5.170 qs_ot_get_p 119 10.4 0.001 0.001 4.829 4.912 sum_up_and_integrate 119 10.3 0.035 0.037 4.627 4.635 integrate_v_rspace 119 11.3 0.003 0.003 4.592 4.600 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.447 4.476 calculate_rho_elec 119 8.7 0.040 0.046 4.446 4.475 init_scf_loop 11 6.9 0.002 0.013 4.151 4.151 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.714 4.088 mp_waitall_1 137007 16.6 3.064 3.853 3.064 3.853 calculate_first_density_matrix 1 7.0 0.000 0.000 3.534 3.535 jit_kernel_multiply 10 15.7 1.244 3.355 1.244 3.355 rs_pw_transfer 974 11.9 0.009 0.009 3.143 3.341 qs_ot_p2m_diag 50 11.0 0.015 0.023 3.167 3.179 prepare_preconditioner 11 7.9 0.000 0.001 3.038 3.049 make_preconditioner 11 8.9 0.002 0.014 3.038 3.049 make_m2s 4572 13.5 0.054 0.056 2.888 3.023 density_rs2pw 119 9.7 0.004 0.004 2.801 2.986 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.627 2.968 calculate_dm_sparse 119 9.5 0.000 0.000 2.905 2.951 make_images 4572 14.5 0.271 0.329 2.780 2.914 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.795 2.795 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.093 2.603 apply_single 119 13.6 0.000 0.000 2.093 2.603 pw_transfer 1439 11.6 0.066 0.070 2.526 2.559 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.433 2.468 cp_fm_diag_elpa 50 13.0 0.013 0.013 2.418 2.426 cp_fm_diag_elpa_base 50 14.0 2.360 2.377 2.402 2.410 potential_pw2rs 119 12.3 0.009 0.009 2.283 2.297 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.224 2.252 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.149 2.150 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 2.042 2.050 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 1.102 1.935 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 1.916 1.929 grid_integrate_task_list 119 12.3 1.818 1.927 1.818 1.927 fft_wrap_pw1pw2_140 487 13.2 0.286 0.299 1.882 1.920 fft3d_ps 1201 14.6 0.560 0.609 1.874 1.904 mp_sum_l 7207 12.9 1.242 1.849 1.242 1.849 ot_diis_step 108 11.5 0.012 0.012 1.831 1.832 acc_transpose_blocks 27432 15.5 0.113 0.116 1.698 1.826 wfi_extrapolate 11 7.9 0.001 0.001 1.561 1.561 make_images_data 4572 15.5 0.045 0.048 1.196 1.381 grid_collocate_task_list 119 9.7 1.221 1.346 1.221 1.346 hybrid_alltoall_any 4725 16.4 0.062 0.152 1.019 1.247 dbcsr_complete_redistribute 329 12.2 0.121 0.149 0.962 1.237 mp_alltoall_z22v 1201 16.6 1.199 1.230 1.199 1.230 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.200 1.215 cp_fm_upper_to_full 72 14.2 0.844 1.201 0.844 1.201 acc_transpose_blocks_kernels 27432 16.5 0.267 0.274 1.074 1.199 mp_alltoall_d11v 2130 13.8 1.060 1.161 1.060 1.161 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.128 1.129 multiply_cannon_sync_h2d 27432 15.5 0.965 1.038 0.965 1.038 cp_fm_cholesky_invert 11 10.9 1.011 1.017 1.011 1.017 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.932 0.937 jit_kernel_transpose 5 15.6 0.807 0.935 0.807 0.935 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.658 0.929 make_images_sizes 4572 15.5 0.005 0.005 0.675 0.925 mp_alltoall_i44 4572 16.5 0.670 0.920 0.670 0.920 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.794 0.872 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=42.884000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=531.545455, yerr=4.142104 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/77ac92570e643352348b65cb9881e203655a06a9_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 609.513472E+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 63495. MP_Allreduce 10074 348. 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 0.012 0.045 30.739 30.750 qs_mol_dyn_low 1 2.0 0.003 0.003 30.364 30.368 qs_forces 11 3.9 0.006 0.011 30.296 30.297 qs_energies 11 4.9 0.001 0.001 28.552 28.558 scf_env_do_scf 11 5.9 0.000 0.001 22.341 22.341 scf_env_do_scf_inner_loop 108 6.5 0.021 0.062 19.805 19.806 velocity_verlet 10 3.0 0.001 0.001 14.699 14.702 dbcsr_multiply_generic 2286 12.5 0.091 0.095 13.317 13.410 qs_scf_new_mos 108 7.5 0.001 0.001 11.619 11.662 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.618 11.661 multiply_cannon 2286 13.5 0.230 0.236 10.687 11.123 ot_scf_mini 108 9.5 0.002 0.002 10.930 10.975 multiply_cannon_loop 2286 14.5 0.328 0.338 9.760 9.942 multiply_cannon_multrec 9144 15.5 1.943 2.150 6.729 6.991 ot_mini 108 10.5 0.001 0.001 6.167 6.216 rebuild_ks_matrix 119 8.3 0.000 0.000 6.072 6.092 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.072 6.092 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.403 5.421 qs_ot_get_derivative 108 11.5 0.001 0.001 4.877 4.921 dbcsr_mm_accdrv_process 12550 15.8 3.492 4.651 4.685 4.831 init_scf_run 11 5.9 0.000 0.001 4.766 4.766 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.765 4.766 sum_up_and_integrate 119 10.3 0.038 0.042 3.735 3.741 integrate_v_rspace 119 11.3 0.003 0.003 3.698 3.705 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.631 3.637 calculate_rho_elec 119 8.7 0.060 0.061 3.630 3.636 calculate_first_density_matrix 1 7.0 0.000 0.000 3.571 3.572 qs_ot_get_p 119 10.4 0.001 0.001 2.969 3.010 init_scf_loop 11 6.9 0.000 0.000 2.514 2.515 jit_kernel_multiply 10 15.7 1.154 2.469 1.154 2.469 calculate_dm_sparse 119 9.5 0.000 0.000 2.152 2.170 mp_waitall_1 115863 16.7 1.635 2.162 1.635 2.162 density_rs2pw 119 9.7 0.004 0.004 1.952 2.091 pw_transfer 1439 11.6 0.066 0.069 2.076 2.086 qs_ot_p2m_diag 50 11.0 0.022 0.023 2.001 2.004 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.982 1.995 rs_pw_transfer 974 11.9 0.008 0.008 1.835 1.961 grid_integrate_task_list 119 12.3 1.849 1.958 1.849 1.958 make_m2s 4572 13.5 0.034 0.035 1.786 1.928 make_images 4572 14.5 0.268 0.302 1.698 1.838 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.819 1.820 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.800 1.801 prepare_preconditioner 11 7.9 0.000 0.000 1.757 1.760 make_preconditioner 11 8.9 0.000 0.000 1.757 1.760 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.657 1.694 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.639 1.671 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.644 1.656 fft_wrap_pw1pw2_140 487 13.2 0.364 0.374 1.616 1.629 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.561 1.567 cp_fm_diag_elpa 50 13.0 0.136 0.138 1.506 1.514 potential_pw2rs 119 12.3 0.010 0.011 1.458 1.461 grid_collocate_task_list 119 9.7 1.273 1.394 1.273 1.394 cp_fm_diag_elpa_base 50 14.0 1.340 1.358 1.367 1.376 fft3d_ps 1201 14.6 0.558 0.571 1.316 1.325 ot_diis_step 108 11.5 0.013 0.013 1.276 1.276 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.227 1.232 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.152 1.180 apply_single 119 13.6 0.000 0.000 1.152 1.180 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.153 1.168 wfi_extrapolate 11 7.9 0.001 0.001 1.145 1.146 parallel_gemm_fm 81 9.0 0.000 0.000 1.083 1.084 parallel_gemm_fm_cosma 81 10.0 1.083 1.084 1.083 1.084 make_basis_sm 11 9.8 0.000 0.000 1.047 1.048 hybrid_alltoall_any 4725 16.4 0.063 0.176 0.805 1.035 make_images_data 4572 15.5 0.038 0.042 0.848 1.024 acc_transpose_blocks 9144 15.5 0.038 0.039 0.961 0.975 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.869 0.924 mp_alltoall_d11v 2130 13.8 0.792 0.922 0.792 0.922 cp_fm_cholesky_invert 11 10.9 0.890 0.893 0.890 0.893 multiply_cannon_sync_h2d 9144 15.5 0.714 0.809 0.714 0.809 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.792 0.794 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.742 0.750 acc_transpose_blocks_kernels 9144 16.5 0.116 0.119 0.723 0.734 qs_env_update_s_mstruct 11 6.9 0.001 0.003 0.679 0.728 mp_allgather_i34 2286 14.5 0.260 0.695 0.260 0.695 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.341 0.676 mp_alltoall_z22v 1201 16.6 0.633 0.658 0.633 0.658 jit_kernel_transpose 5 15.6 0.607 0.618 0.607 0.618 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=30.750000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=575.909091, yerr=7.391082 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/77ac92570e643352348b65cb9881e203655a06a9_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 767.148032E+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 63729. MP_Allreduce 10074 433. 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 0.032 0.132 49.218 49.290 qs_mol_dyn_low 1 2.0 0.003 0.003 48.864 48.871 qs_forces 11 3.9 0.002 0.003 48.799 48.799 qs_energies 11 4.9 0.001 0.001 46.661 46.668 scf_env_do_scf 11 5.9 0.001 0.001 38.766 38.766 scf_env_do_scf_inner_loop 108 6.5 0.004 0.015 29.827 29.827 velocity_verlet 10 3.0 0.001 0.001 26.288 26.293 dbcsr_multiply_generic 2286 12.5 0.101 0.103 20.771 21.163 qs_scf_new_mos 108 7.5 0.001 0.001 18.635 18.740 qs_scf_loop_do_ot 108 8.5 0.001 0.001 18.634 18.739 ot_scf_mini 108 9.5 0.002 0.002 17.448 17.554 multiply_cannon 2286 13.5 0.306 0.321 16.036 17.456 multiply_cannon_loop 2286 14.5 0.344 0.350 14.694 15.833 ot_mini 108 10.5 0.001 0.001 10.397 10.544 multiply_cannon_multrec 9144 15.5 3.501 5.111 9.892 10.306 init_scf_loop 11 6.9 0.000 0.000 8.901 8.902 rebuild_ks_matrix 119 8.3 0.000 0.000 8.232 8.421 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 8.232 8.421 qs_ot_get_derivative 108 11.5 0.001 0.001 8.223 8.329 dbcsr_mm_accdrv_process 12550 15.8 4.988 6.611 6.266 7.969 prepare_preconditioner 11 7.9 0.000 0.000 7.742 7.759 make_preconditioner 11 8.9 0.000 0.000 7.742 7.759 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.291 7.617 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.440 7.608 init_scf_run 11 5.9 0.000 0.001 5.763 5.763 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.763 5.763 mp_waitall_1 94719 16.7 3.798 5.036 3.798 5.036 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.997 5.014 calculate_rho_elec 119 8.7 0.118 0.121 4.996 5.013 sum_up_and_integrate 119 10.3 0.064 0.065 4.664 4.669 integrate_v_rspace 119 11.3 0.003 0.003 4.600 4.605 cp_fm_upper_to_full 72 14.2 3.172 4.546 3.172 4.546 qs_ot_get_p 119 10.4 0.001 0.001 4.120 4.274 calculate_first_density_matrix 1 7.0 0.000 0.000 4.013 4.016 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.334 3.755 calculate_dm_sparse 119 9.5 0.000 0.000 3.450 3.466 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 2.173 3.350 pw_transfer 1439 11.6 0.069 0.070 3.174 3.188 jit_kernel_multiply 10 15.6 1.250 3.147 1.250 3.147 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 3.075 3.090 dbcsr_complete_redistribute 329 12.2 0.347 0.366 2.228 3.044 make_m2s 4572 13.5 0.038 0.038 2.703 2.874 density_rs2pw 119 9.7 0.004 0.004 2.823 2.845 make_images 4572 14.5 0.352 0.382 2.582 2.754 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.257 2.656 apply_single 119 13.6 0.000 0.000 2.256 2.656 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.829 2.639 mp_sum_l 7207 12.9 1.508 2.629 1.508 2.629 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.564 2.569 fft_wrap_pw1pw2_140 487 13.2 0.616 0.620 2.522 2.538 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 2.429 2.441 mp_alltoall_i22 627 13.8 1.483 2.327 1.483 2.327 rs_pw_transfer 974 11.9 0.009 0.009 2.269 2.300 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.251 2.252 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.446 2.238 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.206 2.207 grid_integrate_task_list 119 12.3 2.106 2.149 2.106 2.149 cp_fm_cholesky_invert 11 10.9 2.121 2.124 2.121 2.124 fft3d_ps 1201 14.6 0.599 0.609 2.092 2.107 ot_diis_step 108 11.5 0.014 0.015 2.104 2.105 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.997 2.048 potential_pw2rs 119 12.3 0.014 0.014 1.916 1.926 cp_fm_diag_elpa 50 13.0 0.016 0.024 1.892 1.893 cp_fm_diag_elpa_base 50 14.0 1.716 1.765 1.866 1.867 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.805 1.807 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.776 1.801 wfi_extrapolate 11 7.9 0.001 0.001 1.671 1.671 make_images_data 4572 15.5 0.043 0.045 1.334 1.584 grid_collocate_task_list 119 9.7 1.524 1.553 1.524 1.553 hybrid_alltoall_any 4725 16.4 0.087 0.148 1.347 1.552 mp_alltoall_d11v 2130 13.8 1.358 1.394 1.358 1.394 mp_alltoall_z22v 1201 16.6 1.357 1.390 1.357 1.390 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.260 1.310 acc_transpose_blocks 9144 15.5 0.039 0.039 1.074 1.179 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.139 1.153 multiply_cannon_sync_h2d 9144 15.5 1.048 1.052 1.048 1.052 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 0.994 1.032 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.009 1.028 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=49.290000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=720.727273, yerr=14.447969 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/77ac92570e643352348b65cb9881e203655a06a9_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 503.050240E+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 66218. MP_Allreduce 9696 492. MP_Sync 52 MP_Alltoall 1938 1961874. 