=== 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: 8f18a4e0cc0e0e7c2e20248e70b07b7bca2af436 ################# 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 7.3.2, 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 (12.01.2023) # # \ 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=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.1.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.1 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.3.2 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(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/8f18a4e0cc0e0e7c2e20248e70b07b7bca2af436_performance_tests/01 job id: 44488280 --- 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/8f18a4e0cc0e0e7c2e20248e70b07b7bca2af436_performance_tests/02 job id: 44488281 --- 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/8f18a4e0cc0e0e7c2e20248e70b07b7bca2af436_performance_tests/03 job id: 44488282 --- 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/8f18a4e0cc0e0e7c2e20248e70b07b7bca2af436_performance_tests/04 job id: 44488283 --- 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/8f18a4e0cc0e0e7c2e20248e70b07b7bca2af436_performance_tests/05 job id: 44488284 --- 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/8f18a4e0cc0e0e7c2e20248e70b07b7bca2af436_performance_tests/06 job id: 44488285 --- 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/8f18a4e0cc0e0e7c2e20248e70b07b7bca2af436_performance_tests/07 job id: 44488286 --- 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/8f18a4e0cc0e0e7c2e20248e70b07b7bca2af436_performance_tests/08 job id: 44488287 --- 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/8f18a4e0cc0e0e7c2e20248e70b07b7bca2af436_performance_tests/09 job id: 44488288 --- 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/8f18a4e0cc0e0e7c2e20248e70b07b7bca2af436_performance_tests/10 job id: 44488289 --- 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/8f18a4e0cc0e0e7c2e20248e70b07b7bca2af436_performance_tests/11 job id: 44488290 --- 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/8f18a4e0cc0e0e7c2e20248e70b07b7bca2af436_performance_tests/12 job id: 44488291 --- 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/8f18a4e0cc0e0e7c2e20248e70b07b7bca2af436_performance_tests/13 job id: 44488292 --- 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/8f18a4e0cc0e0e7c2e20248e70b07b7bca2af436_performance_tests/14 job id: 44488293 --- 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/8f18a4e0cc0e0e7c2e20248e70b07b7bca2af436_performance_tests/15 job id: 44488294 --- 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/8f18a4e0cc0e0e7c2e20248e70b07b7bca2af436_performance_tests/16 job id: 44488295 --- 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/8f18a4e0cc0e0e7c2e20248e70b07b7bca2af436_performance_tests/17 job id: 44488296 --- 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/8f18a4e0cc0e0e7c2e20248e70b07b7bca2af436_performance_tests/18 job id: 44488297 --- 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/8f18a4e0cc0e0e7c2e20248e70b07b7bca2af436_performance_tests/19 job id: 44488298 --- 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/8f18a4e0cc0e0e7c2e20248e70b07b7bca2af436_performance_tests/20 job id: 44488299 --- 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/8f18a4e0cc0e0e7c2e20248e70b07b7bca2af436_performance_tests/21 job id: 44488300 --- 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/8f18a4e0cc0e0e7c2e20248e70b07b7bca2af436_performance_tests/22 job id: 44488301 --- 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/8f18a4e0cc0e0e7c2e20248e70b07b7bca2af436_performance_tests/23 job id: 44488302 --- 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/8f18a4e0cc0e0e7c2e20248e70b07b7bca2af436_performance_tests/24 job id: 44488303 --- 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/8f18a4e0cc0e0e7c2e20248e70b07b7bca2af436_performance_tests/25 job id: 44488304 --- 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/8f18a4e0cc0e0e7c2e20248e70b07b7bca2af436_performance_tests/26 job id: 44488305 --- 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 ~~~~~~~ === 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]" === END PLOTS (description) === ============ RESULTS ============ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8f18a4e0cc0e0e7c2e20248e70b07b7bca2af436_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 0.021 0.035 155.668 155.669 farming_run 1 2.0 155.015 155.016 155.639 155.641 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.459290E+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 228 1113141. MP_Allreduce 485 2282278. MP_Sync 27 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 8 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 0.010 0.025 134.996 134.997 qs_energies 1 2.0 0.000 0.000 116.646 116.648 mp2_main 1 3.0 0.000 0.000 114.339 114.342 mp2_gpw_main 1 4.0 0.022 0.029 113.251 113.253 mp2_ri_gpw_compute_in 1 5.0 0.178 0.189 94.241 94.670 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 55.476 55.903 mp2_eri_3c_integrate_gpw 272 7.0 0.152 0.168 41.786 47.180 get_2c_integrals 1 6.0 0.000 0.001 37.948 38.587 integrate_v_rspace 273 8.0 0.436 0.450 25.138 30.339 pw_transfer 6555 10.6 0.369 0.382 27.509 28.214 fft_wrap_pw1pw2 5465 11.4 0.044 0.047 26.150 26.695 grid_integrate_task_list 273 9.0 20.939 26.636 20.939 26.636 fft_wrap_pw1pw2_100 2178 12.4 1.234 1.461 23.667 24.199 compute_2c_integrals 1 7.0 0.013 0.017 19.852 19.861 compute_2c_integrals_loop_lm 1 8.0 0.014 0.021 18.994 19.470 mp2_eri_2c_integrate_gpw 1 9.0 2.362 2.426 18.980 19.458 rpa_ri_compute_en 1 5.0 0.001 0.002 18.899 19.106 cp_fm_cholesky_decompose 12 8.2 18.137 18.803 18.137 18.803 mp_max_i 395 1.7 6.917 18.105 6.917 18.105 mp2_env_release 1 2.0 11.185 18.091 11.185 18.091 cholesky_decomp 1 7.0 0.000 0.000 16.940 17.586 fft3d_s 5443 13.4 16.211 16.626 16.233 16.646 ao_to_mo_and_store_B_mult_1 272 7.0 10.857 15.596 10.857 15.596 calculate_wavefunction 272 8.0 5.426 5.576 12.576 13.225 rpa_num_int 1 6.0 0.001 0.012 10.576 10.585 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.545 10.561 calc_mat_Q 8 8.0 0.000 0.000 9.353 9.447 contract_S_to_Q 8 9.0 0.000 0.000 8.774 8.869 calc_potential_gpw 544 9.5 0.005 0.006 8.284 8.702 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.275 8.606 potential_pw2rs 545 10.0 0.107 0.109 7.734 8.490 parallel_gemm_fm 14 9.1 0.000 0.000 8.356 8.436 parallel_gemm_fm_cosma 14 10.1 8.356 8.436 8.356 8.436 create_integ_mat 1 6.0 0.014 0.027 7.801 7.810 collocate_single_gaussian 272 10.0 0.039 0.042 7.513 7.772 array2fm 1 7.0 0.000 0.000 6.762 7.255 pw_scatter_s 2720 13.7 4.428 4.580 4.428 4.580 pw_gather_s 2722 13.2 3.863 4.178 3.863 4.178 array2fm_buffer_send 1 8.0 2.943 3.083 2.943 3.083 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=113.253515, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2730.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8f18a4e0cc0e0e7c2e20248e70b07b7bca2af436_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 0.054 0.074 405.441 405.441 farming_run 1 2.0 404.404 404.450 405.366 405.396 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.224110E+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 703 408373. MP_Allreduce 1821 23730. MP_Sync 38 MP_Alltoall 77 2368424. 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 0.016 0.052 209.736 209.737 qs_energies 1 2.0 0.002 0.015 209.485 209.492 scf_env_do_scf 1 3.0 0.000 0.000 106.394 106.394 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.473 105.482 rebuild_ks_matrix 4 6.0 0.000 0.000 105.472 105.481 qs_ks_build_kohn_sham_matrix 4 7.0 0.055 0.064 105.472 105.481 hfx_ks_matrix 4 8.0 0.001 0.001 105.062 105.067 integrate_four_center 4 9.0 0.144 0.462 105.061 105.066 mp2_main 1 3.0 0.000 0.004 102.794 102.802 mp2_gpw_main 1 4.0 0.043 0.104 101.875 101.884 integrate_four_center_main 4 10.0 0.092 0.528 96.907 99.633 integrate_four_center_bin 265 11.0 96.815 99.614 96.815 99.614 init_scf_loop 1 4.0 0.000 0.000 91.972 91.972 mp2_ri_gpw_compute_in 1 5.0 0.067 0.133 74.984 76.126 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.514 55.647 mp2_eri_3c_integrate_gpw 91 7.0 0.145 0.158 42.388 47.218 integrate_v_rspace 95 8.0 0.398 0.566 28.768 33.421 pw_transfer 2240 10.6 0.144 0.165 29.930 30.277 fft_wrap_pw1pw2 1868 11.4 0.018 0.020 28.943 29.312 grid_integrate_task_list 95 9.0 24.042 28.916 24.042 28.916 mp2_ri_gpw_compute_en 1 5.0 0.067 0.112 26.728 28.437 ao_to_mo_and_store_B_mult_1 91 7.0 10.444 27.979 10.444 27.979 fft_wrap_pw1pw2_100 730 12.4 1.270 1.412 26.639 26.986 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.830 1.900 24.874 24.886 get_2c_integrals 1 6.0 0.009 0.040 20.311 20.362 compute_2c_integrals 1 7.0 0.006 0.017 19.282 19.300 compute_2c_integrals_loop_lm 1 8.0 0.003 0.010 18.845 19.144 mp2_eri_2c_integrate_gpw 1 9.0 1.739 1.831 18.842 19.143 fft3d_s 1823 13.4 18.428 18.824 18.441 18.837 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.421 14.421 calculate_wavefunction 91 8.0 2.031 2.077 9.758 9.972 mp2_ri_gpw_compute_en_expansio 172 7.0 0.557 0.589 8.759 9.428 potential_pw2rs 186 10.0 0.033 0.035 8.668 9.171 local_gemm 172 8.0 8.202 8.863 8.202 8.863 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.238 8.640 mp2_ri_gpw_compute_en_comm 22 7.0 0.499 0.528 7.895 8.264 collocate_single_gaussian 91 10.0 0.018 0.039 7.876 8.146 calc_potential_gpw 182 9.5 0.002 0.002 7.895 8.103 mp2_ri_gpw_compute_en_ener 172 7.0 6.345 6.436 6.345 6.436 mp_sendrecv_dm3 2068 8.0 5.928 6.277 5.928 6.277 mp_sync 38 10.4 2.861 5.420 2.861 5.420 pw_gather_s 912 13.2 4.921 5.356 4.921 5.356 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.873696, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1511.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8f18a4e0cc0e0e7c2e20248e70b07b7bca2af436_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 452.673536E+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 271. MP_Sync 580 MP_Alltoall 2083 578641. 