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 0.025 0.035 96.295 96.296 qs_mol_dyn_low 1 2.0 0.003 0.003 95.999 96.007 qs_forces 11 3.9 0.014 0.026 95.922 95.923 qs_energies 11 4.9 0.001 0.001 92.764 92.786 scf_env_do_scf 11 5.9 0.000 0.001 81.671 81.673 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 74.413 74.414 dbcsr_multiply_generic 2055 12.4 0.107 0.110 55.035 55.351 qs_scf_new_mos 99 7.5 0.000 0.001 53.459 53.560 qs_scf_loop_do_ot 99 8.5 0.001 0.001 53.458 53.559 ot_scf_mini 99 9.5 0.002 0.002 50.894 51.011 velocity_verlet 10 3.0 0.001 0.001 48.460 48.461 multiply_cannon 2055 13.4 0.185 0.189 44.646 45.651 multiply_cannon_loop 2055 14.4 1.498 1.539 43.208 44.276 ot_mini 99 10.5 0.001 0.001 28.782 28.849 qs_ot_get_derivative 99 11.5 0.001 0.001 21.915 21.982 multiply_cannon_multrec 49320 15.4 12.332 13.110 18.512 19.531 rebuild_ks_matrix 110 8.3 0.000 0.000 16.620 16.749 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 16.620 16.749 qs_ks_update_qs_env 110 7.6 0.001 0.001 14.584 14.711 mp_waitall_1 220248 16.4 13.154 14.171 13.154 14.171 qs_ot_get_p 110 10.4 0.001 0.001 13.232 13.354 multiply_cannon_sync_h2d 49320 15.4 10.100 10.753 10.100 10.753 qs_ot_p2m_diag 48 11.0 0.012 0.019 9.331 9.379 qs_rho_update_rho_low 110 7.6 0.001 0.001 8.822 8.968 calculate_rho_elec 110 8.6 0.021 0.026 8.822 8.967 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 8.320 8.856 init_scf_run 11 5.9 0.000 0.002 8.820 8.821 scf_env_initial_rho_setup 11 6.9 0.001 0.001 8.820 8.820 sum_up_and_integrate 110 10.3 0.036 0.042 8.720 8.733 integrate_v_rspace 110 11.3 0.003 0.004 8.684 8.705 multiply_cannon_metrocomm3 49320 15.4 0.078 0.080 6.932 8.435 cp_dbcsr_syevd 48 12.0 0.002 0.002 8.267 8.268 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.279 7.797 apply_single 110 13.6 0.000 0.000 7.278 7.797 cp_fm_diag_elpa 48 13.0 0.000 0.001 7.382 7.407 cp_fm_diag_elpa_base 48 14.0 7.353 7.383 7.379 7.403 init_scf_loop 11 6.9 0.000 0.000 7.205 7.205 ot_diis_step 99 11.5 0.006 0.006 6.641 6.641 dbcsr_mm_accdrv_process 87628 16.1 1.926 2.013 6.057 6.616 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.397 6.443 rs_pw_transfer 902 11.9 0.011 0.013 5.782 6.274 density_rs2pw 110 9.6 0.004 0.005 5.547 6.012 mp_sum_l 6514 12.8 4.680 5.501 4.680 5.501 make_m2s 4110 13.4 0.061 0.065 5.256 5.402 make_images 4110 14.4 0.179 0.193 5.158 5.305 calculate_dm_sparse 110 9.5 0.000 0.001 5.143 5.291 pw_transfer 1331 11.6 0.056 0.067 5.089 5.252 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 4.999 5.164 prepare_preconditioner 11 7.9 0.000 0.000 4.744 4.770 make_preconditioner 11 8.9 0.000 0.000 4.744 4.770 wfi_extrapolate 11 7.9 0.001 0.001 4.649 4.650 make_full_inverse_cholesky 11 9.9 0.000 0.000 4.465 4.530 fft3d_ps 1111 14.6 0.793 0.887 4.205 4.356 multiply_cannon_metrocomm1 49320 15.4 0.060 0.062 2.958 4.269 potential_pw2rs 110 12.3 0.006 0.007 4.138 4.210 jit_kernel_multiply 13 15.9 3.852 4.172 3.852 4.172 fft_wrap_pw1pw2_140 451 13.1 0.444 0.493 3.955 4.109 calculate_first_density_matrix 1 7.0 0.000 0.000 4.044 4.053 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.719 3.725 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.563 3.660 grid_integrate_task_list 110 12.3 3.216 3.501 3.216 3.501 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.243 3.300 mp_alltoall_z22v 1111 16.6 2.910 3.143 2.910 3.143 mp_alltoall_d11v 2046 13.8 2.611 3.093 2.611 3.093 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 2.869 2.904 cp_fm_cholesky_invert 11 10.9 2.664 2.669 2.664 2.669 make_images_data 4110 15.4 0.043 0.046 2.258 2.515 mp_waitany 14300 13.8 1.972 2.496 1.972 2.496 grid_collocate_task_list 110 9.6 2.082 2.294 2.082 2.294 hybrid_alltoall_any 4261 16.3 0.082 0.475 1.924 2.289 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.170 2.186 acc_transpose_blocks 49320 15.4 0.213 0.226 2.108 2.177 mp_sum_d 3889 11.9 1.497 2.042 1.497 2.042 make_images_sizes 4110 15.4 0.004 0.005 1.459 1.969 mp_alltoall_i44 4110 16.4 1.455 1.965 1.455 1.965 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=96.296000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=477.363636, yerr=2.962576 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/77ac92570e643352348b65cb9881e203655a06a9_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 586.547200E+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 66437. MP_Allreduce 9695 570. MP_Sync 52 MP_Alltoall 1717 2856501. 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 0.026 0.085 73.661 73.663 qs_mol_dyn_low 1 2.0 0.003 0.005 72.958 72.967 qs_forces 11 3.9 0.009 0.016 72.668 72.669 qs_energies 11 4.9 0.003 0.016 69.275 69.285 scf_env_do_scf 11 5.9 0.001 0.001 58.806 58.809 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 50.847 50.868 dbcsr_multiply_generic 2055 12.4 0.114 0.119 39.705 39.903 velocity_verlet 10 3.0 0.029 0.229 36.692 36.693 qs_scf_new_mos 99 7.5 0.001 0.001 34.325 34.439 qs_scf_loop_do_ot 99 8.5 0.001 0.001 34.325 34.438 multiply_cannon 2055 13.4 0.221 0.240 32.891 34.381 ot_scf_mini 99 9.5 0.003 0.003 32.669 32.790 multiply_cannon_loop 2055 14.4 0.920 0.946 31.589 32.773 ot_mini 99 10.5 0.001 0.001 19.463 19.586 multiply_cannon_multrec 24660 15.4 7.621 9.092 15.236 16.810 rebuild_ks_matrix 110 8.3 0.000 0.000 13.995 14.102 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.019 13.995 14.102 qs_ot_get_derivative 99 11.5 0.001 0.001 13.674 13.798 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.293 12.385 mp_waitall_1 176588 16.5 7.887 10.382 7.887 10.382 multiply_cannon_sync_h2d 24660 15.4 7.035 8.158 7.035 8.158 multiply_cannon_metrocomm3 24660 15.4 0.067 0.069 5.270 7.970 init_scf_loop 11 6.9 0.003 0.020 7.924 7.962 dbcsr_mm_accdrv_process 52282 16.1 5.109 6.636 7.454 7.751 init_scf_run 11 5.9 0.000 0.001 7.421 7.421 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.420 7.421 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.682 7.347 apply_single 110 13.6 0.000 0.001 6.681 7.347 qs_ot_get_p 110 10.4 0.002 0.014 6.732 6.874 sum_up_and_integrate 110 10.3 0.053 0.059 6.783 6.800 integrate_v_rspace 110 11.3 0.003 0.003 6.730 6.748 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.468 6.480 calculate_rho_elec 110 8.6 0.040 0.048 6.468 6.479 ot_diis_step 99 11.5 0.010 0.010 5.743 5.743 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.836 5.702 prepare_preconditioner 11 7.9 0.000 0.001 5.662 5.676 make_preconditioner 11 8.9 0.000 0.001 5.662 5.676 make_full_inverse_cholesky 11 9.9 0.001 0.009 5.231 5.389 make_m2s 4110 13.4 0.056 0.059 4.206 4.709 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.648 4.668 make_images 4110 14.4 0.402 0.445 4.099 4.598 cp_dbcsr_syevd 48 12.0 0.003 0.005 4.218 4.218 rs_pw_transfer 902 11.9 0.012 0.013 3.426 4.191 density_rs2pw 110 9.6 0.005 0.006 3.560 4.037 pw_transfer 1331 11.6 0.067 0.076 3.837 3.981 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.877 3.881 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.729 3.870 cp_fm_diag_elpa 48 13.0 0.001 0.002 3.624 3.641 cp_fm_diag_elpa_base 48 14.0 3.570 3.592 3.619 3.635 calculate_first_density_matrix 1 7.0 0.000 0.001 3.593 3.595 jit_kernel_multiply 12 16.3 1.988 3.592 1.988 3.592 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.521 3.572 wfi_extrapolate 11 7.9 0.001 0.001 3.561 3.561 grid_integrate_task_list 110 12.3 3.164 3.354 3.164 3.354 fft_wrap_pw1pw2_140 451 13.1 0.520 0.537 3.160 3.306 calculate_dm_sparse 110 9.5 0.001 0.001 3.219 3.256 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.176 3.229 fft3d_ps 1111 14.6 1.112 1.350 2.704 2.867 make_images_data 4110 15.4 0.046 0.049 2.341 2.852 hybrid_alltoall_any 4261 16.3 0.102 0.445 2.050 2.789 cp_fm_cholesky_invert 11 10.9 2.641 2.648 2.641 2.648 mp_sum_l 6514 12.8 2.008 2.636 2.008 2.636 potential_pw2rs 110 12.3 0.008 0.009 2.548 2.567 grid_collocate_task_list 110 9.6 2.072 2.536 2.072 2.536 qs_energies_init_hamiltonians 11 5.9 0.011 0.023 2.460 2.467 mp_waitany 10164 13.8 1.524 2.298 1.524 2.298 mp_alltoall_d11v 2046 13.8 1.796 2.079 1.796 2.079 rs_pw_transfer_RS2PW_140 121 11.5 0.263 0.661 1.312 2.061 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.946 1.970 mp_sum_d 3889 11.9 1.361 1.908 1.361 1.908 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.858 1.870 acc_transpose_blocks 24660 15.4 0.111 0.115 1.774 1.858 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.795 1.806 multiply_cannon_metrocomm4 22605 15.4 0.075 0.081 0.785 1.679 mp_allgather_i34 2055 14.4 0.631 1.629 0.631 1.629 qs_env_update_s_mstruct 11 6.9 0.006 0.015 1.535 1.617 mp_irecv_dv 57340 16.2 0.659 1.557 0.659 1.557 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.374 1.481 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=73.663000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=554.909091, yerr=7.316911 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/77ac92570e643352348b65cb9881e203655a06a9_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 662.843392E+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 66428. MP_Allreduce 9694 567. 