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 0.036 0.054 59.471 59.479 qs_mol_dyn_low 1 2.0 0.004 0.005 59.155 59.164 qs_forces 11 3.9 0.002 0.002 59.077 59.078 qs_energies 11 4.9 0.003 0.009 57.238 57.260 scf_env_do_scf 11 5.9 0.001 0.001 50.086 50.086 scf_env_do_scf_inner_loop 108 6.5 0.009 0.032 47.384 47.386 dbcsr_multiply_generic 2286 12.5 0.099 0.105 36.312 36.927 qs_scf_new_mos 108 7.5 0.000 0.001 35.330 35.685 qs_scf_loop_do_ot 108 8.5 0.001 0.001 35.330 35.685 ot_scf_mini 108 9.5 0.002 0.005 33.567 33.755 velocity_verlet 10 3.0 0.001 0.002 29.506 29.507 multiply_cannon 2286 13.5 0.182 0.189 27.275 29.221 multiply_cannon_loop 2286 14.5 1.504 1.575 26.227 28.148 ot_mini 108 10.5 0.001 0.001 21.470 21.731 qs_ot_get_derivative 108 11.5 0.001 0.002 18.143 18.362 mp_waitall_1 245248 16.5 10.068 16.159 10.068 16.159 multiply_cannon_metrocomm3 54864 15.5 0.069 0.076 6.311 14.331 multiply_cannon_multrec 54864 15.5 4.119 6.317 7.636 11.064 rebuild_ks_matrix 119 8.3 0.000 0.000 9.312 9.490 qs_ks_build_kohn_sham_matrix 119 9.3 0.011 0.016 9.311 9.489 qs_ks_update_qs_env 119 7.6 0.001 0.001 8.123 8.290 mp_sum_l 7207 12.9 5.826 7.793 5.826 7.793 qs_ot_get_p 119 10.4 0.001 0.005 7.223 7.482 multiply_cannon_sync_h2d 54864 15.5 5.753 6.868 5.753 6.868 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 6.269 6.752 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 6.164 6.283 init_scf_run 11 5.9 0.000 0.001 5.606 5.606 scf_env_initial_rho_setup 11 6.9 0.015 0.023 5.605 5.606 sum_up_and_integrate 119 10.3 0.012 0.015 5.565 5.577 integrate_v_rspace 119 11.3 0.002 0.002 5.553 5.566 qs_rho_update_rho_low 119 7.7 0.001 0.001 5.138 5.249 calculate_rho_elec 119 8.7 0.011 0.017 5.138 5.249 dbcsr_mm_accdrv_process 76910 16.1 1.209 1.865 3.440 4.842 multiply_cannon_metrocomm1 54864 15.5 0.052 0.058 2.478 4.249 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.962 4.018 rs_pw_transfer 974 11.9 0.011 0.012 3.656 3.858 calculate_dm_sparse 119 9.5 0.000 0.000 3.107 3.316 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.048 3.258 apply_single 119 13.6 0.000 0.000 3.048 3.258 cp_dbcsr_syevd 50 12.0 0.002 0.003 3.135 3.135 ot_diis_step 108 11.5 0.006 0.006 3.025 3.029 density_rs2pw 119 9.7 0.004 0.004 2.916 3.025 calculate_first_density_matrix 1 7.0 0.007 0.042 3.004 3.022 make_m2s 4572 13.5 0.054 0.057 2.907 3.005 make_images 4572 14.5 0.133 0.139 2.824 2.922 jit_kernel_multiply 13 15.8 2.168 2.907 2.168 2.907 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.800 2.801 cp_fm_redistribute_end 50 14.0 2.554 2.773 2.561 2.776 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.710 2.717 init_scf_loop 11 6.9 0.003 0.018 2.678 2.683 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.591 2.660 cp_fm_diag_elpa_base 50 14.0 0.211 2.550 0.212 2.561 potential_pw2rs 119 12.3 0.004 0.004 2.538 2.553 pw_transfer 1439 11.6 0.053 0.059 2.451 2.540 wfi_extrapolate 11 7.9 0.001 0.001 2.521 2.522 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 2.374 2.468 mp_alltoall_d11v 2130 13.8 2.189 2.396 2.189 2.396 mp_sum_d 4129 12.0 1.715 2.377 1.715 2.377 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.250 2.315 acc_transpose_blocks 54864 15.5 0.233 0.252 1.848 2.291 fft3d_ps 1201 14.6 0.377 0.484 2.145 2.233 grid_integrate_task_list 119 12.3 2.024 2.111 2.024 2.111 fft_wrap_pw1pw2_140 487 13.2 0.082 0.095 1.829 1.923 make_images_sizes 4572 15.5 0.004 0.005 1.423 1.760 mp_alltoall_i44 4572 16.5 1.419 1.756 1.419 1.756 mp_waitany 12084 13.8 1.559 1.736 1.559 1.736 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.512 1.545 dbcsr_dot_sd 1205 11.9 0.051 0.062 0.988 1.398 grid_collocate_task_list 119 9.7 1.292 1.364 1.292 1.364 prepare_preconditioner 11 7.9 0.000 0.002 1.304 1.333 make_preconditioner 11 8.9 0.000 0.004 1.304 1.332 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.257 1.281 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.202 1.239 mp_alltoall_z22v 1201 16.6 1.147 1.216 1.147 1.216 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=59.479000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=431.090909, yerr=1.239835 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8f18a4e0cc0e0e7c2e20248e70b07b7bca2af436_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 489.545728E+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 104 MP_Alltoall 2060 1621224. 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 0.033 0.062 45.293 45.294 qs_mol_dyn_low 1 2.0 0.003 0.003 44.939 44.948 qs_forces 11 3.9 0.002 0.006 44.867 44.868 qs_energies 11 4.9 0.016 0.061 42.946 42.954 scf_env_do_scf 11 5.9 0.001 0.001 37.268 37.281 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 34.046 34.046 dbcsr_multiply_generic 2286 12.5 0.101 0.103 23.924 24.242 qs_scf_new_mos 108 7.5 0.001 0.001 23.190 23.471 qs_scf_loop_do_ot 108 8.5 0.001 0.001 23.190 23.470 ot_scf_mini 108 9.5 0.002 0.003 22.212 22.438 velocity_verlet 10 3.0 0.001 0.001 21.572 21.574 multiply_cannon 2286 13.5 0.209 0.221 17.334 19.471 multiply_cannon_loop 2286 14.5 0.908 0.982 15.974 17.697 ot_mini 108 10.5 0.001 0.001 13.769 14.076 mp_waitall_1 200699 16.5 7.074 12.257 7.074 12.257 qs_ot_get_derivative 108 11.5 0.001 0.001 11.035 11.254 multiply_cannon_metrocomm3 27432 15.5 0.069 0.072 4.732 11.068 rebuild_ks_matrix 119 8.3 0.000 0.000 8.281 8.633 qs_ks_build_kohn_sham_matrix 119 9.3 0.019 0.044 8.280 8.633 multiply_cannon_multrec 27432 15.5 1.948 4.309 5.854 8.558 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.330 7.630 dbcsr_mm_accdrv_process 47894 16.0 3.064 4.960 3.838 5.701 qs_ot_get_p 119 10.4 0.001 0.001 5.291 5.553 mp_sum_l 7207 12.9 3.007 5.187 3.007 5.187 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 4.195 5.064 sum_up_and_integrate 119 10.3 0.025 0.031 4.972 4.978 integrate_v_rspace 119 11.3 0.003 0.008 4.947 4.955 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.710 4.737 calculate_rho_elec 119 8.7 0.021 0.024 4.709 4.736 init_scf_run 11 5.9 0.000 0.001 4.328 4.328 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.327 4.328 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.191 4.260 apply_single 119 13.6 0.000 0.000 3.191 4.260 rs_pw_transfer 974 11.9 0.010 0.011 3.663 4.110 qs_ot_p2m_diag 50 11.0 0.009 0.013 3.463 3.483 density_rs2pw 119 9.7 0.004 0.004 2.955 3.398 make_m2s 4572 13.5 0.052 0.053 3.151 3.393 make_images 4572 14.5 0.200 0.236 3.063 3.304 init_scf_loop 11 6.9 0.004 0.020 3.187 3.200 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.970 2.971 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.737 2.840 multiply_cannon_sync_h2d 27432 15.5 2.135 2.790 2.135 2.790 ot_diis_step 108 11.5 0.011 0.011 2.647 2.648 potential_pw2rs 119 12.3 0.006 0.006 2.561 2.595 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.579 2.580 cp_fm_redistribute_end 50 14.0 2.135 2.548 2.142 2.549 calculate_first_density_matrix 1 7.0 0.001 0.007 2.503 2.505 pw_transfer 1439 11.6 0.065 0.072 2.429 2.473 cp_fm_diag_elpa_base 50 14.0 0.391 2.384 0.404 2.442 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.337 2.384 calculate_dm_sparse 119 9.5 0.000 0.000 2.151 2.252 fft3d_ps 1201 14.6 0.522 0.576 2.037 2.084 prepare_preconditioner 11 7.9 0.000 0.001 2.000 2.029 make_preconditioner 11 8.9 0.000 0.002 2.000 2.029 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.014 2.018 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.869 1.937 grid_integrate_task_list 119 12.3 1.827 1.908 1.827 1.908 mp_alltoall_d11v 2130 13.8 1.621 1.807 1.621 1.807 make_images_data 4572 15.5 0.044 0.051 1.295 1.774 fft_wrap_pw1pw2_140 487 13.2 0.079 0.085 1.715 1.763 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.704 1.756 wfi_extrapolate 11 7.9 0.001 0.001 1.743 1.743 hybrid_alltoall_any 4725 16.4 0.050 0.112 1.160 1.671 jit_kernel_multiply 8 16.2 0.722 1.586 0.722 1.586 mp_allgather_i34 2286 14.5 0.801 1.557 0.801 1.557 acc_transpose_blocks 27432 15.5 0.108 0.115 1.215 1.532 make_images_sizes 4572 15.5 0.005 0.005 1.046 1.392 mp_alltoall_i44 4572 16.5 1.041 1.388 1.041 1.388 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.325 1.374 grid_collocate_task_list 119 9.7 1.232 1.364 1.232 1.364 multiply_cannon_metrocomm1 27432 15.5 0.032 0.036 0.554 1.359 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.304 1.312 mp_sum_d 4129 12.0 0.828 1.301 0.828 1.301 mp_alltoall_z22v 1201 16.6 1.053 1.166 1.053 1.166 mp_waitany 5720 13.7 0.693 1.156 0.693 1.156 rs_pw_transfer_RS2PW_140 130 11.5 0.140 0.162 0.702 1.135 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.026 1.078 rs_pw_transfer_PW2RS_50 119 14.3 0.587 0.605 0.964 1.042 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 0.997 0.998 dbcsr_dot_sd 1205 11.9 0.072 0.088 0.546 0.966 cp_fm_cholesky_invert 11 10.9 0.919 0.923 0.919 0.923 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=45.294000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=465.272727, yerr=1.354515 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8f18a4e0cc0e0e7c2e20248e70b07b7bca2af436_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 523.530240E+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 3672 62660. MP_Allreduce 10225 303. MP_Sync 104 MP_Alltoall 1821 1607811. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_comm_split 50 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 0.016 0.032 33.093 33.094 qs_mol_dyn_low 1 2.0 0.003 0.004 32.837 32.860 qs_forces 11 3.9 0.001 0.002 32.751 32.752 qs_energies 11 4.9 0.003 0.020 31.149 31.152 scf_env_do_scf 11 5.9 0.000 0.001 26.323 26.324 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 23.698 23.699 dbcsr_multiply_generic 2286 12.5 0.095 0.097 16.655 16.748 velocity_verlet 10 3.0 0.001 0.001 15.701 15.703 qs_scf_new_mos 108 7.5 0.001 0.001 15.209 15.223 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.209 15.223 ot_scf_mini 108 9.5 0.002 0.002 14.477 14.495 multiply_cannon 2286 13.5 0.196 0.204 13.177 13.930 multiply_cannon_loop 2286 14.5 0.638 0.668 12.373 13.218 ot_mini 108 10.5 0.001 0.001 8.914 8.928 qs_ot_get_derivative 108 11.5 0.001 0.001 7.396 7.413 multiply_cannon_multrec 18288 15.5 1.929 3.010 6.761 7.184 rebuild_ks_matrix 119 8.3 0.000 0.000 6.556 6.574 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.555 6.573 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.794 5.811 dbcsr_mm_accdrv_process 38222 16.0 3.996 5.456 4.748 5.611 mp_waitall_1 158411 16.6 2.850 4.255 2.850 4.255 sum_up_and_integrate 119 10.3 0.029 0.030 4.105 4.109 integrate_v_rspace 119 11.3 0.002 0.003 4.075 4.083 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.604 3.638 calculate_rho_elec 119 8.7 0.030 0.031 3.603 3.638 init_scf_run 11 5.9 0.000 0.001 3.610 3.610 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.610 3.610 qs_ot_get_p 119 10.4 0.001 0.001 3.381 3.401 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.753 3.392 multiply_cannon_metrocomm3 18288 15.5 0.045 0.047 1.473 2.798 rs_pw_transfer 974 11.9 0.009 0.010 2.442 2.669 init_scf_loop 11 6.9 0.000 0.001 2.605 2.605 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.062 2.384 apply_single 119 13.6 0.000 0.000 2.062 2.384 density_rs2pw 119 9.7 0.004 0.004 2.116 2.360 calculate_first_density_matrix 1 7.0 0.000 0.002 2.355 2.357 qs_ot_p2m_diag 50 11.0 0.012 0.012 2.284 2.293 pw_transfer 1439 11.6 0.066 0.072 2.043 2.052 make_m2s 4572 13.5 0.044 0.045 1.870 2.023 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.995 1.996 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.949 1.962 jit_kernel_multiply 10 16.0 0.700 1.939 0.700 1.939 make_images 4572 14.5 0.190 0.201 1.786 1.937 grid_integrate_task_list 119 12.3 1.793 1.899 1.793 1.899 calculate_dm_sparse 119 9.5 0.000 0.000 1.862 1.875 potential_pw2rs 119 12.3 0.007 0.008 1.862 1.872 prepare_preconditioner 11 7.9 0.000 0.000 1.780 1.783 make_preconditioner 11 8.9 0.000 0.000 1.780 1.783 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.722 1.724 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.722 1.724 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.631 1.715 cp_fm_redistribute_end 50 14.0 1.284 1.700 1.285 1.701 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.658 1.665 cp_fm_diag_elpa_base 50 14.0 0.398 1.616 0.412 1.657 fft3d_ps 1201 14.6 0.524 0.542 1.628 1.642 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.574 1.581 mp_sum_l 7207 12.9 1.220 1.569 1.220 1.569 multiply_cannon_sync_h2d 18288 15.5 1.366 1.556 1.366 1.556 fft_wrap_pw1pw2_140 487 13.2 0.090 0.094 1.532 1.543 ot_diis_step 108 11.5 0.011 0.011 1.494 1.494 grid_collocate_task_list 119 9.7 1.213 1.370 1.213 1.370 acc_transpose_blocks 18288 15.5 0.076 0.077 1.271 1.295 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.228 1.234 wfi_extrapolate 11 7.9 0.001 0.001 1.203 1.203 multiply_cannon_metrocomm1 18288 15.5 0.029 0.030 0.468 1.146 make_images_data 4572 15.5 0.044 0.048 0.836 0.999 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 0.958 0.959 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.908 0.928 hybrid_alltoall_any 4725 16.4 0.055 0.113 0.726 0.911 acc_transpose_blocks_kernels 18288 16.5 0.210 0.219 0.830 0.846 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.833 0.835 mp_alltoall_d11v 2130 13.8 0.727 0.824 0.727 0.824 mp_alltoall_z22v 1201 16.6 0.699 0.799 0.699 0.799 mp_waitany 9880 13.7 0.525 0.774 0.525 0.774 rs_pw_transfer_RS2PW_140 130 11.5 0.121 0.137 0.520 0.752 cp_fm_cholesky_invert 11 10.9 0.745 0.749 0.745 0.749 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.650 0.743 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=33.094000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=497.363636, yerr=2.603240 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8f18a4e0cc0e0e7c2e20248e70b07b7bca2af436_performance_tests/06/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 114.044384E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3805952 0.0% 0.0% 100.0% average stack size 0.0 0.0 38.6 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 559.337472E+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 3672 62659. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_comm_split 50 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 0.028 0.056 36.303 36.304 qs_mol_dyn_low 1 2.0 0.003 0.005 35.915 35.931 qs_forces 11 3.9 0.002 0.010 35.831 35.832 qs_energies 11 4.9 0.002 0.011 34.093 34.100 scf_env_do_scf 11 5.9 0.001 0.002 28.610 28.611 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 25.059 25.059 dbcsr_multiply_generic 2286 12.5 0.100 0.102 18.202 18.309 velocity_verlet 10 3.0 0.001 0.001 18.139 18.141 qs_scf_new_mos 108 7.5 0.001 0.001 16.340 16.397 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.340 16.397 ot_scf_mini 108 9.5 0.002 0.003 15.397 15.447 multiply_cannon 2286 13.5 0.225 0.271 14.530 14.910 multiply_cannon_loop 2286 14.5 0.935 0.966 13.561 13.908 ot_mini 108 10.5 0.001 0.001 9.347 9.413 multiply_cannon_multrec 27432 15.5 2.339 3.023 8.586 8.961 qs_ot_get_derivative 108 11.5 0.001 0.001 7.526 7.576 dbcsr_mm_accdrv_process 47916 15.9 5.252 6.585 6.154 7.348 rebuild_ks_matrix 119 8.3 0.000 0.001 6.730 6.775 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 6.730 6.774 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.977 6.017 init_scf_run 11 5.9 0.000 0.001 4.012 4.013 scf_env_initial_rho_setup 11 6.9 0.025 0.067 4.012 4.012 sum_up_and_integrate 119 10.3 0.035 0.039 3.917 3.926 integrate_v_rspace 119 11.3 0.002 0.004 3.882 3.891 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.697 3.730 calculate_rho_elec 119 8.7 0.040 0.046 3.696 3.730 init_scf_loop 11 6.9 0.000 0.002 3.529 3.529 qs_ot_get_p 119 10.4 0.001 0.001 3.441 3.516 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.762 3.178 prepare_preconditioner 11 7.9 0.000 0.000 2.643 2.651 make_preconditioner 11 8.9 0.000 0.000 2.643 2.651 mp_waitall_1 137007 16.6 1.988 2.614 1.988 2.614 calculate_first_density_matrix 1 7.0 0.002 0.013 2.568 2.594 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.254 2.580 make_m2s 4572 13.5 0.054 0.055 2.261 2.387 rs_pw_transfer 974 11.9 0.009 0.009 2.165 2.354 density_rs2pw 119 9.7 0.004 0.004 2.099 2.290 make_images 4572 14.5 0.272 0.332 2.155 2.279 pw_transfer 1439 11.6 0.066 0.072 2.179 2.221 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.061 2.193 apply_single 119 13.6 0.000 0.000 2.061 2.193 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.133 2.143 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.086 2.133 calculate_dm_sparse 119 9.5 0.000 0.000 2.040 2.098 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.945 1.972 grid_integrate_task_list 119 12.3 1.818 1.900 1.818 1.900 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.883 1.885 jit_kernel_multiply 10 15.8 0.841 1.828 0.841 1.828 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.813 1.814 fft3d_ps 1201 14.6 0.561 0.615 1.760 1.799 ot_diis_step 108 11.5 0.012 0.012 1.781 1.781 fft_wrap_pw1pw2_140 487 13.2 0.088 0.095 1.710 1.762 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.681 1.693 potential_pw2rs 119 12.3 0.008 0.009 1.665 1.670 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.547 1.549 cp_fm_redistribute_end 50 14.0 1.022 1.518 1.023 1.519 multiply_cannon_metrocomm3 27432 15.5 0.038 0.040 0.836 1.508 acc_transpose_blocks 27432 15.5 0.112 0.114 1.477 1.495 cp_fm_diag_elpa_base 50 14.0 0.472 1.447 0.492 1.487 wfi_extrapolate 11 7.9 0.001 0.001 1.364 1.364 grid_collocate_task_list 119 9.7 1.221 1.336 1.221 1.336 mp_sum_l 7207 12.9 0.995 1.257 0.995 1.257 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.218 1.229 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.201 1.203 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.112 1.127 cp_fm_upper_to_full 72 13.5 0.811 1.122 0.811 1.122 dbcsr_complete_redistribute 329 12.2 0.132 0.167 0.811 1.089 multiply_cannon_sync_h2d 27432 15.5 0.988 1.068 0.988 1.068 make_images_data 4572 15.5 0.045 0.049 0.897 1.050 mp_alltoall_d11v 2130 13.8 0.768 0.963 0.768 0.963 hybrid_alltoall_any 4725 16.4 0.061 0.151 0.783 0.959 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.791 0.870 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.594 0.868 acc_transpose_blocks_kernels 27432 16.5 0.269 0.277 0.852 0.864 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.844 0.848 mp_sum_d 4127 12.0 0.614 0.835 0.614 0.835 mp_alltoall_z22v 1201 16.6 0.783 0.823 0.783 0.823 cp_fm_cholesky_invert 11 10.9 0.805 0.809 0.805 0.809 mp_alltoall_i22 627 13.8 0.430 0.727 0.430 0.727 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=36.304000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=529.818182, yerr=4.238743 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8f18a4e0cc0e0e7c2e20248e70b07b7bca2af436_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 599.769088E+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 3672 62658. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_comm_split 50 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 0.057 0.072 29.216 29.218 qs_mol_dyn_low 1 2.0 0.003 0.004 28.897 28.906 qs_forces 11 3.9 0.002 0.002 28.820 28.821 qs_energies 11 4.9 0.001 0.001 27.101 27.104 scf_env_do_scf 11 5.9 0.000 0.001 22.182 22.183 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 19.557 19.557 velocity_verlet 10 3.0 0.001 0.001 14.783 14.787 dbcsr_multiply_generic 2286 12.5 0.093 0.096 12.412 12.551 qs_scf_new_mos 108 7.5 0.001 0.001 11.311 11.340 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.311 11.339 ot_scf_mini 108 9.5 0.002 0.002 10.627 10.656 multiply_cannon 2286 13.5 0.233 0.239 9.780 10.307 multiply_cannon_loop 2286 14.5 0.331 0.342 8.857 9.117 rebuild_ks_matrix 119 8.3 0.000 0.000 6.145 6.169 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 6.145 6.168 multiply_cannon_multrec 9144 15.5 1.605 1.830 5.770 6.029 ot_mini 108 10.5 0.001 0.001 5.931 5.966 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.491 5.514 qs_ot_get_derivative 108 11.5 0.001 0.001 4.617 4.644 dbcsr_mm_accdrv_process 12550 15.8 3.150 3.767 4.064 4.155 sum_up_and_integrate 119 10.3 0.038 0.040 3.780 3.786 integrate_v_rspace 119 11.3 0.003 0.003 3.743 3.748 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.659 3.665 calculate_rho_elec 119 8.7 0.059 0.061 3.658 3.665 init_scf_run 11 5.9 0.000 0.001 3.472 3.472 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.472 3.472 qs_ot_get_p 119 10.4 0.001 0.001 2.908 2.945 init_scf_loop 11 6.9 0.000 0.000 2.603 2.603 calculate_first_density_matrix 1 7.0 0.000 0.000 2.281 2.283 pw_transfer 1439 11.6 0.066 0.070 2.235 2.245 mp_waitall_1 115863 16.7 1.698 2.224 1.698 2.224 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.141 2.154 density_rs2pw 119 9.7 0.004 0.004 1.995 2.136 make_m2s 4572 13.5 0.034 0.035 1.822 1.973 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.955 1.958 grid_integrate_task_list 119 12.3 1.851 1.929 1.851 1.929 rs_pw_transfer 974 11.9 0.008 0.008 1.786 1.929 make_images 4572 14.5 0.269 0.303 1.732 1.882 prepare_preconditioner 11 7.9 0.000 0.000 1.831 1.835 make_preconditioner 11 8.9 0.000 0.000 1.831 1.835 fft3d_ps 1201 14.6 0.566 0.577 1.799 1.807 fft_wrap_pw1pw2_140 487 13.2 0.087 0.090 1.764 1.777 calculate_dm_sparse 119 9.5 0.000 0.000 1.750 1.769 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.755 1.755 jit_kernel_multiply 10 15.7 0.876 1.754 0.876 1.754 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.712 1.741 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.662 1.663 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.541 1.553 potential_pw2rs 119 12.3 0.010 0.011 1.494 1.500 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.488 1.498 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.471 1.472 cp_fm_redistribute_end 50 14.0 0.735 1.448 0.737 1.449 cp_fm_diag_elpa_base 50 14.0 0.666 1.371 0.711 1.430 grid_collocate_task_list 119 9.7 1.274 1.394 1.274 1.394 ot_diis_step 108 11.5 0.013 0.013 1.300 1.300 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.233 1.233 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.176 1.201 apply_single 119 13.6 0.000 0.000 1.176 1.201 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.172 1.186 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.163 1.168 wfi_extrapolate 11 7.9 0.001 0.001 1.139 1.139 make_images_data 4572 15.5 0.039 0.042 0.847 1.045 hybrid_alltoall_any 4725 16.4 0.062 0.175 0.805 1.043 acc_transpose_blocks 9144 15.5 0.038 0.039 0.953 0.965 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.869 0.924 cp_fm_cholesky_invert 11 10.9 0.913 0.916 0.913 0.916 mp_alltoall_d11v 2130 13.8 0.778 0.838 0.778 0.838 multiply_cannon_sync_h2d 9144 15.5 0.717 0.785 0.717 0.785 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.777 0.779 mp_alltoall_z22v 1201 16.6 0.680 0.740 0.680 0.740 qs_env_update_s_mstruct 11 6.9 0.001 0.001 0.685 0.739 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.728 0.736 acc_transpose_blocks_kernels 9144 16.5 0.117 0.120 0.713 0.722 multiply_cannon_metrocomm3 9144 15.5 0.019 0.020 0.444 0.716 mp_allgather_i34 2286 14.5 0.255 0.660 0.255 0.660 yz_to_x 606 15.1 0.265 0.276 0.622 0.643 x_to_yz 595 16.2 0.276 0.289 0.599 0.607 jit_kernel_transpose 5 15.6 0.595 0.605 0.595 0.605 dbcsr_complete_redistribute 329 12.2 0.167 0.180 0.562 0.593 mp_waitany 5200 13.7 0.463 0.589 0.463 0.589 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=29.218000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=570.272727, yerr=3.816017 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8f18a4e0cc0e0e7c2e20248e70b07b7bca2af436_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 742.367232E+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 0.019 0.035 42.527 42.529 qs_mol_dyn_low 1 2.0 0.003 0.004 42.267 42.274 qs_forces 11 3.9 0.002 0.002 42.199 42.200 qs_energies 11 4.9 0.002 0.003 40.173 40.177 scf_env_do_scf 11 5.9 0.001 0.001 34.296 34.296 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 26.250 26.251 velocity_verlet 10 3.0 0.001 0.001 23.893 23.900 dbcsr_multiply_generic 2286 12.5 0.101 0.103 17.573 17.767 qs_scf_new_mos 108 7.5 0.001 0.001 16.170 16.278 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.169 16.277 ot_scf_mini 108 9.5 0.002 0.002 15.076 15.186 multiply_cannon 2286 13.5 0.305 0.315 13.648 14.583 multiply_cannon_loop 2286 14.5 0.344 0.350 12.377 13.346 ot_mini 108 10.5 0.001 0.001 8.901 9.027 multiply_cannon_multrec 9144 15.5 3.338 4.567 8.543 8.633 init_scf_loop 11 6.9 0.000 0.000 8.015 8.016 rebuild_ks_matrix 119 8.3 0.000 0.000 7.523 7.678 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 7.522 7.677 prepare_preconditioner 11 7.9 0.000 0.000 6.998 7.013 make_preconditioner 11 8.9 0.000 0.000 6.998 7.013 qs_ot_get_derivative 108 11.5 0.001 0.001 6.857 6.969 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.808 6.949 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.560 6.888 dbcsr_mm_accdrv_process 12550 15.8 4.260 5.623 5.082 6.356 cp_fm_upper_to_full 72 14.2 3.163 4.543 3.163 4.543 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.385 4.392 calculate_rho_elec 119 8.7 0.118 0.121 4.385 4.391 sum_up_and_integrate 119 10.3 0.065 0.066 4.162 4.168 integrate_v_rspace 119 11.3 0.003 0.003 4.097 4.104 init_scf_run 11 5.9 0.000 0.001 3.767 3.767 scf_env_initial_rho_setup 11 6.9 0.001 0.002 3.767 3.767 qs_ot_get_p 119 10.4 0.001 0.001 3.431 3.579 mp_waitall_1 94719 16.7 2.473 3.454 2.473 3.454 pw_transfer 1439 11.6 0.069 0.070 2.933 2.940 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.538 2.929 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.834 2.840 dbcsr_complete_redistribute 329 12.2 0.287 0.293 1.997 2.794 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.674 2.479 make_m2s 4572 13.5 0.038 0.038 2.306 2.477 fft3d_ps 1201 14.6 0.605 0.615 2.457 2.463 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.161 2.457 apply_single 119 13.6 0.000 0.000 2.161 2.457 fft_wrap_pw1pw2_140 487 13.2 0.095 0.098 2.365 2.373 make_images 4572 14.5 0.354 0.382 2.185 2.356 density_rs2pw 119 9.7 0.004 0.004 2.301 2.315 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 1.385 2.290 calculate_first_density_matrix 1 7.0 0.000 0.000 2.287 2.288 mp_alltoall_i22 627 13.8 1.474 2.286 1.474 2.286 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.433 2.230 calculate_dm_sparse 119 9.5 0.000 0.000 2.179 2.196 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.118 2.119 grid_integrate_task_list 119 12.3 2.060 2.076 2.060 2.076 ot_diis_step 108 11.5 0.014 0.014 2.016 2.016 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.937 1.937 mp_sum_l 7207 12.9 1.076 1.859 1.076 1.859 qs_energies_init_hamiltonians 11 5.9 0.002 0.003 1.838 1.838 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.829 1.829 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.710 1.766 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.624 1.644 cp_fm_cholesky_invert 11 10.9 1.584 1.588 1.584 1.588 rs_pw_transfer 974 11.9 0.009 0.009 1.565 1.583 potential_pw2rs 119 12.3 0.014 0.015 1.551 1.554 grid_collocate_task_list 119 9.7 1.491 1.502 1.491 1.502 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.502 1.502 cp_fm_diag_elpa_base 50 14.0 1.358 1.409 1.500 1.500 wfi_extrapolate 11 7.9 0.001 0.001 1.406 1.406 hybrid_alltoall_any 4725 16.4 0.087 0.147 1.117 1.362 make_images_data 4572 15.5 0.043 0.046 1.071 1.289 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.234 1.242 qs_env_update_s_mstruct 11 6.9 0.002 0.004 1.172 1.187 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.154 1.176 mp_alltoall_d11v 2130 13.8 1.093 1.114 1.093 1.114 multiply_cannon_sync_h2d 9144 15.5 1.045 1.047 1.045 1.047 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 0.986 1.036 jit_kernel_multiply 6 15.5 0.795 1.032 0.795 1.032 yz_to_x 606 15.1 0.460 0.470 0.994 1.008 qs_create_task_list 11 7.9 0.030 0.045 0.988 1.002 generate_qs_task_list 11 8.9 0.370 0.389 0.958 1.001 mp_alltoall_z22v 1201 16.6 0.899 0.946 0.899 0.946 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.927 0.942 acc_transpose_blocks 9144 15.5 0.038 0.038 0.911 0.917 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=42.529000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=697.272727, yerr=12.424982 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8f18a4e0cc0e0e7c2e20248e70b07b7bca2af436_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 502.177792E+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 3521 65372. MP_Allreduce 9840 486. MP_Sync 100 MP_Alltoall 1938 1386274. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_comm_split 48 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 0.032 0.056 92.232 92.233 qs_mol_dyn_low 1 2.0 0.004 0.007 91.913 91.923 qs_forces 11 3.9 0.002 0.004 89.874 89.879 qs_energies 11 4.9 0.001 0.002 86.964 86.977 scf_env_do_scf 11 5.9 0.000 0.001 77.638 77.640 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 71.288 71.289 dbcsr_multiply_generic 2055 12.4 0.109 0.112 55.400 55.829 qs_scf_new_mos 99 7.5 0.000 0.001 52.162 52.345 qs_scf_loop_do_ot 99 8.5 0.001 0.001 52.162 52.345 ot_scf_mini 99 9.5 0.002 0.002 49.545 49.687 velocity_verlet 10 3.0 0.001 0.001 47.192 47.213 multiply_cannon 2055 13.4 0.175 0.179 43.824 45.155 multiply_cannon_loop 2055 14.4 1.553 1.597 42.634 43.951 ot_mini 99 10.5 0.001 0.001 29.743 29.906 qs_ot_get_derivative 99 11.5 0.001 0.001 22.729 22.923 multiply_cannon_multrec 49320 15.4 11.883 12.803 16.979 17.697 rebuild_ks_matrix 110 8.3 0.000 0.001 15.540 15.739 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.011 15.540 15.738 mp_waitall_1 220248 16.4 13.530 15.017 13.530 15.017 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.724 13.903 qs_ot_get_p 110 10.4 0.001 0.001 10.866 10.974 multiply_cannon_sync_h2d 49320 15.4 9.717 10.409 9.717 10.409 multiply_cannon_metrocomm3 49320 15.4 0.078 0.083 7.127 8.307 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.642 8.190 sum_up_and_integrate 110 10.3 0.036 0.043 7.844 7.859 integrate_v_rspace 110 11.3 0.003 0.003 7.807 7.830 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.563 7.709 calculate_rho_elec 110 8.6 0.021 0.025 7.562 7.708 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 7.588 7.678 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.214 7.654 apply_single 110 13.6 0.000 0.001 7.213 7.653 mp_sum_l 6514 12.8 6.145 7.137 6.145 7.137 init_scf_run 11 5.9 0.000 0.001 7.127 7.127 scf_env_initial_rho_setup 11 6.9 0.000 0.001 7.126 7.127 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.806 6.837 ot_diis_step 99 11.5 0.006 0.006 6.704 6.706 init_scf_loop 11 6.9 0.000 0.000 6.319 6.320 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.699 5.700 dbcsr_mm_accdrv_process 87628 16.1 2.109 2.217 4.977 5.271 rs_pw_transfer 902 11.9 0.011 0.013 4.471 5.113 make_m2s 4110 13.4 0.061 0.065 4.966 5.102 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.076 5.080 cp_fm_redistribute_end 48 14.0 4.443 5.048 4.449 5.052 make_images 4110 14.4 0.177 0.191 4.870 5.010 multiply_cannon_metrocomm1 49320 15.4 0.060 0.064 3.959 4.994 cp_fm_diag_elpa_base 48 14.0 0.593 4.755 0.596 4.788 density_rs2pw 110 9.6 0.004 0.005 4.074 4.724 wfi_extrapolate 11 7.9 0.001 0.001 4.173 4.173 calculate_dm_sparse 110 9.5 0.000 0.001 4.021 4.143 prepare_preconditioner 11 7.9 0.000 0.000 4.025 4.043 make_preconditioner 11 8.9 0.000 0.000 4.025 4.043 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.796 3.835 pw_transfer 1331 11.6 0.055 0.068 3.724 3.789 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.636 3.703 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.661 3.665 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.603 3.663 grid_integrate_task_list 110 12.3 3.262 3.423 3.262 3.423 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.292 3.355 fft3d_ps 1111 14.6 0.784 0.889 3.164 3.216 potential_pw2rs 110 12.3 0.006 0.007 3.167 3.191 mp_alltoall_d11v 2046 13.8 2.658 3.067 2.658 3.067 fft_wrap_pw1pw2_140 451 13.1 0.170 0.191 2.970 3.045 calculate_first_density_matrix 1 7.0 0.000 0.000 2.821 2.844 mp_waitany 14300 13.8 2.146 2.844 2.146 2.844 jit_kernel_multiply 13 15.9 2.584 2.618 2.584 2.618 make_images_data 4110 15.4 0.042 0.045 2.114 2.366 grid_collocate_task_list 110 9.6 2.084 2.309 2.084 2.309 acc_transpose_blocks 49320 15.4 0.226 0.242 2.226 2.302 cp_fm_cholesky_invert 11 10.9 2.253 2.257 2.253 2.257 mp_sum_d 3883 11.9 1.731 2.238 1.731 2.238 hybrid_alltoall_any 4261 16.3 0.081 0.475 1.851 2.155 mp_sum_dm 438 4.9 2.081 2.116 2.081 2.116 md_write_output 11 3.9 0.021 1.994 0.023 2.014 md_output 10 3.0 0.000 0.000 0.024 2.011 update_particle_set 20 4.0 0.000 0.000 1.989 2.010 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.960 1.976 mp_alltoall_z22v 1111 16.6 1.655 1.921 1.655 1.921 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=92.233000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=477.000000, yerr=2.044949 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8f18a4e0cc0e0e7c2e20248e70b07b7bca2af436_performance_tests/10/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 390.715586E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 5019072 0.0% 0.0% 100.0% average stack size 0.0 0.0 196.1 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 593.420288E+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 3521 65587. MP_Allreduce 9839 562. MP_Sync 100 MP_Alltoall 1717 2463018. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 MP_comm_split 48 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 0.055 0.072 76.592 76.594 qs_mol_dyn_low 1 2.0 0.004 0.007 74.886 76.170 qs_forces 11 3.9 0.002 0.006 74.726 74.729 qs_energies 11 4.9 0.003 0.018 71.219 71.226 scf_env_do_scf 11 5.9 0.001 0.001 61.986 61.990 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 53.327 53.341 dbcsr_multiply_generic 2055 12.4 0.114 0.119 40.034 40.270 velocity_verlet 10 3.0 0.001 0.001 39.135 39.138 qs_scf_new_mos 99 7.5 0.001 0.001 35.285 35.447 qs_scf_loop_do_ot 99 8.5 0.001 0.002 35.284 35.447 ot_scf_mini 99 9.5 0.003 0.007 33.524 33.673 multiply_cannon 2055 13.4 0.223 0.243 31.653 33.106 multiply_cannon_loop 2055 14.4 0.936 0.953 30.147 31.069 ot_mini 99 10.5 0.001 0.001 19.611 19.773 multiply_cannon_multrec 24660 15.4 7.550 9.423 13.790 15.679 rebuild_ks_matrix 110 8.3 0.000 0.001 15.097 15.204 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.016 15.097 15.204 qs_ot_get_derivative 99 11.5 0.001 0.001 13.614 13.763 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.305 13.400 mp_waitall_1 176588 16.5 8.794 11.237 8.794 11.237 init_scf_loop 11 6.9 0.001 0.007 8.625 8.705 multiply_cannon_metrocomm3 24660 15.4 0.071 0.072 5.355 8.391 multiply_cannon_sync_h2d 24660 15.4 6.959 8.288 6.959 8.288 sum_up_and_integrate 110 10.3 0.051 0.058 7.488 7.505 integrate_v_rspace 110 11.3 0.002 0.004 7.436 7.456 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.610 7.342 apply_single 110 13.6 0.000 0.001 6.609 7.342 qs_ot_get_p 110 10.4 0.001 0.001 6.982 7.139 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.970 6.980 calculate_rho_elec 110 8.6 0.039 0.047 6.970 6.979 init_scf_run 11 5.9 0.000 0.001 6.625 6.626 scf_env_initial_rho_setup 11 6.9 0.004 0.010 6.625 6.625 dbcsr_mm_accdrv_process 52282 16.1 4.803 5.870 6.082 6.387 prepare_preconditioner 11 7.9 0.000 0.001 6.241 6.262 make_preconditioner 11 8.9 0.000 0.001 6.240 6.262 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.758 5.944 ot_diis_step 99 11.5 0.010 0.010 5.939 5.941 make_m2s 4110 13.4 0.056 0.059 5.357 5.902 make_images 4110 14.4 0.399 0.441 5.248 5.790 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 5.009 5.771 qs_ot_p2m_diag 48 11.0 0.028 0.044 4.641 4.663 pw_transfer 1331 11.6 0.066 0.075 4.489 4.644 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 4.382 4.540 density_rs2pw 110 9.6 0.004 0.004 3.932 4.489 rs_pw_transfer 902 11.9 0.012 0.013 3.588 4.188 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.159 4.159 fft3d_ps 1111 14.6 1.104 1.319 3.732 3.877 wfi_extrapolate 11 7.9 0.001 0.001 3.820 3.821 fft_wrap_pw1pw2_140 451 13.1 0.203 0.219 3.628 3.789 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.550 3.631 cp_dbcsr_sm_fm_multiply 37 9.5 0.003 0.012 3.541 3.544 make_images_data 4110 15.4 0.046 0.050 2.841 3.406 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.372 3.379 grid_integrate_task_list 110 12.3 3.149 3.330 3.149 3.330 cp_fm_redistribute_end 48 14.0 2.512 3.311 2.514 3.312 hybrid_alltoall_any 4261 16.3 0.101 0.442 2.455 3.260 mp_sum_l 6514 12.8 2.404 3.191 2.404 3.191 cp_fm_diag_elpa_base 48 14.0 0.756 3.078 0.787 3.163 calculate_dm_sparse 110 9.5 0.001 0.001 3.089 3.121 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.045 3.100 potential_pw2rs 110 12.3 0.008 0.009 2.990 3.022 cp_fm_cholesky_invert 11 10.9 2.904 2.923 2.904 2.923 mp_alltoall_d11v 2046 13.8 2.461 2.796 2.461 2.796 calculate_first_density_matrix 1 7.0 0.004 0.030 2.700 2.705 grid_collocate_task_list 110 9.6 2.048 2.466 2.048 2.466 qs_ot_get_orbitals 99 10.5 0.001 0.001 2.117 2.143 mp_waitany 10164 13.8 1.487 2.059 1.487 2.059 mp_allgather_i34 2055 14.4 0.836 2.054 0.836 2.054 mp_alltoall_z22v 1111 16.6 1.935 2.052 1.935 2.052 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 2.023 2.026 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.963 1.976 jit_kernel_multiply 10 16.3 0.930 1.940 0.930 1.940 dbcsr_complete_redistribute 325 12.2 0.237 0.296 1.482 1.779 rs_pw_transfer_RS2PW_140 121 11.5 0.208 0.218 1.155 1.761 multiply_cannon_metrocomm4 22605 15.4 0.072 0.076 0.790 1.612 acc_transpose_blocks 24660 15.4 0.111 0.114 1.568 1.602 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.585 1.596 cp_fm_cholesky_decompose 22 10.9 1.532 1.539 1.532 1.539 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=76.594000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=558.636364, yerr=8.552222 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8f18a4e0cc0e0e7c2e20248e70b07b7bca2af436_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 660.496384E+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 3521 65578. MP_Allreduce 9838 559. MP_Sync 100 MP_Alltoall 1496 4511006. MP_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 MP_comm_split 48 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 0.047 0.064 65.385 65.388 qs_mol_dyn_low 1 2.0 0.007 0.030 64.991 65.002 qs_forces 11 3.9 0.004 0.019 64.859 64.864 qs_energies 11 4.9 0.003 0.015 61.466 61.475 scf_env_do_scf 11 5.9 0.001 0.002 52.841 52.841 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 43.376 43.383 velocity_verlet 10 3.0 0.001 0.001 35.141 35.143 dbcsr_multiply_generic 2055 12.4 0.109 0.112 29.702 29.946 qs_scf_new_mos 99 7.5 0.001 0.001 27.261 27.365 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.261 27.364 ot_scf_mini 99 9.5 0.003 0.006 25.961 26.083 multiply_cannon 2055 13.4 0.213 0.225 22.555 23.749 multiply_cannon_loop 2055 14.4 0.618 0.636 21.194 22.280 ot_mini 99 10.5 0.001 0.001 14.428 14.552 rebuild_ks_matrix 110 8.3 0.000 0.000 13.103 13.227 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.015 13.102 13.227 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.524 11.639 multiply_cannon_multrec 16440 15.4 3.956 4.991 9.750 10.722 qs_ot_get_derivative 99 11.5 0.001 0.001 9.931 10.053 mp_waitall_1 139946 16.5 7.474 9.960 7.474 9.960 init_scf_loop 11 6.9 0.001 0.005 9.431 9.437 prepare_preconditioner 11 7.9 0.000 0.000 7.492 7.507 make_preconditioner 11 8.9 0.000 0.001 7.492 7.507 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.788 7.182 sum_up_and_integrate 110 10.3 0.060 0.064 7.098 7.113 integrate_v_rspace 110 11.3 0.003 0.008 7.038 7.055 multiply_cannon_metrocomm3 16440 