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 0.049 0.069 69.277 69.278 qs_mol_dyn_low 1 2.0 0.005 0.022 68.565 68.586 qs_forces 11 3.9 0.003 0.003 68.431 68.433 qs_energies 11 4.9 0.003 0.015 65.101 65.108 scf_env_do_scf 11 5.9 0.000 0.001 54.559 54.559 scf_env_do_scf_inner_loop 99 6.5 0.009 0.037 44.410 44.411 velocity_verlet 10 3.0 0.001 0.001 35.516 35.520 dbcsr_multiply_generic 2055 12.4 0.108 0.111 31.454 32.021 qs_scf_new_mos 99 7.5 0.001 0.001 28.122 28.230 qs_scf_loop_do_ot 99 8.5 0.001 0.001 28.121 28.229 ot_scf_mini 99 9.5 0.002 0.003 26.822 26.942 multiply_cannon 2055 13.4 0.211 0.227 24.155 25.910 multiply_cannon_loop 2055 14.4 0.614 0.628 22.807 23.895 ot_mini 99 10.5 0.001 0.001 14.675 14.795 rebuild_ks_matrix 110 8.3 0.000 0.000 13.073 13.240 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 13.073 13.239 multiply_cannon_multrec 16440 15.4 3.760 4.580 10.914 11.749 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.516 11.665 mp_waitall_1 139946 16.5 8.162 11.546 8.162 11.546 qs_ot_get_derivative 99 11.5 0.001 0.001 10.014 10.138 init_scf_loop 11 6.9 0.000 0.000 10.102 10.103 init_scf_run 11 5.9 0.000 0.001 8.077 8.078 scf_env_initial_rho_setup 11 6.9 0.001 0.001 8.077 8.077 multiply_cannon_metrocomm3 16440 15.4 0.042 0.045 4.898 8.070 prepare_preconditioner 11 7.9 0.000 0.000 7.778 7.799 make_preconditioner 11 8.9 0.000 0.000 7.778 7.799 qs_ot_get_p 110 10.4 0.001 0.001 7.389 7.694 make_full_inverse_cholesky 11 9.9 0.000 0.000 7.059 7.433 dbcsr_mm_accdrv_process 34862 16.1 5.966 6.660 7.012 7.212 sum_up_and_integrate 110 10.3 0.061 0.062 7.050 7.064 integrate_v_rspace 110 11.3 0.003 0.003 6.989 7.003 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.735 6.749 calculate_rho_elec 110 8.6 0.059 0.060 6.734 6.748 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.149 5.604 apply_single 110 13.6 0.000 0.000 5.148 5.604 qs_ot_p2m_diag 48 11.0 0.042 0.044 5.214 5.222 make_m2s 4110 13.4 0.050 0.051 4.666 5.125 make_images 4110 14.4 0.393 0.512 4.551 5.010 density_rs2pw 110 9.6 0.005 0.005 3.725 5.002 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.814 4.815 calculate_first_density_matrix 1 7.0 0.000 0.000 4.698 4.700 ot_diis_step 99 11.5 0.011 0.011 4.606 4.607 rs_pw_transfer 902 11.9 0.010 0.011 3.297 4.568 multiply_cannon_sync_h2d 16440 15.4 3.621 4.227 3.621 4.227 pw_transfer 1331 11.6 0.066 0.074 4.169 4.194 cp_fm_diag_elpa 48 13.0 0.001 0.001 4.136 4.143 cp_fm_diag_elpa_base 48 14.0 4.063 4.094 4.134 4.140 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 4.062 4.091 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.385 4.055 calculate_dm_sparse 110 9.5 0.001 0.001 3.888 3.919 fft_wrap_pw1pw2_140 451 13.1 0.635 0.642 3.358 3.388 grid_integrate_task_list 110 12.3 3.153 3.379 3.153 3.379 cp_fm_cholesky_invert 11 10.9 3.307 3.316 3.307 3.316 wfi_extrapolate 11 7.9 0.001 0.001 3.276 3.276 make_images_data 4110 15.4 0.043 0.048 2.695 3.228 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.144 3.145 hybrid_alltoall_any 4261 16.3 0.105 0.372 2.329 3.015 fft3d_ps 1111 14.6 1.092 1.102 2.889 2.912 mp_sum_l 6514 12.8 2.033 2.832 2.033 2.832 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 2.786 2.804 potential_pw2rs 110 12.3 0.011 0.011 2.717 2.734 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.680 2.731 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.617 2.685 mp_waitany 17072 13.8 1.266 2.587 1.266 2.587 mp_alltoall_d11v 2046 13.8 2.130 2.567 2.130 2.567 multiply_cannon_metrocomm4 14385 15.4 0.046 0.049 0.908 2.533 jit_kernel_multiply 8 16.5 0.650 2.478 0.650 2.478 grid_collocate_task_list 110 9.6 2.079 2.460 2.079 2.460 mp_irecv_dv 48980 15.7 0.837 2.405 0.837 2.405 rs_pw_transfer_RS2PW_140 121 11.5 0.175 0.179 0.927 2.204 dbcsr_complete_redistribute 325 12.2 0.320 0.427 1.737 2.199 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.984 1.985 cp_fm_cholesky_decompose 22 10.9 1.921 1.942 1.921 1.942 mp_allgather_i34 2055 14.4 0.676 1.887 0.676 1.887 cp_fm_upper_to_full 70 14.2 1.408 1.818 1.408 1.818 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.755 1.772 mp_alltoall_z22v 1111 16.6 1.588 1.621 1.588 1.621 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.152 1.598 parallel_gemm_fm 81 9.0 0.000 0.000 1.514 1.516 parallel_gemm_fm_cosma 81 10.0 1.514 1.516 1.514 1.516 make_basis_sm 11 9.8 0.000 0.000 1.509 1.512 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.357 1.497 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.403 1.419 rs_gather_matrices 110 12.3 0.234 0.265 1.037 1.409 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=69.278000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=626.545455, yerr=9.930336 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/77ac92570e643352348b65cb9881e203655a06a9_performance_tests/12/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 601.317074E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 4916280 0.0% 0.0% 100.0% average stack size 0.0 0.0 200.2 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 734.228480E+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 66426. MP_Allreduce 9694 608. 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 0.271 0.723 73.631 73.662 qs_mol_dyn_low 1 2.0 0.003 0.004 72.446 72.452 qs_forces 11 3.9 0.003 0.003 72.298 72.300 qs_energies 11 4.9 0.004 0.030 68.781 68.785 scf_env_do_scf 11 5.9 0.001 0.002 59.062 59.065 scf_env_do_scf_inner_loop 99 6.5 0.052 0.065 46.019 46.020 velocity_verlet 10 3.0 0.001 0.001 39.804 39.809 dbcsr_multiply_generic 2055 12.4 0.124 0.137 31.487 31.711 qs_scf_new_mos 99 7.5 0.001 0.001 29.771 29.912 qs_scf_loop_do_ot 99 8.5 0.001 0.001 29.771 29.911 ot_scf_mini 99 9.5 0.003 0.004 28.103 28.242 multiply_cannon 2055 13.4 0.244 0.276 23.837 24.880 multiply_cannon_loop 2055 14.4 0.883 0.909 22.171 22.870 ot_mini 99 10.5 0.001 0.001 15.093 15.262 multiply_cannon_multrec 24660 15.4 4.193 6.763 13.339 14.600 rebuild_ks_matrix 110 8.3 0.000 0.000 12.953 13.062 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.953 13.062 init_scf_loop 11 6.9 0.001 0.005 12.988 12.989 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.468 11.564 prepare_preconditioner 11 7.9 0.000 0.000 10.962 10.976 make_preconditioner 11 8.9 0.000 0.001 10.962 10.976 qs_ot_get_derivative 99 11.5 0.001 0.001 10.820 10.959 make_full_inverse_cholesky 11 9.9 0.000 0.001 9.150 10.633 dbcsr_mm_accdrv_process 52304 16.0 7.319 8.605 8.998 9.879 qs_ot_get_p 110 10.4 0.001 0.001 7.969 8.107 mp_waitall_1 121746 16.5 5.096 7.675 5.096 7.675 sum_up_and_integrate 110 10.3 0.067 0.071 7.109 7.121 integrate_v_rspace 110 11.3 0.003 0.003 7.041 7.060 init_scf_run 11 5.9 0.000 0.001 6.794 6.794 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.794 6.794 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.750 6.761 calculate_rho_elec 110 8.6 0.078 0.082 6.750 6.761 make_m2s 4110 13.4 0.060 0.061 5.825 6.342 make_images 4110 14.4 0.577 0.693 5.682 6.194 qs_ot_p2m_diag 48 11.0 0.055 0.064 5.776 5.791 cp_dbcsr_syevd 48 12.0 0.003 0.006 5.265 5.266 cp_fm_upper_to_full 70 14.2 3.382 4.907 3.382 4.907 cp_fm_diag_elpa 48 13.0 0.001 0.001 4.552 4.563 cp_fm_diag_elpa_base 48 14.0 4.383 4.449 4.547 4.558 pw_transfer 1331 11.6 0.066 0.076 4.288 4.337 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 4.181 4.234 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.104 4.220 apply_single 110 13.6 0.000 0.000 4.104 4.220 ot_diis_step 99 11.5 0.011 0.012 4.219 4.219 dbcsr_complete_redistribute 325 12.2 0.424 0.500 2.932 4.093 density_rs2pw 110 9.6 0.005 0.006 3.586 4.093 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.725 3.820 calculate_dm_sparse 110 9.5 0.001 0.001 3.746 3.779 multiply_cannon_metrocomm3 24660 15.4 0.036 0.037 1.618 3.697 fft_wrap_pw1pw2_140 451 13.1 0.666 0.689 3.460 3.517 rs_pw_transfer 902 11.9 0.010 0.011 2.985 3.500 calculate_first_density_matrix 1 7.0 0.000 0.003 3.490 3.493 make_images_data 4110 15.4 0.046 0.050 2.925 3.476 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.289 3.434 grid_integrate_task_list 110 12.3 3.263 3.412 3.263 3.412 hybrid_alltoall_any 4261 16.3 0.120 0.456 2.530 3.374 cp_fm_cholesky_invert 11 10.9 3.342 3.352 3.342 3.352 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.291 3.343 multiply_cannon_sync_h2d 24660 15.4 3.151 3.323 3.151 3.323 wfi_extrapolate 11 7.9 0.001 0.001 3.203 3.203 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.100 3.104 fft3d_ps 1111 14.6 1.086 1.122 2.946 2.976 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.802 2.937 mp_alltoall_i22 605 13.7 1.714 2.863 1.714 2.863 potential_pw2rs 110 12.3 0.012 0.013 2.657 2.675 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.615 2.657 jit_kernel_multiply 10 15.8 1.345 2.472 1.345 2.472 mp_alltoall_d11v 2046 13.8 2.144 2.448 2.144 2.448 grid_collocate_task_list 110 9.6 2.180 2.421 2.180 2.421 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 2.326 2.344 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.318 2.319 cp_fm_cholesky_decompose 22 10.9 2.028 2.086 2.028 2.086 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.801 1.830 acc_transpose_blocks 24660 15.4 0.108 0.111 1.637 1.758 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.733 1.744 mp_alltoall_z22v 1111 16.6 1.645 1.728 1.645 1.728 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.604 1.702 mp_waitany 13376 13.8 1.182 1.684 1.182 1.684 mp_allgather_i34 2055 14.4 0.769 1.682 0.769 1.682 mp_sum_l 6514 12.8 1.180 1.664 1.180 1.664 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=73.662000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=692.818182, yerr=9.934496 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/77ac92570e643352348b65cb9881e203655a06a9_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 833.802240E+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 66424. MP_Allreduce 9694 649. 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 0.329 0.368 62.919 62.982 qs_mol_dyn_low 1 2.0 0.003 0.003 61.846 61.851 qs_forces 11 3.9 0.003 0.003 61.764 61.765 qs_energies 11 4.9 0.001 0.001 57.916 57.921 scf_env_do_scf 11 5.9 0.000 0.001 48.751 48.752 scf_env_do_scf_inner_loop 99 6.5 0.003 0.012 39.606 39.607 velocity_verlet 10 3.0 0.001 0.001 33.950 33.954 dbcsr_multiply_generic 2055 12.4 0.105 0.109 24.436 24.595 qs_scf_new_mos 99 7.5 0.001 0.001 23.508 23.569 qs_scf_loop_do_ot 99 8.5 0.001 0.001 23.508 23.569 ot_scf_mini 99 9.5 0.002 0.002 22.190 22.238 multiply_cannon 2055 13.4 0.247 0.258 18.275 19.763 multiply_cannon_loop 2055 14.4 0.318 0.331 16.760 17.197 rebuild_ks_matrix 110 8.3 0.000 0.000 12.809 12.856 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 12.809 12.855 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.416 11.457 ot_mini 99 10.5 0.001 0.001 11.381 11.422 init_scf_loop 11 6.9 0.000 0.000 9.089 9.091 multiply_cannon_multrec 8220 15.4 3.265 4.804 7.812 8.953 mp_waitall_1 103326 16.6 6.830 8.345 6.830 8.345 qs_ot_get_derivative 99 11.5 0.001 0.001 7.410 7.458 prepare_preconditioner 11 7.9 0.000 0.000 7.103 7.108 make_preconditioner 11 8.9 0.000 0.000 7.103 7.108 sum_up_and_integrate 110 10.3 0.080 0.081 6.998 7.008 integrate_v_rspace 110 11.3 0.003 0.003 6.918 6.927 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.834 6.851 calculate_rho_elec 110 8.6 0.114 0.115 6.834 6.851 qs_ot_get_p 110 10.4 0.001 0.001 6.699 6.772 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.644 6.739 init_scf_run 11 5.9 0.000 0.002 5.886 5.886 scf_env_initial_rho_setup 11 6.9 0.004 0.005 5.886 5.886 dbcsr_mm_accdrv_process 17442 15.9 3.017 4.167 4.416 5.393 qs_ot_p2m_diag 48 11.0 0.081 0.084 5.030 5.035 make_m2s 4110 13.4 0.039 0.040 4.636 4.881 make_images 4110 14.4 0.638 0.690 4.505 4.750 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.693 4.694 multiply_cannon_metrocomm3 8220 15.4 0.017 0.018 3.202 4.573 pw_transfer 1331 11.6 0.067 0.072 4.419 4.439 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.311 4.330 cp_fm_diag_elpa 48 13.0 0.001 0.001 4.004 4.011 cp_fm_diag_elpa_base 48 14.0 3.945 3.967 4.001 4.007 ot_diis_step 99 11.5 0.012 0.012 3.938 3.939 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.819 3.869 apply_single 110 13.6 0.000 0.000 3.819 3.869 density_rs2pw 110 9.6 0.004 0.005 3.488 3.860 cp_fm_cholesky_invert 11 10.9 3.744 3.748 3.744 3.748 fft_wrap_pw1pw2_140 451 13.1 0.833 0.843 3.631 3.669 