15.4 0.044 0.045 4.208 6.820 qs_ot_get_p 110 10.4 0.001 0.001 6.496 6.619 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.586 6.596 calculate_rho_elec 110 8.6 0.058 0.059 6.586 6.596 init_scf_run 11 5.9 0.000 0.001 6.054 6.055 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.054 6.054 dbcsr_mm_accdrv_process 34862 16.1 4.663 5.354 5.649 5.858 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.032 5.433 apply_single 110 13.6 0.000 0.000 5.032 5.432 make_m2s 4110 13.4 0.050 0.051 4.727 5.135 make_images 4110 14.4 0.397 0.524 4.613 5.019 density_rs2pw 110 9.6 0.004 0.004 3.558 4.738 ot_diis_step 99 11.5 0.011 0.011 4.465 4.467 qs_ot_p2m_diag 48 11.0 0.042 0.044 4.429 4.432 multiply_cannon_sync_h2d 16440 15.4 3.681 4.322 3.681 4.322 rs_pw_transfer 902 11.9 0.010 0.011 3.129 4.288 pw_transfer 1331 11.6 0.066 0.073 4.135 4.145 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.028 4.040 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.033 4.034 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.350 4.028 fft_wrap_pw1pw2_140 451 13.1 0.213 0.216 3.386 3.400 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.375 3.377 grid_integrate_task_list 110 12.3 3.157 3.357 3.157 3.357 cp_fm_redistribute_end 48 14.0 2.111 3.345 2.113 3.346 fft3d_ps 1111 14.6 1.093 1.104 3.326 3.339 cp_fm_diag_elpa_base 48 14.0 1.158 3.158 1.222 3.277 wfi_extrapolate 11 7.9 0.001 0.001 3.224 3.224 make_images_data 4110 15.4 0.043 0.049 2.576 3.124 cp_fm_cholesky_invert 11 10.9 3.046 3.052 3.046 3.052 hybrid_alltoall_any 4261 16.3 0.105 0.375 2.285 3.031 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 2.881 2.881 calculate_first_density_matrix 1 7.0 0.000 0.003 2.740 2.742 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.687 2.740 mp_alltoall_d11v 2046 13.8 2.254 2.740 2.254 2.740 potential_pw2rs 110 12.3 0.011 0.011 2.651 2.669 calculate_dm_sparse 110 9.5 0.001 0.001 2.620 2.652 mp_waitany 17072 13.8 1.423 2.633 1.423 2.633 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.482 2.527 grid_collocate_task_list 110 9.6 2.084 2.468 2.084 2.468 mp_sum_l 6514 12.8 1.813 2.350 1.813 2.350 multiply_cannon_metrocomm4 14385 15.4 0.044 0.048 0.859 2.316 rs_pw_transfer_RS2PW_140 121 11.5 0.174 0.179 1.039 2.195 mp_irecv_dv 48980 15.7 0.790 2.186 0.790 2.186 dbcsr_complete_redistribute 325 12.2 0.325 0.370 1.592 2.114 qs_energies_init_hamiltonians 11 5.9 0.002 0.003 2.052 2.057 cp_fm_upper_to_full 70 13.6 1.471 2.004 1.471 2.004 cp_fm_cholesky_decompose 22 10.9 1.824 1.840 1.824 1.840 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.802 1.818 mp_allgather_i34 2055 14.4 0.685 1.696 0.685 1.696 copy_fm_to_dbcsr 174 11.2 0.001 0.002 1.083 1.603 rs_gather_matrices 110 12.3 0.234 0.266 1.127 1.587 jit_kernel_multiply 8 16.6 0.597 1.516 0.597 1.516 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.500 1.514 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.354 1.497 mp_alltoall_z22v 1111 16.6 1.465 1.496 1.465 1.496 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.446 1.458 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=65.388000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=624.272727, yerr=8.781084 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8f18a4e0cc0e0e7c2e20248e70b07b7bca2af436_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 736.559104E+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 3521 65576. MP_Allreduce 9838 600. MP_Sync 100 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 MP_comm_split 48 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.037 0.070 67.785 67.787 qs_mol_dyn_low 1 2.0 0.005 0.007 67.455 67.465 qs_forces 11 3.9 0.002 0.002 67.355 67.357 qs_energies 11 4.9 0.002 0.005 63.903 63.908 scf_env_do_scf 11 5.9 0.001 0.005 55.234 55.237 scf_env_do_scf_inner_loop 99 6.5 0.003 0.006 43.040 43.041 velocity_verlet 10 3.0 0.001 0.001 38.316 38.319 dbcsr_multiply_generic 2055 12.4 0.116 0.118 30.096 30.375 qs_scf_new_mos 99 7.5 0.001 0.001 27.385 27.487 qs_scf_loop_do_ot 99 8.5 0.001 0.002 27.385 27.487 ot_scf_mini 99 9.5 0.003 0.006 25.749 25.838 multiply_cannon 2055 13.4 0.245 0.260 22.747 24.063 multiply_cannon_loop 2055 14.4 0.886 0.908 21.212 21.849 ot_mini 99 10.5 0.001 0.001 14.590 14.695 multiply_cannon_multrec 24660 15.4 4.238 6.723 12.655 13.851 rebuild_ks_matrix 110 8.3 0.000 0.000 12.701 12.809 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.016 12.700 12.809 init_scf_loop 11 6.9 0.001 0.005 12.149 12.150 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.238 11.336 qs_ot_get_derivative 99 11.5 0.001 0.001 10.386 10.478 prepare_preconditioner 11 7.9 0.001 0.002 10.296 10.313 make_preconditioner 11 8.9 0.001 0.005 10.295 10.313 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.494 9.976 dbcsr_mm_accdrv_process 52304 16.0 6.929 8.519 8.273 9.208 sum_up_and_integrate 110 10.3 0.066 0.069 6.841 6.852 integrate_v_rspace 110 11.3 0.003 0.003 6.774 6.789 mp_waitall_1 121746 16.5 4.738 6.513 4.738 6.513 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.366 6.377 calculate_rho_elec 110 8.6 0.077 0.081 6.366 6.376 qs_ot_get_p 110 10.4 0.001 0.001 6.057 6.187 make_m2s 4110 13.4 0.059 0.062 5.635 6.000 make_images 4110 14.4 0.576 0.697 5.494 5.857 init_scf_run 11 5.9 0.000 0.001 5.769 5.770 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.769 5.769 cp_fm_upper_to_full 70 13.8 3.295 4.702 3.295 4.702 qs_ot_p2m_diag 48 11.0 0.054 0.063 4.202 4.216 ot_diis_step 99 11.5 0.011 0.011 4.162 4.163 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.044 4.132 apply_single 110 13.6 0.000 0.000 4.044 4.132 pw_transfer 1331 11.6 0.065 0.074 4.087 4.120 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.981 4.018 dbcsr_complete_redistribute 325 12.2 0.414 0.464 2.723 3.892 density_rs2pw 110 9.6 0.004 0.004 3.238 3.744 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.709 3.710 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.380 3.428 fft_wrap_pw1pw2_140 451 13.1 0.202 0.213 3.367 3.415 multiply_cannon_metrocomm3 24660 15.4 0.036 0.037 1.523 3.393 grid_integrate_task_list 110 12.3 3.256 3.373 3.256 3.373 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.181 3.337 fft3d_ps 1111 14.6 1.096 1.137 3.272 3.301 hybrid_alltoall_any 4261 16.3 0.119 0.459 2.478 3.291 multiply_cannon_sync_h2d 24660 15.4 3.164 3.278 3.164 3.278 make_images_data 4110 15.4 0.046 0.050 2.862 3.240 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.144 3.196 wfi_extrapolate 11 7.9 0.001 0.001 3.130 3.130 rs_pw_transfer 902 11.9 0.010 0.011 2.604 3.127 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.054 3.059 cp_fm_redistribute_end 48 14.0 1.520 3.009 1.522 3.011 cp_fm_diag_elpa_base 48 14.0 1.404 2.867 1.483 2.977 cp_fm_cholesky_invert 11 10.9 2.961 2.974 2.961 2.974 calculate_dm_sparse 110 9.5 0.001 0.001 2.936 2.968 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.791 2.932 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.917 2.919 mp_alltoall_i22 605 13.7 1.682 2.847 1.682 2.847 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.533 2.578 calculate_first_density_matrix 1 7.0 0.000 0.002 2.544 2.547 potential_pw2rs 110 12.3 0.012 0.013 2.435 2.442 grid_collocate_task_list 110 9.6 2.182 2.433 2.182 2.433 jit_kernel_multiply 13 15.9 1.014 2.376 1.014 2.376 qs_energies_init_hamiltonians 11 5.9 0.023 0.028 2.312 2.318 mp_alltoall_d11v 2046 13.8 2.019 2.287 2.019 2.287 cp_fm_cholesky_decompose 22 10.9 1.766 1.813 1.766 1.813 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.769 1.801 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.714 1.724 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.596 1.698 mp_allgather_i34 2055 14.4 0.618 1.629 0.618 1.629 acc_transpose_blocks 24660 15.4 0.106 0.109 1.554 1.604 multiply_cannon_metrocomm4 20550 15.4 0.057 0.059 0.842 1.570 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.547 1.564 mp_waitany 13376 13.8 1.096 1.559 1.096 1.559 mp_sum_l 6514 12.8 1.031 1.540 1.031 1.540 mp_irecv_dv 62702 16.1 0.745 1.491 0.745 1.491 mp_alltoall_z22v 1111 16.6 1.315 1.419 1.315 1.419 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=67.787000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=693.363636, yerr=9.393157 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8f18a4e0cc0e0e7c2e20248e70b07b7bca2af436_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 820.379648E+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 3521 65574. MP_Allreduce 9838 640. MP_Sync 100 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 MP_comm_split 48 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 0.040 0.044 59.313 59.315 qs_mol_dyn_low 1 2.0 0.003 0.004 58.985 58.996 qs_forces 11 3.9 0.002 0.002 58.874 58.876 qs_energies 11 4.9 0.001 0.001 55.123 55.130 scf_env_do_scf 11 5.9 0.000 0.001 46.336 46.337 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 37.725 37.726 velocity_verlet 10 3.0 0.001 0.001 33.328 33.334 dbcsr_multiply_generic 2055 12.4 0.105 0.107 23.729 23.873 qs_scf_new_mos 99 7.5 0.001 0.001 21.892 21.949 qs_scf_loop_do_ot 99 8.5 0.001 0.001 21.891 21.948 ot_scf_mini 99 9.5 0.002 0.002 20.609 20.636 multiply_cannon 2055 13.4 0.250 0.264 17.801 18.931 multiply_cannon_loop 2055 14.4 0.322 0.335 16.332 16.638 rebuild_ks_matrix 110 8.3 0.000 0.000 12.445 12.483 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.015 12.444 12.483 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.068 11.100 ot_mini 99 10.5 0.001 0.001 10.986 11.008 init_scf_loop 11 6.9 0.000 0.000 8.561 8.562 mp_waitall_1 103326 16.6 6.567 8.541 6.567 8.541 multiply_cannon_multrec 8220 15.4 3.215 4.437 7.449 8.484 qs_ot_get_derivative 99 11.5 0.001 0.001 7.107 7.133 sum_up_and_integrate 110 10.3 0.079 0.081 6.818 6.831 prepare_preconditioner 11 7.9 0.000 0.000 6.808 6.813 make_preconditioner 11 8.9 0.000 0.000 6.808 6.813 integrate_v_rspace 110 11.3 0.003 0.003 6.738 6.750 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.691 