grid_integrate_task_list 110 12.3 3.352 3.528 3.352 3.528 make_images_data 4110 15.4 0.038 0.044 2.635 3.082 hybrid_alltoall_any 4261 16.3 0.200 0.862 2.507 3.016 rs_pw_transfer 902 11.9 0.010 0.011 2.618 3.015 wfi_extrapolate 11 7.9 0.001 0.001 3.003 3.003 multiply_cannon_sync_h2d 8220 15.4 2.883 2.970 2.883 2.970 fft3d_ps 1111 14.6 1.140 1.189 2.857 2.879 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.840 2.842 calculate_first_density_matrix 1 7.0 0.000 0.000 2.769 2.773 qs_energies_init_hamiltonians 11 5.9 0.017 0.026 2.729 2.730 calculate_dm_sparse 110 9.5 0.001 0.001 2.669 2.727 grid_collocate_task_list 110 9.6 2.280 2.544 2.280 2.544 potential_pw2rs 110 12.3 0.015 0.015 2.430 2.444 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.411 2.427 mp_alltoall_d11v 2046 13.8 2.017 2.302 2.017 2.302 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.211 2.238 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.076 2.096 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.773 1.990 mp_allgather_i34 2055 14.4 0.701 1.908 0.701 1.908 cp_fm_cholesky_decompose 22 10.9 1.862 1.874 1.862 1.874 dbcsr_complete_redistribute 325 12.2 0.547 0.569 1.651 1.759 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.729 1.735 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.539 1.665 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.568 1.583 mp_waitany 9240 13.8 1.129 1.541 1.129 1.541 mp_alltoall_z22v 1111 16.6 1.482 1.515 1.482 1.515 jit_kernel_multiply 7 15.6 1.084 1.480 1.084 1.480 qs_create_task_list 11 7.9 0.007 0.014 1.242 1.347 generate_qs_task_list 11 8.9 0.374 0.442 1.235 1.346 multiply_cannon_metrocomm1 8220 15.4 0.021 0.022 0.981 1.333 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.292 1.320 rs_gather_matrices 110 12.3 0.328 0.372 1.041 1.301 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=62.982000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=783.181818, yerr=14.332532 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/77ac92570e643352348b65cb9881e203655a06a9_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 1.376432E+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 67104. MP_Allreduce 9672 819. 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 0.079 0.095 94.655 94.665 qs_mol_dyn_low 1 2.0 0.003 0.003 94.074 94.082 qs_forces 11 3.9 0.003 0.004 93.988 93.989 qs_energies 11 4.9 0.001 0.001 89.673 89.675 scf_env_do_scf 11 5.9 0.001 0.001 78.099 78.099 velocity_verlet 10 3.0 0.001 0.001 59.093 59.099 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 48.142 48.144 dbcsr_multiply_generic 2055 12.4 0.119 0.123 31.272 31.409 init_scf_loop 11 6.9 0.000 0.000 29.869 29.873 qs_scf_new_mos 99 7.5 0.001 0.001 28.836 28.864 qs_scf_loop_do_ot 99 8.5 0.001 0.001 28.836 28.863 prepare_preconditioner 11 7.9 0.000 0.000 27.690 27.695 make_preconditioner 11 8.9 0.000 0.000 27.690 27.695 make_full_inverse_cholesky 11 9.9 0.000 0.000 21.852 27.144 ot_scf_mini 99 9.5 0.002 0.002 27.014 27.037 multiply_cannon 2055 13.4 0.343 0.356 23.422 24.215 multiply_cannon_loop 2055 14.4 0.339 0.341 21.485 21.787 cp_fm_upper_to_full 70 14.2 12.629 18.168 12.629 18.168 rebuild_ks_matrix 110 8.3 0.001 0.001 14.857 14.910 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 14.856 14.909 ot_mini 99 10.5 0.001 0.001 14.560 14.585 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.442 13.489 multiply_cannon_multrec 8220 15.4 4.537 4.716 10.820 10.982 dbcsr_complete_redistribute 325 12.2 1.021 1.047 7.586 10.790 qs_ot_get_derivative 99 11.5 0.001 0.001 9.812 9.834 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.459 9.658 mp_waitall_1 84994 16.7 8.631 9.601 8.631 9.601 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.823 8.996 mp_alltoall_i22 605 13.7 5.456 8.699 5.456 8.699 qs_rho_update_rho_low 110 7.6 0.001 0.001 8.285 8.313 calculate_rho_elec 110 8.6 0.225 0.226 8.284 8.312 sum_up_and_integrate 110 10.3 0.150 0.151 7.749 7.764 integrate_v_rspace 110 11.3 0.004 0.004 7.599 7.613 init_scf_run 11 5.9 0.000 0.001 7.252 7.252 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.252 7.252 qs_ot_get_p 110 10.4 0.001 0.001 6.995 7.018 make_m2s 4110 13.4 0.043 0.043 5.888 6.468 dbcsr_mm_accdrv_process 11614 15.7 3.379 3.642 6.141 6.417 cp_fm_cholesky_invert 11 10.9 6.405 6.411 6.405 6.411 make_images 4110 14.4 0.877 0.915 5.695 6.279 pw_transfer 1331 11.6 0.075 0.076 5.582 5.594 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 5.465 5.476 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 5.231 5.476 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.839 5.301 apply_single 110 13.6 0.000 0.000 4.838 5.300 qs_ot_p2m_diag 48 11.0 0.151 0.156 5.185 5.194 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.715 4.716 ot_diis_step 99 11.5 0.015 0.016 4.711 4.712 fft_wrap_pw1pw2_140 451 13.1 1.338 1.343 4.655 4.666 calculate_dm_sparse 110 9.5 0.001 0.001 4.380 4.394 density_rs2pw 110 9.6 0.004 0.004 4.215 4.228 make_images_data 4110 15.4 0.041 0.044 3.340 4.125 hybrid_alltoall_any 4261 16.3 0.257 0.557 3.279 4.120 cp_fm_diag_elpa 48 13.0 0.210 0.210 4.034 4.034 multiply_cannon_sync_h2d 8220 15.4 3.952 3.957 3.952 3.957 cp_fm_diag_elpa_base 48 14.0 3.274 3.470 3.820 3.821 grid_integrate_task_list 110 12.3 3.690 3.758 3.690 3.758 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.678 3.679 wfi_extrapolate 11 7.9 0.001 0.001 3.588 3.588 calculate_first_density_matrix 1 7.0 0.000 0.000 3.533 3.535 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.075 3.527 fft3d_ps 1111 14.6 1.311 1.321 3.394 3.408 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.067 3.071 jit_kernel_multiply 10 15.1 2.557 2.900 2.557 2.900 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 2.877 2.893 potential_pw2rs 110 12.3 0.021 0.021 2.734 2.741 grid_collocate_task_list 110 9.6 2.654 2.667 2.654 2.667 rs_pw_transfer 902 11.9 0.010 0.011 2.632 2.653 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.572 2.583 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.488 2.513 mp_alltoall_d11v 2046 13.8 2.301 2.422 2.301 2.422 cp_fm_cholesky_decompose 22 10.9 2.289 2.304 2.289 2.304 qs_env_update_s_mstruct 11 6.9 0.001 0.001 2.203 2.257 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.100 2.191 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.021 2.026 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.923 1.951 qs_create_task_list 11 7.9 0.001 0.001 1.904 1.951 generate_qs_task_list 11 8.9 0.731 0.785 1.903 1.950 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=94.665000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1242.000000, yerr=58.668561 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/77ac92570e643352348b65cb9881e203655a06a9_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 630.050816E+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 57767. MP_Allreduce 11004 802. MP_Sync 87 MP_Alltoall 2226 3325623. MP_SendRecv 24320 18752. MP_ISendRecv 24320 18752. MP_Wait 42476 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.039 0.054 223.143 223.144 qs_mol_dyn_low 1 2.0 0.003 0.003 222.433 222.447 qs_forces 11 3.9 0.005 0.006 222.267 222.268 qs_energies 11 4.9 0.001 0.001 216.473 216.485 scf_env_do_scf 11 5.9 0.001 0.001 199.221 199.225 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 175.983 175.985 qs_scf_new_mos 117 7.6 0.001 0.001 133.103 133.365 qs_scf_loop_do_ot 117 8.6 0.001 0.001 133.102 133.364 velocity_verlet 10 3.0 0.001 0.001 131.856 131.858 dbcsr_multiply_generic 2507 12.6 0.189 0.202 127.677 128.187 ot_scf_mini 117 9.6 0.003 0.003 126.476 126.741 multiply_cannon 2507 13.6 0.241 0.249 102.760 104.399 multiply_cannon_loop 2507 14.6 2.059 2.111 100.246 101.660 ot_mini 117 10.6 0.001 0.001 68.437 68.731 qs_ot_get_derivative 117 11.6 0.001 0.001 43.480 43.754 multiply_cannon_multrec 60168 15.6 33.328 34.964 42.025 43.537 qs_ot_get_p 128 10.4 0.001 0.001 35.706 36.051 rebuild_ks_matrix 128 8.3 0.001 0.001 35.276 35.648 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.017 35.275 35.647 mp_waitall_1 267128 16.5 30.926 33.871 30.926 33.871 qs_ks_update_qs_env 128 7.6 0.001 0.001 31.685 31.995 multiply_cannon_sync_h2d 60168 15.6 27.247 29.208 27.247 29.208 qs_ot_p2m_diag 83 11.4 0.079 0.091 28.567 28.622 cp_dbcsr_syevd 83 12.4 0.004 0.005 25.842 25.845 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.489 25.261 apply_single 128 13.6 0.001 0.001 24.489 25.261 ot_diis_step 117 11.6 0.007 0.008 24.677 24.678 init_scf_loop 11 6.9 0.000 0.000 23.145 23.146 cp_fm_diag_elpa 83 13.4 0.000 0.000 22.078 22.112 cp_fm_diag_elpa_base 83 14.4 21.972 22.018 22.074 22.106 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 20.703 20.878 prepare_preconditioner 11 7.9 0.000 0.000 18.401 18.444 make_preconditioner 11 8.9 0.000 0.000 18.401 18.443 multiply_cannon_metrocomm3 60168 15.6 0.112 0.117 16.097 17.869 make_full_inverse_cholesky 11 9.9 0.000 0.000 17.552 17.753 sum_up_and_integrate 128 10.3 0.089 0.106 15.607 15.621 qs_rho_update_rho_low 128 7.7 0.001 0.001 15.399 15.537 calculate_rho_elec 128 8.7 0.046 0.065 15.398 15.536 integrate_v_rspace 128 11.3 0.004 0.005 15.518 15.534 make_m2s 5014 13.6 0.103 0.110 14.805 15.165 make_images 5014 14.6 0.403 0.422 14.624 15.000 init_scf_run 11 5.9 0.000 0.002 12.958 12.958 scf_env_initial_rho_setup 11 6.9 0.001 0.003 12.957 12.958 density_rs2pw 128 9.7 0.006 0.007 8.750 11.960 rs_pw_transfer 1046 11.9 0.016 0.019 7.766 11.016 cp_fm_cholesky_invert 11 10.9 10.715 10.722 10.715 10.722 multiply_cannon_metrocomm1 60168 15.6 0.087 0.091 6.948 10.550 mp_sum_l 7870 13.0 8.836 9.917 8.836 9.917 wfi_extrapolate 11 7.9 0.001 0.001 9.459 9.459 pw_transfer 1547 11.6 0.075 0.088 8.745 9.037 dbcsr_mm_accdrv_process 124484 16.2 3.153 3.263 8.260 8.858 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 8.541 8.838 calculate_dm_sparse 128 9.5 0.001 0.001 8.497 8.612 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.341 8.561 make_images_data 5014 15.6 0.066 0.072 7.200 8.173 qs_ot_get_orbitals 117 10.6 0.001 0.001 8.094 8.173 grid_integrate_task_list 128 12.3 6.972 7.671 6.972 7.671 hybrid_alltoall_any 5200 16.5 0.292 2.264 6.228 7.363 fft_wrap_pw1pw2_140 523 13.2 1.255 1.315 6.937 7.238 cp_dbcsr_sm_fm_multiply 37 9.5 0.003 0.003 6.854 6.861 fft3d_ps 1291 14.7 2.148 2.784 6.330 6.748 mp_alltoall_d11v 2415 14.1 4.733 6.087 4.733 6.087 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.936 6.024 mp_waitany 16020 13.9 2.737 5.979 2.737 5.979 potential_pw2rs 128 12.3 0.009 0.011 5.869 5.974 grid_collocate_task_list 128 9.7 4.556 5.809 4.556 5.809 cp_fm_cholesky_decompose 22 10.9 5.397 5.410 5.397 5.410 rs_pw_transfer_RS2PW_140 139 11.5 0.279 0.296 2.132 5.358 mp_sum_d 4470 12.1 3.533 4.473 3.533 4.473 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=223.144000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=596.727273, yerr=6.743224 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/77ac92570e643352348b65cb9881e203655a06a9_performance_tests/16/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 2.183246E+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 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 830.709760E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100942E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834332048 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58363. MP_Allreduce 10978 967. MP_Sync 87 MP_Alltoall 1969 4982848. MP_SendRecv 12032 47072. MP_ISendRecv 12032 47072. MP_Wait 25916 MP_ISend 11748 212467. MP_IRecv 11748 212467. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.076 0.118 204.254 204.259 qs_mol_dyn_low 1 2.0 0.003 0.004 203.599 203.610 qs_forces 11 3.9 0.014 0.025 203.480 203.481 qs_energies 11 4.9 0.007 0.026 196.451 196.470 scf_env_do_scf 11 5.9 0.001 0.002 178.548 178.559 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 143.115 143.118 velocity_verlet 10 3.0 0.001 0.001 126.665 126.666 dbcsr_multiply_generic 2507 12.6 0.187 0.194 100.023 101.181 qs_scf_new_mos 117 7.6 0.001 0.002 100.522 100.978 qs_scf_loop_do_ot 117 8.6 0.001 0.002 100.521 100.977 ot_scf_mini 117 9.6 0.004 0.005 95.674 96.114 multiply_cannon 2507 13.6 0.476 0.526 79.030 83.431 multiply_cannon_loop 2507 14.6 1.243 1.279 75.084 77.723 ot_mini 117 10.6 0.001 0.001 51.307 51.772 mp_waitall_1 214728 16.6 25.982 40.696 25.982 40.696 multiply_cannon_multrec 30084 15.6 22.039 26.605 32.208 36.969 rebuild_ks_matrix 128 8.3 0.001 0.001 34.956 35.463 qs_ks_build_kohn_sham_matrix 128 9.3 0.038 0.185 34.955 35.463 init_scf_loop 11 6.9 0.097 0.309 35.327 35.329 qs_ks_update_qs_env 128 7.6 0.001 0.001 31.429 31.919 prepare_preconditioner 11 7.9 0.005 0.030 30.576 30.742 make_preconditioner 11 8.9 0.008 0.060 30.572 30.742 make_full_inverse_cholesky 11 9.9 0.000 0.001 29.175 29.869 qs_ot_get_derivative 117 11.6 0.001 0.002 29.328 29.776 multiply_cannon_metrocomm3 30084 15.6 0.090 0.096 16.083 29.421 qs_ot_get_p 128 10.4 0.001 0.001 25.878 26.528 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.023 23.089 apply_single 128 13.6 0.001 0.001 22.023 23.088 ot_diis_step 117 11.6 0.014 0.015 21.802 21.804 multiply_cannon_sync_h2d 30084 15.6 19.287 21.763 19.287 21.763 qs_ot_p2m_diag 83 11.4 0.188 0.216 20.729 20.762 cp_dbcsr_syevd 83 12.4 0.042 0.297 19.532 19.534 cp_fm_cholesky_invert 11 10.9 17.643 17.656 17.643 17.656 make_m2s 5014 13.6 0.088 0.093 15.176 16.922 make_images 5014 14.6 1.166 1.357 14.969 16.717 cp_fm_diag_elpa 83 13.4 0.181 0.292 16.020 16.082 sum_up_and_integrate 128 10.3 0.116 0.134 15.963 15.997 integrate_v_rspace 128 11.3 0.004 0.004 15.847 15.886 cp_fm_diag_elpa_base 83 14.4 15.459 15.602 15.755 15.775 qs_rho_update_rho_low 128 7.7 0.001 0.001 15.456 15.490 calculate_rho_elec 128 8.7 0.088 0.104 15.455 15.490 init_scf_run 11 5.9 0.000 0.002 12.774 12.776 scf_env_initial_rho_setup 11 6.9 0.001 0.002 12.774 12.776 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 11.387 11.699 make_images_data 5014 15.6 0.063 0.071 9.168 11.286 density_rs2pw 128 9.7 0.007 0.015 8.586 10.943 hybrid_alltoall_any 5200 16.5 0.343 1.504 7.795 10.741 multiply_cannon_metrocomm4 27577 15.6 0.099 0.114 3.751 10.674 dbcsr_mm_accdrv_process 62242 16.2 4.635 5.429 9.626 10.535 mp_irecv_dv 69486 16.3 3.555 10.289 3.555 10.289 pw_transfer 1547 11.6 0.086 0.100 9.797 9.856 fft_wrap_pw1pw2 1291 12.7 0.011 0.011 9.571 9.634 rs_pw_transfer 1046 11.9 0.014 0.016 7.059 9.435 wfi_extrapolate 11 7.9 0.001 0.001 8.805 8.806 fft_wrap_pw1pw2_140 523 13.2 1.343 1.379 8.275 8.355 grid_integrate_task_list 128 12.3 7.132 7.685 7.132 7.685 cp_fm_cholesky_decompose 22 10.9 7.575 7.658 7.575 7.658 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.760 7.575 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.973 6.985 fft3d_ps 1291 14.7 2.812 2.977 6.871 6.934 calculate_dm_sparse 128 9.5 0.001 0.001 6.527 6.680 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 6.041 6.171 potential_pw2rs 128 12.3 0.016 0.019 5.952 5.996 grid_collocate_task_list 128 9.7 4.706 5.869 4.706 5.869 mp_alltoall_d11v 2415 14.1 5.066 5.860 5.066 5.860 mp_sum_l 7870 13.0 4.072 5.792 4.072 5.792 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.387 5.447 mp_allgather_i34 2507 14.6 2.133 5.376 2.133 5.376 mp_waitany 11748 13.9 2.563 4.929 2.563 4.929 dbcsr_complete_redistribute 395 12.7 0.766 0.861 3.620 4.500 rs_pw_transfer_RS2PW_140 139 11.5 0.355 0.382 2.113 4.491 mp_sum_d 4465 12.1 2.748 4.156 2.748 4.156 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=204.259000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=791.090909, yerr=2.842999 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/77ac92570e643352348b65cb9881e203655a06a9_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420241154048 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 11528896499712 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 93.514757E+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 6755939872 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 932.769792E+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 931531083728 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 58214. MP_Allreduce 11002 1006. 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 0.063 0.136 179.553 179.567 qs_mol_dyn_low 1 2.0 0.003 0.003 178.675 178.691 qs_forces 11 3.9 0.004 0.005 178.569 178.573 qs_energies 11 4.9 0.007 0.049 171.922 171.932 scf_env_do_scf 11 5.9 0.002 0.012 154.589 154.590 scf_env_do_scf_inner_loop 117 6.6 0.004 0.010 119.292 119.294 velocity_verlet 10 3.0 0.001 0.001 113.343 113.345 dbcsr_multiply_generic 2507 12.6 0.178 0.184 83.690 84.796 qs_scf_new_mos 117 7.6 0.001 0.002 81.061 81.421 qs_scf_loop_do_ot 117 8.6 0.001 0.001 81.061 81.420 ot_scf_mini 117 9.6 0.004 0.004 76.934 77.320 multiply_cannon 2507 13.6 0.501 0.517 63.967 68.405 multiply_cannon_loop 2507 14.6 0.853 0.883 60.916 63.512 ot_mini 117 10.6 0.001 0.001 43.408 43.810 mp_waitall_1 170520 16.6 25.243 35.335 25.243 35.335 init_scf_loop 11 6.9 0.001 0.004 35.200 35.201 rebuild_ks_matrix 128 8.3 0.001 0.001 30.831 31.344 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.018 30.831 31.344 prepare_preconditioner 11 7.9 0.000 0.001 31.127 31.167 make_preconditioner 11 8.9 0.000 0.001 31.127 31.167 make_full_inverse_cholesky 11 9.9 0.001 0.003 28.795 30.196 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.751 28.226 multiply_cannon_multrec 20056 15.6 13.530 16.426 24.130 27.069 multiply_cannon_metrocomm3 20056 15.6 0.058 0.062 15.444 25.207 qs_ot_get_derivative 117 11.6 0.002 0.002 23.676 24.070 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 20.008 21.157 apply_single 128 13.6 0.001 0.001 20.008 21.157 ot_diis_step 117 11.6 0.018 0.018 19.631 19.631 qs_ot_get_p 128 10.4 0.001 0.001 18.965 19.419 multiply_cannon_sync_h2d 20056 15.6 14.220 15.644 14.220 15.644 make_m2s 5014 13.6 0.081 0.086 14.499 15.395 make_images 5014 14.6 1.176 1.255 14.266 15.151 qs_ot_p2m_diag 83 11.4 0.266 0.273 14.767 14.778 sum_up_and_integrate 128 10.3 0.133 0.144 14.693 14.720 cp_fm_cholesky_invert 11 10.9 14.602 14.610 14.602 14.610 integrate_v_rspace 128 11.3 0.004 0.005 14.560 14.588 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.311 14.337 calculate_rho_elec 128 8.7 0.130 0.145 14.311 14.336 cp_dbcsr_syevd 83 12.4 0.006 0.009 13.752 13.753 init_scf_run 11 5.9 0.001 0.014 12.198 12.199 scf_env_initial_rho_setup 11 6.9 0.010 0.013 12.198 12.198 cp_fm_diag_elpa 83 13.4 0.001 0.002 10.541 10.563 cp_fm_diag_elpa_base 83 14.4 10.128 10.284 10.536 10.557 make_images_data 5014 15.6 0.060 0.068 8.948 10.202 dbcsr_mm_accdrv_process 41502 16.2 5.871 7.162 10.058 10.176 hybrid_alltoall_any 5200 16.5 0.434 1.985 7.680 9.652 multiply_cannon_metrocomm4 17549 15.6 0.063 0.075 3.455 9.352 density_rs2pw 128 9.7 0.006 0.007 7.197 9.176 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 8.834 9.116 mp_irecv_dv 50230 16.2 3.332 9.103 3.332 9.103 pw_transfer 1547 11.6 0.086 0.102 8.881 8.988 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.656 8.768 grid_integrate_task_list 128 12.3 7.252 7.779 7.252 7.779 fft_wrap_pw1pw2_140 523 13.2 1.412 1.435 7.626 7.752 cp_fm_cholesky_decompose 22 10.9 7.505 7.540 7.505 7.540 wfi_extrapolate 11 7.9 0.001 0.001 7.402 7.402 rs_pw_transfer 1046 11.9 0.014 0.015 5.229 7.245 cp_fm_upper_to_full 105 14.8 5.648 7.159 5.648 7.159 calculate_dm_sparse 128 9.5 0.001 0.001 6.944 7.038 dbcsr_complete_redistribute 395 12.7 1.163 1.191 4.533 6.250 fft3d_ps 1291 14.7 2.741 2.964 5.780 5.863 grid_collocate_task_list 128 9.7 4.900 5.796 4.900 5.796 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.652 5.660 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.606 5.474 mp_alltoall_d11v 2415 14.1 4.338 5.208 4.338 5.208 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.376 5.090 mp_sum_l 7870 13.0 3.424 4.993 3.424 4.993 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.778 4.895 potential_pw2rs 128 12.3 0.020 0.022 4.783 4.799 calculate_first_density_matrix 1 7.0 0.001 0.004 4.573 4.583 mp_allgather_i34 2507 14.6 1.569 4.554 1.569 4.554 mp_waitany 11748 13.9 2.347 4.424 2.347 4.424 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.032 4.070 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.312 4.000 rs_pw_transfer_RS2PW_140 139 11.5 0.334 0.363 1.912 3.922 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 3.794 3.795 mp_alltoall_i22 716 14.1 1.917 3.777 1.917 3.777 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.582 3.622 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=179.567000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=885.363636, yerr=9.412494 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/77ac92570e643352348b65cb9881e203655a06a9_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 4.353788E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 5977344 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.142575E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1143192 MPI messages size (bytes): total size 2.023815E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.770319E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 319024 36042702848 131072 < size <= 4194304 715736 785529176064 4194304 < size <= 16777216 70320 665379241840 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58354. MP_Allreduce 10977 1091. MP_Sync 87 MP_Alltoall 1712 12503107. MP_SendRecv 5888 75008. MP_ISendRecv 5888 75008. MP_Wait 22442 MP_ISend 14952 244818. MP_IRecv 14952 244818. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.034 0.180 204.576 204.612 qs_mol_dyn_low 1 2.0 0.003 0.004 203.519 203.537 qs_forces 11 3.9 0.004 0.005 203.373 203.379 qs_energies 11 4.9 0.003 0.011 196.055 196.064 scf_env_do_scf 11 5.9 0.003 0.030 177.875 177.882 velocity_verlet 10 3.0 0.001 0.001 133.225 133.229 scf_env_do_scf_inner_loop 117 6.6 0.016 0.038 127.106 127.109 qs_scf_new_mos 117 7.6 0.001 0.001 88.207 88.504 qs_scf_loop_do_ot 117 8.6 0.001 0.001 88.206 88.503 ot_scf_mini 117 9.6 0.003 0.004 83.685 83.976 dbcsr_multiply_generic 2507 12.6 0.200 0.204 81.756 82.454 multiply_cannon 2507 13.6 0.557 0.595 56.411 59.529 multiply_cannon_loop 2507 14.6 1.177 1.211 52.301 54.243 init_scf_loop 11 6.9 0.004 0.023 50.615 50.616 prepare_preconditioner 11 7.9 0.000 0.000 46.348 46.371 make_preconditioner 11 8.9 0.002 0.007 46.348 46.370 make_full_inverse_cholesky 11 9.9 0.000 0.001 39.961 44.998 ot_mini 117 10.6 0.001 0.001 44.217 44.498 multiply_cannon_multrec 30084 15.6 14.103 19.551 26.840 31.888 rebuild_ks_matrix 128 8.3 0.001 0.001 30.760 31.039 qs_ks_build_kohn_sham_matrix 128 9.3 0.022 0.062 30.760 31.038 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.728 27.973 mp_waitall_1 147882 16.7 18.126 27.411 18.126 27.411 qs_ot_get_p 128 10.4 0.001 0.001 25.584 25.892 qs_ot_get_derivative 117 11.6 0.002 0.002 24.498 24.794 make_m2s 5014 13.6 0.097 0.101 20.615 21.589 make_images 5014 14.6 1.958 2.255 20.305 21.281 qs_ot_p2m_diag 83 11.4 0.343 0.391 21.210 21.265 cp_dbcsr_syevd 83 12.4 0.010 0.042 19.867 19.869 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 19.107 19.619 apply_single 128 13.6 0.001 0.001 19.107 19.618 ot_diis_step 117 11.6 0.017 0.019 19.580 19.581 cp_fm_cholesky_invert 11 10.9 19.080 19.092 19.080 19.092 cp_fm_upper_to_full 105 14.8 11.416 16.848 11.416 16.848 cp_fm_diag_elpa 83 13.4 0.005 0.007 16.219 16.238 cp_fm_diag_elpa_base 83 14.4 15.182 15.509 16.198 16.213 multiply_cannon_metrocomm3 30084 15.6 0.047 0.049 6.640 15.844 sum_up_and_integrate 128 10.3 0.140 0.151 15.472 15.496 qs_rho_update_rho_low 128 7.7 0.001 0.001 15.345 15.384 calculate_rho_elec 128 8.7 0.173 0.189 15.344 15.383 integrate_v_rspace 128 11.3 0.004 0.004 15.331 15.360 dbcsr_complete_redistribute 395 12.7 1.526 1.650 9.142 12.907 dbcsr_mm_accdrv_process 62264 16.2 7.571 8.611 12.313 12.904 multiply_cannon_sync_h2d 30084 15.6 11.691 12.743 11.691 12.743 make_images_data 5014 15.6 0.063 0.068 10.968 12.687 init_scf_run 11 5.9 0.000 0.002 11.892 11.893 scf_env_initial_rho_setup 11 6.9 0.001 0.002 11.892 11.893 copy_fm_to_dbcsr 209 11.7 0.002 0.002 7.678 11.450 hybrid_alltoall_any 5200 16.5 0.525 2.211 9.789 11.403 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 10.050 10.260 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.362 10.073 pw_transfer 1547 11.6 0.088 0.102 9.422 9.520 mp_alltoall_i22 716 14.1 5.692 9.471 5.692 9.471 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 9.194 9.299 density_rs2pw 128 9.7 0.006 0.007 7.756 9.263 cp_fm_cholesky_decompose 22 10.9 8.769 8.870 8.769 8.870 fft_wrap_pw1pw2_140 523 13.2 1.555 1.586 8.010 8.132 grid_integrate_task_list 128 12.3 7.486 8.043 7.486 8.043 wfi_extrapolate 11 7.9 0.001 0.001 7.943 7.943 rs_pw_transfer 1046 11.9 0.013 0.014 6.038 7.710 multiply_cannon_metrocomm4 25070 15.6 0.084 0.097 2.810 7.373 mp_irecv_dv 76098 16.2 2.657 7.097 2.657 7.097 calculate_dm_sparse 128 9.5 0.001 0.001 6.384 6.475 fft3d_ps 1291 14.7 2.834 2.905 6.119 6.194 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.905 5.931 mp_alltoall_d11v 2415 14.1 5.070 5.864 5.070 5.864 grid_collocate_task_list 128 9.7 5.069 5.716 5.069 5.716 potential_pw2rs 128 12.3 0.023 0.023 5.246 5.301 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.913 5.005 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.796 4.874 qs_energies_init_hamiltonians 11 5.9 0.006 0.022 4.674 4.709 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.337 4.389 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=204.612000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1079.090909, yerr=20.597681 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/77ac92570e643352348b65cb9881e203655a06a9_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410023282688 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 11444707676160 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 92.796579E+12 0.0% 0.0% 100.0% flops max/rank 5.820059E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500928 0.0% 0.0% 100.0% number of processed stacks 1944496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3448.5 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.506415E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 238560 MPI messages size (bytes): total size 1.321104E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.537828E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 112800 59139686400 4194304 < size <= 16777216 104112 545846722560 16777216 < size 20064 716108638608 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8852 52. MP_Alltoall 9584 804353. MP_ISend 39716 2104723. MP_IRecv 39716 2103824. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4001 58222. MP_Allreduce 11000 1175. MP_Sync 86 MP_Alltoall 1700 18828189. MP_SendRecv 3810 122880. MP_ISendRecv 3810 122880. MP_Wait 16000 MP_ISend 10600 423612. MP_IRecv 10600 423612. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.048 0.082 191.977 191.978 qs_mol_dyn_low 1 2.0 0.003 0.003 191.337 191.346 qs_forces 11 3.9 0.004 0.004 191.211 191.217 qs_energies 11 4.9 0.001 0.001 183.514 183.522 scf_env_do_scf 11 5.9 0.001 0.001 164.620 164.622 scf_env_do_scf_inner_loop 116 6.6 0.009 0.010 123.581 123.582 velocity_verlet 10 3.0 0.001 0.001 122.979 122.985 qs_scf_new_mos 116 7.6 0.001 0.001 83.751 83.858 qs_scf_loop_do_ot 116 8.6 0.001 0.001 83.750 83.857 ot_scf_mini 116 9.6 0.003 0.004 79.271 79.378 dbcsr_multiply_generic 2485 12.5 0.190 0.201 76.943 77.632 multiply_cannon 2485 13.5 0.580 0.610 56.165 60.483 multiply_cannon_loop 2485 14.5 0.438 0.450 50.890 52.337 ot_mini 116 10.6 0.001 0.001 41.671 41.795 init_scf_loop 11 6.9 0.000 0.000 40.874 40.876 prepare_preconditioner 11 7.9 0.000 0.000 36.137 36.171 make_preconditioner 11 8.9 0.000 0.000 36.137 36.171 mp_waitall_1 124680 16.7 27.168 35.773 27.168 35.773 make_full_inverse_cholesky 11 9.9 0.000 0.000 33.968 34.256 rebuild_ks_matrix 127 8.3 0.001 0.001 30.913 31.021 qs_ks_build_kohn_sham_matrix 127 9.3 0.018 0.019 30.912 31.021 qs_ks_update_qs_env 127 7.6 0.001 0.001 28.070 28.174 qs_ot_get_p 127 10.4 0.001 0.001 24.733 24.867 qs_ot_get_derivative 116 11.6 0.001 0.002 21.631 21.739 multiply_cannon_multrec 9940 15.5 10.337 14.848 18.225 21.175 cp_fm_cholesky_invert 11 10.9 21.041 21.049 21.041 21.049 qs_ot_p2m_diag 82 11.4 0.490 0.495 20.311 20.334 ot_diis_step 116 11.6 0.020 0.020 19.933 19.934 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.472 19.836 apply_single 127 13.6 0.001 0.001 19.471 19.836 make_m2s 4970 13.5 0.065 0.069 16.768 19.769 multiply_cannon_metrocomm3 9940 15.5 0.022 0.024 12.044 19.572 make_images 4970 14.5 2.268 2.552 16.465 19.467 cp_dbcsr_syevd 82 12.4 0.005 0.005 18.978 18.981 qs_rho_update_rho_low 127 7.7 0.001 0.001 16.213 16.279 calculate_rho_elec 127 8.7 0.254 0.265 16.212 16.278 sum_up_and_integrate 127 10.3 0.181 0.190 15.730 15.781 integrate_v_rspace 127 11.3 0.004 0.004 15.549 15.609 cp_fm_diag_elpa 82 13.4 0.011 0.013 15.313 15.321 cp_fm_diag_elpa_base 82 14.4 15.007 15.129 15.292 15.300 make_images_data 4970 15.5 0.050 0.060 10.119 13.258 hybrid_alltoall_any 5155 16.4 0.794 3.630 9.802 12.828 multiply_cannon_sync_h2d 9940 15.5 11.450 12.137 11.450 12.137 init_scf_run 11 5.9 0.000 0.002 12.041 12.041 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.041 12.041 pw_transfer 1535 11.6 0.086 0.096 9.947 9.990 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 9.724 9.770 cp_fm_cholesky_decompose 22 10.9 9.354 9.447 9.354 9.447 qs_ot_get_derivative_diag 76 12.4 0.002 0.003 8.701 8.753 multiply_cannon_metrocomm1 9940 15.5 0.028 0.029 5.324 8.744 density_rs2pw 127 9.7 0.006 0.006 7.686 8.573 fft_wrap_pw1pw2_140 519 13.2 1.888 1.923 8.388 8.439 mp_allgather_i34 2485 14.5 3.488 8.348 3.488 8.348 dbcsr_mm_accdrv_process 20590 16.1 2.773 3.601 7.524 8.236 grid_integrate_task_list 127 12.3 7.641 8.214 7.641 8.214 wfi_extrapolate 11 7.9 0.001 0.001 7.869 7.869 mp_alltoall_d11v 2401 14.1 5.687 6.902 5.687 6.902 calculate_dm_sparse 127 9.5 0.001 0.001 6.524 6.623 dbcsr_complete_redistribute 393 12.7 2.470 2.543 5.863 6.331 fft3d_ps 1281 14.7 2.745 2.826 6.240 6.281 rs_pw_transfer 1038 11.9 0.012 0.013 5.138 6.065 grid_collocate_task_list 127 9.7 5.351 6.007 5.351 6.007 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.949 5.960 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.261 5.262 potential_pw2rs 127 12.3 0.026 0.026 5.053 5.076 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.438 4.489 multiply_cannon_metrocomm4 7455 15.5 0.024 0.027 1.860 4.346 copy_fm_to_dbcsr 208 11.6 0.001 0.002 3.946 4.343 mp_irecv_dv 28618 15.9 1.825 4.264 1.825 4.264 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.950 4.006 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.916 3.977 calculate_first_density_matrix 1 7.0 0.000 0.000 3.938 3.941 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.578 3.843 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=191.978000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1407.000000, yerr=42.661458 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/77ac92570e643352348b65cb9881e203655a06a9_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 1971389988864 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 1992007581696 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 15239176888320 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176888320 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 94.241596E+12 0.0% 0.0% 100.0% flops max/rank 11.787141E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806515136 0.0% 0.0% 100.0% number of processed stacks 1981760 0.0% 0.0% 100.0% average stack size 0.0 0.0 3434.6 marketing flops 145.659521E+12 ------------------------------------------------------------------------------- # multiplications 2533 max memory usage/rank 2.948051E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101320 MPI messages size (bytes): total size 1.145104E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.301851E+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 45808 35567697920 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592330608 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4065 58259. MP_Allreduce 11160 1510. MP_Sync 86 MP_Alltoall 1724 36993608. 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 0.031 0.094 323.153 323.191 qs_mol_dyn_low 1 2.0 0.003 0.003 322.473 322.481 qs_forces 11 3.9 0.005 0.005 322.365 322.366 qs_energies 11 4.9 0.001 0.001 312.776 312.782 scf_env_do_scf 11 5.9 0.001 0.001 289.056 289.064 velocity_verlet 10 3.0 0.001 0.001 227.263 227.270 scf_env_do_scf_inner_loop 118 6.6 0.090 0.470 158.851 158.853 init_scf_loop 11 6.9 0.000 0.000 129.931 129.936 prepare_preconditioner 11 7.9 0.000 0.000 124.583 124.618 make_preconditioner 11 8.9 0.000 0.000 124.583 124.618 make_full_inverse_cholesky 11 9.9 0.000 0.000 100.864 121.650 qs_scf_new_mos 118 7.6 0.001 0.001 104.346 104.498 qs_scf_loop_do_ot 118 8.6 0.001 0.001 104.345 104.497 ot_scf_mini 118 9.6 0.004 0.004 99.354 99.495 dbcsr_multiply_generic 2533 12.6 0.225 0.238 90.250 91.333 cp_fm_upper_to_full 104 14.8 52.175 74.862 52.175 74.862 multiply_cannon 2533 13.6 0.711 0.752 62.674 63.521 multiply_cannon_loop 2533 14.6 0.473 0.482 58.162 60.679 ot_mini 118 10.6 0.001 0.001 49.783 49.925 dbcsr_complete_redistribute 393 12.7 4.003 4.045 29.727 42.353 rebuild_ks_matrix 129 8.3 0.001 0.001 40.476 40.646 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.018 40.476 40.646 copy_fm_to_dbcsr 208 11.6 0.002 0.002 26.169 38.758 cp_fm_cholesky_invert 11 10.9 37.784 37.792 37.784 37.792 qs_ks_update_qs_env 129 7.6 0.001 0.001 37.339 37.500 mp_waitall_1 104740 16.8 32.227 36.905 32.227 36.905 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 23.673 36.135 mp_alltoall_i22 716 14.1 21.505 34.126 21.505 34.126 qs_ot_get_p 129 10.4 0.001 0.001 33.300 33.411 qs_ot_p2m_diag 82 11.4 0.868 0.874 28.466 28.509 qs_ot_get_derivative 118 11.6 0.002 0.002 28.101 28.238 cp_dbcsr_syevd 82 12.4 0.006 0.006 26.560 26.563 make_m2s 5066 13.6 0.077 0.080 22.203 23.817 make_images 5066 14.6 3.819 3.988 21.720 23.337 multiply_cannon_metrocomm3 10132 15.6 0.023 0.024 20.680 22.963 cp_fm_diag_elpa 82 13.4 0.013 0.013 22.791 22.792 cp_fm_diag_elpa_base 82 14.4 18.316 19.981 22.772 22.773 qs_rho_update_rho_low 129 7.7 0.001 0.001 22.125 22.168 calculate_rho_elec 129 8.7 0.485 0.485 22.125 22.167 ot_diis_step 118 11.6 