6.710 calculate_rho_elec 110 8.6 0.115 0.116 6.691 6.709 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.370 6.455 qs_ot_get_p 110 10.4 0.001 0.001 5.559 5.581 init_scf_run 11 5.9 0.000 0.001 5.564 5.564 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.563 5.564 dbcsr_mm_accdrv_process 17442 15.9 2.860 3.771 4.103 5.004 make_m2s 4110 13.4 0.038 0.040 4.460 4.731 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 3.194 4.723 make_images 4110 14.4 0.640 0.696 4.330 4.598 pw_transfer 1331 11.6 0.066 0.071 4.373 4.384 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.266 4.280 qs_ot_p2m_diag 48 11.0 0.081 0.084 4.028 4.032 ot_diis_step 99 11.5 0.012 0.012 3.851 3.851 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.746 3.787 apply_single 110 13.6 0.000 0.000 3.746 3.787 density_rs2pw 110 9.6 0.004 0.004 3.398 3.781 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.709 3.710 fft_wrap_pw1pw2_140 451 13.1 0.216 0.219 3.620 3.638 grid_integrate_task_list 110 12.3 3.361 3.523 3.361 3.523 fft3d_ps 1111 14.6 1.148 1.168 3.507 3.517 cp_fm_cholesky_invert 11 10.9 3.376 3.380 3.376 3.380 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.120 3.121 cp_fm_redistribute_end 48 14.0 0.801 3.080 0.806 3.081 cp_fm_diag_elpa_base 48 14.0 2.089 2.867 2.266 3.033 multiply_cannon_sync_h2d 8220 15.4 2.921 3.026 2.921 3.026 wfi_extrapolate 11 7.9 0.001 0.001 3.016 3.016 make_images_data 4110 15.4 0.038 0.043 2.483 2.948 hybrid_alltoall_any 4261 16.3 0.199 0.860 2.381 2.852 rs_pw_transfer 902 11.9 0.010 0.010 2.454 2.840 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.757 2.759 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.667 2.668 calculate_dm_sparse 110 9.5 0.001 0.001 2.576 2.617 grid_collocate_task_list 110 9.6 2.278 2.538 2.278 2.538 calculate_first_density_matrix 1 7.0 0.000 0.000 2.439 2.440 potential_pw2rs 110 12.3 0.015 0.015 2.318 2.325 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.244 2.259 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.058 2.074 mp_alltoall_d11v 2046 13.8 1.802 2.056 1.802 2.056 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.992 2.006 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.773 1.992 cp_fm_cholesky_decompose 22 10.9 1.960 1.980 1.960 1.980 mp_allgather_i34 2055 14.4 0.625 1.754 0.625 1.754 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.717 1.723 dbcsr_complete_redistribute 325 12.2 0.563 0.584 1.619 1.720 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.589 1.712 mp_waitany 9240 13.8 1.171 1.581 1.171 1.581 multiply_cannon_metrocomm1 8220 15.4 0.021 0.022 0.977 1.530 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.501 1.511 qs_create_task_list 11 7.9 0.001 0.001 1.277 1.384 generate_qs_task_list 11 8.9 0.381 0.448 1.276 1.383 jit_kernel_multiply 8 15.7 0.937 1.383 0.937 1.383 mp_alltoall_z22v 1111 16.6 1.246 1.269 1.246 1.269 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.240 1.268 rs_pw_transfer_RS2PW_140 121 11.5 0.162 0.166 0.847 1.236 x_to_yz 550 16.1 0.574 0.586 1.187 1.195 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=59.315000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=776.909091, yerr=8.027844 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8f18a4e0cc0e0e7c2e20248e70b07b7bca2af436_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 1.369936E+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 0.024 0.062 140.167 140.168 qs_mol_dyn_low 1 2.0 0.005 0.020 110.678 139.739 qs_forces 11 3.9 0.002 0.002 106.443 106.449 qs_energies 11 4.9 0.001 0.001 102.230 102.247 scf_env_do_scf 11 5.9 0.001 0.001 77.760 77.760 velocity_verlet 10 3.0 0.001 0.001 72.120 72.128 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 48.494 48.496 md_output 10 3.0 0.000 0.000 4.160 33.260 mp_max_l 33 2.8 29.062 33.214 29.062 33.214 dbcsr_multiply_generic 2055 12.4 0.129 0.135 30.306 30.431 init_scf_loop 11 6.9 0.000 0.000 29.190 29.193 qs_scf_new_mos 99 7.5 0.001 0.001 28.779 28.842 qs_scf_loop_do_ot 99 8.5 0.001 0.001 28.779 28.841 prepare_preconditioner 11 7.9 0.000 0.000 27.058 27.065 make_preconditioner 11 8.9 0.000 0.000 27.058 27.065 ot_scf_mini 99 9.5 0.002 0.002 26.948 26.995 make_full_inverse_cholesky 11 9.9 0.000 0.000 21.328 26.512 dbcsr_complete_redistribute 325 12.2 1.027 1.045 21.377 24.440 copy_fm_to_dbcsr 174 11.2 0.001 0.001 20.273 23.334 multiply_cannon 2055 13.4 0.347 0.371 22.300 23.125 md_write_output 11 3.9 2.763 22.102 2.768 22.129 multiply_cannon_loop 2055 14.4 0.343 0.346 20.390 20.780 init_scf_run 11 5.9 0.000 0.001 20.076 20.078 scf_env_initial_rho_setup 11 6.9 0.001 0.001 20.076 20.078 cp_fm_upper_to_full 70 14.2 12.577 18.000 12.577 18.000 wfi_extrapolate 11 7.9 0.001 0.001 17.497 17.499 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 16.958 16.965 mp_alltoall_d11v 2046 13.8 16.245 16.321 16.245 16.321 rebuild_ks_matrix 110 8.3 0.001 0.001 14.998 15.073 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 14.997 15.073 ot_mini 99 10.5 0.001 0.001 14.671 14.733 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.638 13.709 write_restart 10 4.0 1.393 11.143 1.395 11.145 qs_ot_get_derivative 99 11.5 0.001 0.001 9.855 9.909 multiply_cannon_multrec 8220 15.4 4.381 4.596 9.627 9.742 mp_waitall_1 84994 16.7 8.647 9.626 8.647 9.626 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.715 8.752 qs_rho_update_rho_low 110 7.6 0.001 0.001 8.419 8.453 calculate_rho_elec 110 8.6 0.227 0.228 8.418 8.453 mp_alltoall_i22 605 13.7 5.359 8.429 5.359 8.429 sum_up_and_integrate 110 10.3 0.151 0.153 7.863 7.883 integrate_v_rspace 110 11.3 0.003 0.004 7.712 7.731 qs_ot_get_p 110 10.4 0.001 0.001 6.735 6.774 make_m2s 4110 13.4 0.043 0.043 5.803 6.388 make_images 4110 14.4 0.877 0.913 5.614 6.199 cp_fm_cholesky_invert 11 10.9 5.957 5.962 5.957 5.962 pw_transfer 1331 11.6 0.075 0.075 5.871 5.879 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 5.754 5.763 multiply_cannon_metrocomm3 8220 15.4 0.018 0.019 5.345 5.656 dbcsr_mm_accdrv_process 11614 15.7 3.266 3.343 5.101 5.355 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.859 5.307 apply_single 110 13.6 0.000 0.000 4.859 5.307 fft3d_ps 1111 14.6 1.304 1.312 4.947 4.954 qs_ot_p2m_diag 48 11.0 0.151 0.155 4.927 4.936 fft_wrap_pw1pw2_140 451 13.1 0.221 0.223 4.879 4.886 ot_diis_step 99 11.5 0.015 0.016 4.780 4.780 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.441 4.442 density_rs2pw 110 9.6 0.004 0.004 4.312 4.346 hybrid_alltoall_any 4261 16.3 0.257 0.558 3.223 4.073 make_images_data 4110 15.4 0.041 0.044 3.284 4.040 multiply_cannon_sync_h2d 8220 15.4 3.952 3.960 3.952 3.960 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.775 3.776 cp_fm_diag_elpa_base 48 14.0 3.235 3.424 3.772 3.772 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.760 3.767 grid_integrate_task_list 110 12.3 3.687 3.739 3.687 3.739 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.103 3.577 calculate_dm_sparse 110 9.5 0.001 0.001 3.266 3.285 potential_pw2rs 110 12.3 0.021 0.021 2.828 2.834 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.773 2.807 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=140.168000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1226.272727, yerr=69.146867 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8f18a4e0cc0e0e7c2e20248e70b07b7bca2af436_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 632.905728E+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 4085 56760. MP_Allreduce 11253 785. MP_Sync 170 MP_Alltoall 2226 2520945. MP_SendRecv 24320 18752. MP_ISendRecv 24320 18752. MP_Wait 42476 MP_comm_split 83 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 0.040 0.114 235.680 235.712 qs_mol_dyn_low 1 2.0 0.005 0.014 235.070 235.085 qs_forces 11 3.9 0.003 0.004 233.458 233.461 qs_energies 11 4.9 0.003 0.012 227.800 227.819 scf_env_do_scf 11 5.9 0.001 0.007 210.957 210.960 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 189.580 189.582 dbcsr_multiply_generic 2507 12.6 0.182 0.204 139.297 139.967 velocity_verlet 10 3.0 0.001 0.002 137.798 137.815 qs_scf_new_mos 117 7.6 0.001 0.001 136.501 136.715 qs_scf_loop_do_ot 117 8.6 0.001 0.002 136.500 136.715 ot_scf_mini 117 9.6 0.003 0.004 129.860 129.991 multiply_cannon 2507 13.6 0.238 0.245 111.945 116.371 multiply_cannon_loop 2507 14.6 2.116 2.172 109.716 114.248 ot_mini 117 10.6 0.001 0.001 66.597 66.751 mp_waitall_1 267128 16.5 39.904 46.134 39.904 46.134 multiply_cannon_multrec 60168 15.6 32.989 35.071 41.373 43.283 qs_ot_get_derivative 117 11.6 0.001 0.001 41.784 41.976 qs_ot_get_p 128 10.4 0.001 0.001 41.331 41.570 qs_ot_p2m_diag 83 11.4 0.077 0.091 34.345 34.444 rebuild_ks_matrix 128 8.3 0.001 0.001 34.220 34.424 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.018 34.220 34.423 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.799 30.982 multiply_cannon_metrocomm3 60168 15.6 0.115 0.120 17.669 30.537 multiply_cannon_sync_h2d 60168 15.6 27.118 28.627 27.118 28.627 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.249 24.861 apply_single 128 13.6 0.001 0.001 24.249 24.861 ot_diis_step 117 11.6 0.008 0.008 24.468 24.469 mp_sum_l 7870 13.0 11.961 23.246 11.961 23.246 multiply_cannon_metrocomm1 60168 15.6 0.088 0.093 15.442 22.706 init_scf_loop 11 6.9 0.001 0.005 21.294 21.295 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.625 19.733 cp_dbcsr_syevd 83 12.4 0.004 0.005 17.673 17.674 prepare_preconditioner 11 7.9 0.000 0.001 16.610 16.643 make_preconditioner 11 8.9 0.000 0.002 16.610 16.643 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.790 15.949 sum_up_and_integrate 128 10.3 0.089 0.106 14.694 14.713 cp_fm_diag_elpa 83 13.4 0.000 0.000 14.672 14.679 integrate_v_rspace 128 11.3 0.003 0.004 14.604 14.625 cp_fm_redistribute_end 83 14.4 11.594 14.584 11.605 14.586 make_m2s 5014 13.6 0.105 0.111 14.143 14.581 make_images 5014 14.6 0.400 0.419 13.962 14.409 cp_fm_diag_elpa_base 83 14.4 2.935 14.170 2.967 14.291 