0.022 0.023 21.624 21.625 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 20.696 21.121 apply_single 129 13.6 0.001 0.001 20.696 21.121 sum_up_and_integrate 129 10.3 0.323 0.326 21.040 21.119 integrate_v_rspace 129 11.3 0.004 0.004 20.717 20.797 multiply_cannon_multrec 10132 15.6 10.713 12.585 18.872 19.231 multiply_cannon_sync_h2d 10132 15.6 15.783 15.803 15.783 15.803 make_images_data 5066 15.6 0.059 0.065 12.427 15.072 hybrid_alltoall_any 5251 16.5 1.308 3.061 12.212 14.878 init_scf_run 11 5.9 0.000 0.001 13.458 13.459 scf_env_initial_rho_setup 11 6.9 0.008 0.009 13.458 13.459 pw_transfer 1559 11.6 0.094 0.095 12.940 12.948 fft_wrap_pw1pw2 1301 12.7 0.011 0.011 12.703 12.713 fft_wrap_pw1pw2_140 527 13.2 3.097 3.119 11.103 11.112 cp_fm_cholesky_decompose 22 10.9 10.684 10.701 10.684 10.701 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 10.561 10.668 mp_alltoall_d11v 2417 14.1 9.462 10.417 9.462 10.417 dbcsr_mm_accdrv_process 20950 16.1 4.202 6.502 7.915 10.029 wfi_extrapolate 11 7.9 0.001 0.001 9.616 9.616 density_rs2pw 129 9.7 0.006 0.006 9.253 9.390 grid_integrate_task_list 129 12.3 8.644 8.829 8.644 8.829 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 8.175 8.176 fft3d_ps 1301 14.7 2.868 2.878 7.743 7.797 calculate_dm_sparse 129 9.5 0.001 0.001 6.939 6.970 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.781 6.846 rs_scatter_matrices 140 9.7 3.782 4.730 6.448 6.732 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=323.191000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2658.000000, yerr=150.692342 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/77ac92570e643352348b65cb9881e203655a06a9_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 1.261670E+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 255669. MP_Allreduce 3059 6274. MP_Sync 4 MP_Alltoall 54 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 0.038 0.061 85.452 85.453 qs_energies 1 2.0 0.000 0.000 84.876 84.894 ls_scf 1 3.0 0.000 0.000 83.975 83.994 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.732 72.934 multiply_cannon 111 7.7 0.017 0.020 56.011 57.351 multiply_cannon_loop 111 8.7 0.206 0.219 52.575 54.147 ls_scf_main 1 4.0 0.000 0.000 52.292 52.292 density_matrix_trs4 2 5.0 0.002 0.003 46.776 46.859 ls_scf_init_scf 1 4.0 0.000 0.000 28.619 28.620 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.531 27.591 mp_waitall_1 11031 10.9 22.704 25.725 22.704 25.725 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.426 25.446 multiply_cannon_multrec 2664 9.7 8.111 8.873 15.548 17.300 multiply_cannon_sync_h2d 2664 9.7 13.514 15.715 13.514 15.715 make_m2s 222 7.7 0.008 0.011 13.076 13.668 make_images 222 8.7 0.099 0.110 13.053 13.648 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 9.820 12.554 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.548 9.455 make_images_data 222 9.7 0.004 0.005 7.673 8.203 dbcsr_mm_accdrv_process 4760 10.4 0.517 0.652 7.057 8.110 hybrid_alltoall_any 227 10.6 0.215 1.822 6.566 7.781 dbcsr_mm_accdrv_process_sort 4760 11.4 6.325 7.186 6.325 7.186 calculate_norms 4752 9.8 5.491 6.137 5.491 6.137 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.049 5.197 mp_sum_l 807 5.4 3.110 4.449 3.110 4.449 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.052 4.022 mp_irecv_dv 6231 10.9 2.034 3.989 2.034 3.989 make_images_sizes 222 9.7 0.000 0.000 0.688 3.306 mp_alltoall_i44 222 10.7 0.688 3.306 0.688 3.306 arnoldi_extremal 4 6.8 0.000 0.000 3.242 3.270 arnoldi_normal_ev 4 7.8 0.005 0.013 3.242 3.270 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.324 3.261 build_subspace 16 8.4 0.009 0.012 3.134 3.136 ls_scf_post 1 4.0 0.000 0.000 3.064 3.082 ls_scf_store_result 1 5.0 0.000 0.000 2.874 2.921 dbcsr_special_finalize 555 9.7 0.005 0.006 2.365 2.803 dbcsr_merge_single_wm 555 10.7 0.461 0.608 2.357 2.795 make_images_pack 222 9.7 2.203 2.647 2.205 2.649 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.315 2.575 dbcsr_sort_data 658 11.4 2.152 2.506 2.152 2.506 dbcsr_matrix_vector_mult_local 304 10.0 2.069 2.480 2.071 2.482 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.272 2.331 buffer_matrices_ensure_size 222 8.7 1.755 2.063 1.755 2.063 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.790 1.792 rebuild_ks_matrix 3 7.3 0.000 0.000 1.781 1.782 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.781 1.782 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.453000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1126.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/77ac92570e643352348b65cb9881e203655a06a9_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 2.110861E+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 266696. MP_Allreduce 3058 10339. 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 0.030 0.056 92.506 92.515 qs_energies 1 2.0 0.000 0.000 91.867 91.870 ls_scf 1 3.0 0.000 0.000 90.433 90.437 dbcsr_multiply_generic 111 6.7 0.015 0.015 76.023 76.405 multiply_cannon 111 7.7 0.028 0.062 53.734 57.348 ls_scf_main 1 4.0 0.000 0.000 55.716 55.721 multiply_cannon_loop 111 8.7 0.115 0.123 50.437 54.039 density_matrix_trs4 2 5.0 0.002 0.012 49.888 50.075 mp_waitall_1 9105 10.9 22.120 31.253 22.120 31.253 ls_scf_init_scf 1 4.0 0.000 0.000 31.130 31.131 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.875 29.981 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.462 27.475 multiply_cannon_multrec 1332 9.7 13.041 16.775 22.389 27.381 multiply_cannon_metrocomm3 1332 9.7 0.006 0.007 12.213 21.388 make_m2s 222 7.7 0.006 0.007 15.736 16.397 make_images 222 8.7 1.574 1.964 15.707 16.367 dbcsr_mm_accdrv_process 4041 10.4 0.295 0.517 8.949 10.520 make_images_data 222 9.7 0.004 0.004 9.200 10.194 dbcsr_mm_accdrv_process_sort 4041 11.4 8.519 10.069 8.519 10.069 hybrid_alltoall_any 227 10.6 0.520 2.465 8.636 9.706 mp_sum_l 807 5.4 5.273 8.868 5.273 8.868 multiply_cannon_metrocomm4 1221 9.7 0.007 0.008 3.233 7.826 mp_irecv_dv 3311 11.0 3.214 7.768 3.214 7.768 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.092 7.113 calculate_norms 2376 9.8 5.983 6.801 5.983 6.801 multiply_cannon_sync_h2d 1332 9.7 4.717 6.103 4.717 6.103 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.110 5.297 arnoldi_extremal 4 6.8 0.000 0.000 4.870 4.890 arnoldi_normal_ev 4 7.8 0.001 0.005 4.869 4.890 build_subspace 16 8.4 0.014 0.021 4.607 4.610 ls_scf_post 1 4.0 0.000 0.000 3.588 3.592 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.230 3.503 ls_scf_store_result 1 5.0 0.000 0.000 3.268 3.410 dbcsr_matrix_vector_mult_local 304 10.0 2.745 3.235 2.748 3.237 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 1.328 3.140 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.654 2.765 mp_allgather_i34 111 8.7 0.996 2.444 0.996 2.444 make_images_pack 222 9.7 2.023 2.427 2.026 2.429 dbcsr_sort_data 436 11.2 1.829 2.106 1.829 2.106 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.900 1.902 rebuild_ks_matrix 3 7.3 0.000 0.000 1.887 1.889 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.887 1.889 dbcsr_data_new 4174 10.1 1.607 1.868 1.607 1.868 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=92.515000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1762.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/77ac92570e643352348b65cb9881e203655a06a9_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 2.777514E+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 265470. MP_Allreduce 3058 11181. 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 0.040 0.089 95.764 95.774 qs_energies 1 2.0 0.000 0.000 94.911 94.923 ls_scf 1 3.0 0.003 0.012 93.013 93.029 dbcsr_multiply_generic 111 6.7 0.015 0.016 77.293 77.594 ls_scf_main 1 4.0 0.001 0.012 58.157 58.165 multiply_cannon 111 7.7 0.037 0.060 53.308 57.516 multiply_cannon_loop 111 8.7 0.100 0.113 49.690 53.544 density_matrix_trs4 2 5.0 0.003 0.017 52.136 52.253 mp_waitall_1 7281 11.0 24.457 35.045 24.457 35.045 ls_scf_init_scf 1 4.0 0.001 0.003 31.258 31.260 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.864 29.933 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.413 27.430 multiply_cannon_multrec 888 9.7 12.622 15.229 21.260 24.470 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.332 23.631 make_m2s 222 7.7 0.006 0.006 17.264 18.498 make_images 222 8.7 1.970 2.282 17.226 18.461 make_images_data 222 9.7 0.003 0.004 10.040 11.051 hybrid_alltoall_any 227 10.6 0.623 2.860 9.716 11.024 dbcsr_mm_accdrv_process 3754 10.4 0.246 0.414 8.171 9.422 dbcsr_mm_accdrv_process_sort 3754 11.4 7.801 9.008 7.801 9.008 mp_sum_l 807 5.4 5.186 9.006 5.186 9.006 multiply_cannon_sync_h2d 888 9.7 6.028 7.417 6.028 7.417 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.020 7.069 multiply_cannon_metrocomm1 888 9.7 0.002 0.003 3.704 6.905 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.432 6.593 mp_irecv_dv 2335 11.1 2.416 6.546 2.416 6.546 arnoldi_extremal 4 6.8 0.000 0.000 5.190 5.208 arnoldi_normal_ev 4 7.8 0.001 0.005 5.190 5.208 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.963 5.152 build_subspace 16 8.4 0.014 0.020 4.885 4.890 calculate_norms 1584 9.8 4.294 4.703 4.294 4.703 mp_allgather_i34 111 8.7 1.455 3.959 1.455 3.959 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.514 3.806 ls_scf_post 1 4.0 0.003 0.012 3.595 3.611 dbcsr_matrix_vector_mult_local 304 10.0 3.023 3.594 3.026 3.596 ls_scf_store_result 1 5.0 0.000 0.000 3.306 3.407 ls_scf_dm_to_ks 2 5.0 0.000 0.007 2.874 2.984 dbcsr_sort_data 325 11.1 1.934 2.123 1.934 2.123 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.091 2.093 make_images_pack 222 9.7 1.829 2.089 1.831 2.092 rebuild_ks_matrix 3 7.3 0.000 0.000 2.073 2.075 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.014 2.073 2.075 make_images_sizes 222 9.7 0.000 0.000 0.846 2.065 mp_alltoall_i44 222 10.7 0.846 2.065 0.846 2.065 dbcsr_data_release 9322 10.9 1.298 1.970 1.298 1.970 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=95.774000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2172.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/77ac92570e643352348b65cb9881e203655a06a9_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 3.346166E+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 266696. MP_Allreduce 3058 13371. 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 0.043 0.074 99.847 99.848 qs_energies 1 2.0 0.000 0.000 98.385 98.400 ls_scf 1 3.0 0.006 0.023 96.577 96.597 dbcsr_multiply_generic 111 6.7 0.017 0.018 79.433 79.658 ls_scf_main 1 4.0 0.005 0.036 59.867 59.874 multiply_cannon 111 7.7 0.068 0.114 52.087 56.389 density_matrix_trs4 2 5.0 0.006 0.035 53.627 53.731 multiply_cannon_loop 111 8.7 0.114 0.124 47.003 50.312 ls_scf_init_scf 1 4.0 0.000 0.002 33.415 33.424 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.739 31.836 mp_waitall_1 6369 11.0 23.590 30.584 23.590 30.584 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.180 29.195 multiply_cannon_multrec 1332 9.7 14.091 17.068 22.040 24.502 make_m2s 222 7.7 0.007 0.008 21.532 23.012 make_images 222 8.7 3.138 3.600 21.481 22.965 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.721 17.467 make_images_data 222 9.7 0.004 0.004 12.142 13.780 hybrid_alltoall_any 227 10.6 0.797 3.747 11.466 13.485 dbcsr_mm_accdrv_process 3641 10.4 0.188 0.364 7.595 9.146 dbcsr_mm_accdrv_process_sort 3641 11.4 7.223 8.724 7.223 8.724 mp_sum_l 807 5.4 4.230 7.375 4.230 7.375 multiply_cannon_sync_h2d 1332 9.7 5.486 6.400 5.486 6.400 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.076 6.064 mp_irecv_dv 3229 10.9 2.053 5.993 2.053 5.993 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.230 5.711 arnoldi_extremal 4 6.8 0.000 0.000 5.489 5.500 arnoldi_normal_ev 4 7.8 0.001 0.005 5.489 5.500 build_subspace 16 8.4 0.014 0.021 5.153 5.160 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.554 5.010 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.624 4.823 mp_allgather_i34 111 8.7 2.204 4.669 2.204 4.669 calculate_norms 2376 9.8 4.182 4.488 4.182 4.488 dbcsr_matrix_vector_mult 304 9.0 0.010 0.020 3.708 4.047 dbcsr_matrix_vector_mult_local 304 10.0 3.195 3.685 3.197 3.687 dbcsr_sort_data 658 11.4 3.089 3.472 3.089 3.472 ls_scf_post 1 4.0 0.004 0.018 3.290 3.304 dbcsr_special_finalize 555 9.7 0.006 0.007 2.835 3.262 dbcsr_merge_single_wm 555 10.7 0.539 0.668 2.826 3.254 ls_scf_dm_to_ks 2 5.0 0.000 0.006 3.032 3.098 ls_scf_store_result 1 5.0 0.000 0.000 3.014 3.064 dbcsr_data_release 10477 10.7 1.573 2.365 1.573 2.365 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.147 2.148 rebuild_ks_matrix 3 7.3 0.000 0.000 2.124 2.125 qs_ks_build_kohn_sham_matrix 3 8.3 0.010 0.030 2.124 2.125 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=99.848000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2747.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/77ac92570e643352348b65cb9881e203655a06a9_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 4.686586E+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 265558. MP_Allreduce 3049 15663. 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 0.041 0.066 100.560 100.564 qs_energies 1 2.0 0.000 0.000 99.592 99.596 ls_scf 1 3.0 0.000 0.000 97.401 97.404 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.653 78.865 ls_scf_main 1 4.0 0.000 0.000 62.646 62.647 multiply_cannon 111 7.7 0.077 0.176 55.934 61.297 density_matrix_trs4 2 5.0 0.002 0.003 55.398 55.470 multiply_cannon_loop 111 8.7 0.070 0.082 51.417 52.962 mp_waitall_1 5436 11.0 27.133 33.205 27.133 33.205 ls_scf_init_scf 1 4.0 0.000 0.000 31.075 31.081 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.777 29.812 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.581 27.588 multiply_cannon_multrec 444 9.7 14.078 16.582 21.212 23.603 make_m2s 222 7.7 0.004 0.005 17.889 20.387 make_images 222 8.7 3.724 4.406 17.827 20.328 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.587 16.819 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.232 15.394 make_images_data 222 9.7 0.003 0.004 10.112 12.576 hybrid_alltoall_any 227 10.6 0.788 3.772 9.945 12.389 multiply_cannon_sync_h2d 444 9.7 6.579 8.136 6.579 8.136 dbcsr_mm_accdrv_process 3003 10.4 0.200 0.345 6.837 7.982 dbcsr_mm_accdrv_process_sort 3003 11.4 6.520 7.634 6.520 7.634 mp_allgather_i34 111 8.7 2.744 6.970 2.744 6.970 arnoldi_extremal 4 6.8 0.000 0.000 6.002 6.010 arnoldi_normal_ev 4 7.8 0.006 0.030 6.002 6.010 build_subspace 16 8.4 0.015 0.020 5.575 5.582 mp_sum_l 807 5.4 2.899 4.932 2.899 4.932 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.739 4.882 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.290 4.439 dbcsr_matrix_vector_mult_local 304 10.0 3.700 4.175 3.702 4.177 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.608 3.850 mp_irecv_dv 1241 11.2 1.589 3.826 1.589 3.826 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.986 3.729 make_images_sizes 222 9.7 0.000 0.000 1.131 3.688 calculate_norms 792 9.8 3.535 3.688 3.535 3.688 mp_alltoall_i44 222 10.7 1.131 3.688 1.131 3.688 ls_scf_post 1 4.0 0.000 0.000 3.680 3.683 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.549 3.668 ls_scf_store_result 1 5.0 0.000 0.000 3.439 3.498 dbcsr_finalize 304 7.8 0.062 0.078 2.199 2.270 qs_energies_init_hamiltonians 1 3.0 0.000 0.001 2.175 2.176 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.123 2.125 dbcsr_merge_all 275 8.9 0.474 0.518 2.048 2.114 rebuild_ks_matrix 3 7.3 0.000 0.000 2.090 2.092 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.090 2.092 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=100.564000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3680.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/77ac92570e643352348b65cb9881e203655a06a9_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 8.712413E+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 284111. MP_Allreduce 3043 21950. 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 0.092 0.115 109.104 109.113 qs_energies 1 2.0 0.000 0.000 107.465 107.478 ls_scf 1 3.0 0.000 0.000 104.320 104.332 dbcsr_multiply_generic 111 6.7 0.023 0.028 77.777 77.936 ls_scf_main 1 4.0 0.000 0.000 66.298 66.299 density_matrix_trs4 2 5.0 0.026 0.195 57.130 57.187 multiply_cannon 111 7.7 0.126 0.158 50.029 51.994 multiply_cannon_loop 111 8.7 0.067 0.069 46.553 47.489 ls_scf_init_scf 1 4.0 0.000 0.000 34.339 34.340 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.782 32.797 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.969 29.979 mp_waitall_1 4527 11.1 22.224 26.085 22.224 26.085 make_m2s 222 7.7 0.005 0.005 23.909 24.926 make_images 222 8.7 4.600 5.008 23.802 24.818 multiply_cannon_multrec 444 9.7 17.888 18.670 22.542 23.117 hybrid_alltoall_any 227 10.6 1.663 3.650 12.931 15.598 make_images_data 222 9.7 0.003 0.003 13.133 15.595 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.497 11.047 multiply_cannon_sync_h2d 444 9.7 8.852 8.906 8.852 8.906 arnoldi_extremal 4 6.8 0.000 0.000 7.475 7.503 arnoldi_normal_ev 4 7.8 0.019 0.046 7.475 7.502 build_subspace 16 8.4 0.026 0.036 6.865 6.875 dbcsr_matrix_vector_mult 304 9.0 0.017 0.033 5.474 5.645 dbcsr_matrix_vector_mult_local 304 10.0 5.042 5.374 5.044 5.376 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.994 5.237 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.137 5.229 dbcsr_mm_accdrv_process 1814 10.4 0.229 0.316 4.490 4.626 dbcsr_mm_accdrv_process_sort 1814 11.4 4.193 4.332 4.193 4.332 ls_scf_post 1 4.0 0.000 0.000 3.683 3.695 mp_allgather_i34 111 8.7 1.127 3.569 1.127 3.569 make_images_sizes 222 9.7 0.000 0.000 1.421 3.474 mp_alltoall_i44 222 10.7 1.421 3.474 1.421 3.474 ls_scf_store_result 1 5.0 0.000 0.000 3.425 3.438 calculate_norms 792 9.8 3.236 3.272 3.236 3.272 dbcsr_finalize 304 7.8 0.082 0.089 3.093 3.182 qs_energies_init_hamiltonians 1 3.0 0.012 0.018 3.115 3.115 dbcsr_merge_all 275 8.9 0.891 0.919 2.877 2.962 dbcsr_complete_redistribute 5 7.6 1.434 1.466 2.771 2.889 dbcsr_data_release 12724 10.6 2.341 2.843 2.341 2.843 matrix_ls_to_qs 2 6.0 0.000 0.000 2.414 2.533 dbcsr_sort_data 325 11.1 2.444 2.514 2.444 2.514 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.478 2.479 rebuild_ks_matrix 3 7.3 0.000 0.000 2.411 2.413 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.411 2.413 dbcsr_new_transposed 4 7.5 0.257 0.290 2.366 2.390 dbcsr_frobenius_norm 74 6.6 2.054 2.122 2.182 2.231 dbcsr_add_d 103 6.2 0.000 0.000 2.132 2.200 dbcsr_add_anytype 103 7.2 0.859 0.890 2.132 2.199 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=109.113000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6903.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/77ac92570e643352348b65cb9881e203655a06a9_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 590.204928E+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 76559. MP_Allreduce 13152 2094. MP_Sync 1064 MP_Alltoall 2588 897994675. 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 0.084 0.304 249.992 250.008 qs_mol_dyn_low 1 2.0 0.005 0.038 248.087 248.109 qs_forces 5 3.8 0.019 0.035 247.963 247.980 qs_energies 5 4.8 0.002 0.041 244.720 244.754 scf_env_do_scf 5 5.8 0.002 0.024 229.283 229.297 scf_env_do_scf_inner_loop 105 6.6 0.004 0.022 199.345 199.360 qs_scf_new_mos 105 7.6 0.001 0.002 153.702 153.966 qs_scf_loop_do_ot 105 8.6 0.001 0.008 153.702 153.965 ot_scf_mini 105 9.6 0.003 0.017 143.992 144.156 dbcsr_multiply_generic 1445 12.2 0.126 0.158 132.822 133.333 multiply_cannon 1445 13.2 0.276 0.305 113.035 115.520 multiply_cannon_loop 1445 14.2 2.366 2.493 110.951 112.178 velocity_verlet 4 3.0 0.001 0.011 110.408 110.416 ot_mini 105 10.6 0.002 0.045 60.028 60.146 qs_ot_get_p 112 10.4 0.001 0.001 54.137 54.481 multiply_cannon_multrec 69360 15.2 31.825 37.708 40.416 45.969 qs_ot_p2m_diag 40 11.0 0.020 0.031 42.837 42.957 mp_waitall_1 488190 16.1 34.827 42.533 34.827 42.533 cp_dbcsr_syevd 40 12.0 0.004 0.050 39.387 39.390 qs_ot_get_derivative 55 11.6 0.001 0.001 38.308 38.448 multiply_cannon_sync_h2d 69360 15.2 30.722 34.811 30.722 34.811 cp_fm_syevd 40 13.0 0.013 0.031 33.654 33.817 multiply_cannon_metrocomm3 69360 15.2 0.191 0.200 24.489 32.801 rebuild_ks_matrix 110 8.4 0.000 0.000 31.328 31.499 qs_ks_build_kohn_sham_matrix 110 9.4 0.014 0.047 31.327 31.499 init_scf_loop 7 6.6 0.002 0.026 29.890 29.929 qs_ks_update_qs_env 112 7.6 0.001 0.001 28.842 28.999 cp_fm_redistribute_end 40 14.0 14.291 28.498 14.298 28.501 cp_fm_syevd_base 40 14.0 14.191 28.402 14.191 28.402 prepare_preconditioner 7 7.6 0.000 0.000 24.422 24.455 make_preconditioner 7 8.6 0.000 0.006 24.422 24.455 apply_preconditioner_dbcsr 62 12.6 0.000 0.001 22.940 23.196 apply_single 62 13.6 0.000 0.000 22.940 23.196 qs_rho_update_rho_low 110 7.6 0.001 0.009 21.075 21.384 calculate_rho_elec 110 8.6 0.031 0.041 21.075 21.383 ot_new_cg_direction 55 11.6 0.004 0.119 21.043 21.046 rs_pw_transfer 690 11.5 0.011 0.039 16.431 17.601 density_rs2pw 110 9.6 0.006 0.008 15.024 16.192 make_full_inverse_cholesky 7 9.6 0.000 0.001 16.107 16.177 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.606 14.810 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 13.761 13.869 pw_transfer 1645 12.4 0.083 0.103 12.827 13.072 mp_sum_l 4684 12.4 12.113 12.994 12.113 12.994 fft_wrap_pw1pw2 1425 13.5 0.013 0.016 12.684 12.933 init_scf_run 5 5.8 0.000 0.001 12.561 12.562 scf_env_initial_rho_setup 5 6.8 0.001 0.003 12.561 12.562 calculate_dm_sparse 110 9.5 0.000 0.001 11.689 11.858 cp_fm_cholesky_invert 7 10.6 10.998 11.014 10.998 11.014 fft_wrap_pw1pw2_240 915 15.0 1.189 1.287 10.670 10.908 qs_vxc_create 110 10.4 0.011 0.071 10.767 10.818 sum_up_and_integrate 60 10.3 0.028 0.032 10.496 10.508 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 10.422 10.485 integrate_v_rspace 60 11.3 0.002 0.002 10.468 10.481 check_diag 80 13.5 8.604 8.905 10.221 10.419 dbcsr_mm_accdrv_process 154766 15.8 4.007 4.159 8.460 9.407 fft3d_pb 915 16.0 2.386 2.640 8.732 9.025 make_full_single_inverse 7 9.6 0.013 0.269 8.009 8.042 make_m2s 2890 13.2 0.076 0.084 7.247 7.801 xc_rho_set_and_dset_create 110 12.4 0.079 0.103 7.502 7.760 make_images 2890 14.2 0.242 0.260 7.144 7.697 xc_vxc_pw_create 60 11.3 0.039 0.049 7.252 7.311 calculate_first_density_matrix 1 7.0 0.000 0.004 7.262 7.277 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.016 6.899 6.918 potential_pw2rs 60 12.3 0.003 0.003 6.562 6.622 multiply_cannon_metrocomm1 69360 15.2 0.090 0.095 4.092 6.620 xc_pw_derive 510 13.4 0.005 0.007 6.470 6.552 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.001 6.436 6.482 acc_transpose_blocks 69360 15.2 0.361 0.379 6.064 6.386 mp_alltoall_z22v 2340 17.7 5.836 6.212 5.836 6.212 rs_pw_transfer_RS2PW_30 110 11.6 1.598 1.691 5.197 5.891 mp_waitany 7680 13.5 4.618 5.738 4.618 5.738 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=250.008000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=559.800000, yerr=2.993326 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 77ac92570e643352348b65cb9881e203655a06a9 Summary: empty Status: OK