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.133 14.253 calculate_rho_elec 128 8.7 0.045 0.064 14.132 14.253 init_scf_run 11 5.9 0.000 0.001 12.613 12.614 scf_env_initial_rho_setup 11 6.9 0.001 0.002 12.613 12.614 mp_bcast_b 1988 13.7 12.296 12.570 12.296 12.570 external_control 127 7.2 0.128 12.211 12.392 12.539 density_rs2pw 128 9.7 0.006 0.007 7.516 10.727 cp_fm_cholesky_invert 11 10.9 9.529 9.537 9.529 9.537 rs_pw_transfer 1046 11.9 0.016 0.018 6.208 9.399 wfi_extrapolate 11 7.9 0.001 0.001 9.139 9.139 calculate_dm_sparse 128 9.5 0.001 0.001 8.472 8.585 pw_transfer 1547 11.6 0.076 0.106 8.233 8.493 dbcsr_mm_accdrv_process 124484 16.2 3.322 3.476 7.946 8.476 fft_wrap_pw1pw2 1291 12.7 0.010 0.013 8.027 8.290 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.996 8.101 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.932 8.021 make_images_data 5014 15.6 0.065 0.072 6.900 7.898 grid_integrate_task_list 128 12.3 7.079 7.561 7.079 7.561 hybrid_alltoall_any 5200 16.5 0.289 2.257 6.019 7.378 fft_wrap_pw1pw2_140 523 13.2 0.444 0.511 6.873 7.196 fft3d_ps 1291 14.7 2.114 2.894 6.813 6.991 cp_dbcsr_sm_fm_multiply 37 9.5 0.003 0.003 6.806 6.813 mp_waitany 16020 13.9 2.835 6.129 2.835 6.129 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 5.869 5.952 mp_alltoall_d11v 2415 14.1 4.305 5.898 4.305 5.898 grid_collocate_task_list 128 9.7 4.544 5.774 4.544 5.774 rs_pw_transfer_RS2PW_140 139 11.5 0.278 0.296 2.187 5.405 potential_pw2rs 128 12.3 0.009 0.011 5.016 5.040 cp_fm_cholesky_decompose 22 10.9 4.846 4.865 4.846 4.865 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=235.712000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=597.000000, yerr=8.135221 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8f18a4e0cc0e0e7c2e20248e70b07b7bca2af436_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 829.038592E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100943E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834845264 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4085 57203. MP_Allreduce 11252 946. MP_Sync 170 MP_Alltoall 1969 5622776. MP_SendRecv 12032 47072. MP_ISendRecv 12032 47072. MP_Wait 25916 MP_comm_split 83 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 0.037 0.120 225.242 225.254 qs_mol_dyn_low 1 2.0 0.005 0.011 224.561 224.576 qs_forces 11 3.9 0.004 0.011 210.481 210.483 qs_energies 11 4.9 0.003 0.007 203.687 203.697 scf_env_do_scf 11 5.9 0.001 0.003 187.131 187.141 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 153.747 153.750 velocity_verlet 10 3.0 0.001 0.002 150.700 150.994 dbcsr_multiply_generic 2507 12.6 0.186 0.192 112.086 113.359 qs_scf_new_mos 117 7.6 0.001 0.001 110.643 110.974 qs_scf_loop_do_ot 117 8.6 0.001 0.002 110.642 110.973 ot_scf_mini 117 9.6 0.004 0.006 91.728 92.134 multiply_cannon 2507 13.6 0.479 0.529 77.814 82.446 multiply_cannon_loop 2507 14.6 1.267 1.304 74.489 77.213 mp_waitall_1 214728 16.6 38.311 52.095 38.311 52.095 ot_mini 117 10.6 0.001 0.006 50.260 50.615 multiply_cannon_multrec 30084 15.6 22.039 26.196 31.632 36.266 rebuild_ks_matrix 128 8.3 0.001 0.001 33.686 34.134 qs_ks_build_kohn_sham_matrix 128 9.3 0.018 0.025 33.686 34.133 init_scf_loop 11 6.9 0.002 0.010 33.295 33.300 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.367 30.779 make_m2s 5014 13.6 0.090 0.095 28.156 29.638 make_images 5014 14.6 1.156 1.340 27.945 29.425 prepare_preconditioner 11 7.9 0.000 0.001 28.850 28.906 make_preconditioner 11 8.9 0.000 0.002 28.850 28.906 qs_ot_get_derivative 117 11.6 0.001 0.003 28.281 28.667 multiply_cannon_metrocomm3 30084 15.6 0.096 0.102 15.940 28.346 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.592 28.126 make_images_data 5014 15.6 0.064 0.072 22.134 23.835 qs_ot_get_p 128 10.4 0.001 0.002 23.107 23.542 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 21.972 23.147 apply_single 128 13.6 0.001 0.001 21.971 23.146 hybrid_alltoall_any 5200 16.5 0.342 1.518 20.545 23.069 ot_diis_step 117 11.6 0.015 0.024 21.802 21.804 multiply_cannon_sync_h2d 30084 15.6 19.390 21.260 19.390 21.260 calculate_dm_sparse 128 9.5 0.001 0.001 20.570 20.702 qs_ot_p2m_diag 83 11.4 0.188 0.216 18.109 18.147 cp_fm_cholesky_invert 11 10.9 16.970 16.983 16.970 16.983 cp_dbcsr_syevd 83 12.4 0.005 0.006 16.918 16.919 sum_up_and_integrate 128 10.3 0.114 0.132 15.017 15.049 integrate_v_rspace 128 11.3 0.004 0.005 14.903 14.937 mp_sum_dm 438 4.9 13.858 14.206 13.858 14.206 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.162 14.193 calculate_rho_elec 128 8.7 0.088 0.110 14.161 14.192 md_output 10 3.0 0.000 0.000 0.299 14.057 md_write_output 11 3.9 0.001 0.039 0.295 14.054 update_particle_set 20 4.0 0.000 0.000 13.757 14.051 write_trajectory 44 4.9 0.292 13.991 0.293 14.015 cp_fm_diag_elpa 83 13.4 0.000 0.001 13.684 13.695 cp_fm_redistribute_end 83 14.4 7.988 13.609 8.002 13.614 cp_fm_diag_elpa_base 83 14.4 5.368 13.095 5.591 13.458 init_scf_run 11 5.9 0.000 0.001 11.621 11.623 scf_env_initial_rho_setup 11 6.9 0.001 0.003 11.621 11.622 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 11.227 11.499 multiply_cannon_metrocomm4 27577 15.6 0.094 0.106 3.727 10.312 density_rs2pw 128 9.7 0.006 0.007 7.597 10.064 mp_irecv_dv 69486 16.3 3.538 9.934 3.538 9.934 dbcsr_mm_accdrv_process 62242 16.2 4.446 5.260 9.050 9.631 pw_transfer 1547 11.6 0.087 0.106 8.954 9.032 fft_wrap_pw1pw2 1291 12.7 0.011 0.012 8.728 8.806 rs_pw_transfer 1046 11.9 0.014 0.016 6.075 8.514 wfi_extrapolate 11 7.9 0.001 0.003 8.388 8.388 fft_wrap_pw1pw2_140 523 13.2 0.472 0.528 7.576 7.671 grid_integrate_task_list 128 12.3 7.181 7.580 7.181 7.580 fft3d_ps 1291 14.7 2.755 2.926 7.093 7.133 cp_fm_cholesky_decompose 22 10.9 7.035 7.131 7.035 7.131 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.207 6.971 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.124 6.133 mp_sum_l 7870 13.0 4.374 5.934 4.374 5.934 grid_collocate_task_list 128 9.7 4.712 5.867 4.712 5.867 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.419 5.503 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.300 5.424 potential_pw2rs 128 12.3 0.015 0.018 5.317 5.337 mp_waitany 11748 13.9 2.577 5.067 2.577 5.067 mp_allgather_i34 2507 14.6 1.898 4.752 1.898 4.752 mp_alltoall_d11v 2415 14.1 3.923 4.607 3.923 4.607 rs_pw_transfer_RS2PW_140 139 11.5 0.354 0.379 2.150 4.584 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=225.254000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=788.727273, yerr=5.361325 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8f18a4e0cc0e0e7c2e20248e70b07b7bca2af436_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963542011904 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019182452736 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019182452736 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796564E+12 0.0% 0.0% 100.0% flops max/rank 2.906045E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499488 0.0% 0.0% 100.0% number of processed stacks 3951168 0.0% 0.0% 100.0% average stack size 0.0 0.0 1697.1 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 935.882752E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1033760 MPI messages size (bytes): total size 2.695213E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.607194E+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 279168 36591108096 131072 < size <= 4194304 654272 987691483136 4194304 < size <= 16777216 65184 925172558800 16777216 < size 28448 745747251200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4077 57304. MP_Allreduce 11231 986. MP_Sync 168 MP_Alltoall 1700 9383497. MP_SendRecv 7874 75008. MP_ISendRecv 7874 75008. MP_Wait 21654 MP_comm_split 82 MP_ISend 11660 275234. MP_IRecv 11660 275234. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.053 0.062 211.763 211.765 qs_mol_dyn_low 1 2.0 0.005 0.007 211.300 211.314 qs_forces 11 3.9 0.003 0.005 192.006 192.012 qs_energies 11 4.9 0.001 0.002 185.370 185.379 scf_env_do_scf 11 5.9 0.001 0.001 169.958 169.959 velocity_verlet 10 3.0 0.001 0.001 144.806 145.407 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 134.162 134.163 qs_scf_new_mos 116 7.6 0.001 0.001 88.157 96.359 qs_scf_loop_do_ot 116 8.6 0.001 0.001 88.156 96.358 dbcsr_multiply_generic 2485 12.5 0.180 0.185 95.400 96.327 ot_scf_mini 116 9.6 0.003 0.004 84.048 92.339 multiply_cannon 2485 13.5 0.497 0.518 69.215 77.462 multiply_cannon_loop 2485 14.5 0.853 0.878 65.993 74.604 ot_mini 116 10.6 0.001 0.001 48.808 57.075 mp_waitall_1 169034 16.6 32.331 47.089 32.331 47.089 rebuild_ks_matrix 127 8.3 0.001 0.001 38.574 45.103 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.018 38.574 45.102 qs_ks_update_qs_env 127 7.6 0.001 0.001 35.517 42.011 multiply_cannon_metrocomm3 19880 15.5 0.060 0.063 22.684 37.731 qs_ot_get_derivative 116 11.6 0.001 0.002 29.230 37.521 init_scf_loop 11 6.9 0.000 0.000 35.693 35.695 prepare_preconditioner 11 7.9 0.000 0.000 31.610 31.653 make_preconditioner 11 8.9 0.000 0.000 31.610 31.653 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.236 30.659 multiply_cannon_multrec 19880 15.5 13.451 16.779 22.071 25.339 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 15.279 23.413 qs_ot_get_p 127 10.4 0.001 0.001 20.598 21.044 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.664 20.654 apply_single 127 13.6 0.001 0.001 19.664 20.654 ot_diis_step 116 11.6 0.017 0.018 19.446 19.447 mp_sum_dm 438 4.9 18.726 19.373 18.726 19.373 update_particle_set 20 4.0 0.000 0.000 18.635 19.236 md_write_output 11 3.9 0.601 19.200 0.603 19.235 mp_sum_l 7804 13.0 9.717 18.760 9.717 18.760 md_output 10 3.0 0.000 0.000 0.584 18.520 qs_ot_p2m_diag 82 11.4 0.262 0.269 16.221 16.237 multiply_cannon_sync_h2d 19880 15.5 14.281 15.929 14.281 15.929 make_m2s 4970 13.5 0.080 0.085 14.658 15.574 make_images 4970 14.5 1.170 1.261 14.427 15.342 cp_dbcsr_syevd 82 12.4 0.005 0.005 15.164 15.167 cp_fm_cholesky_invert 11 10.9 14.846 14.856 14.846 14.856 sum_up_and_integrate 127 10.3 0.130 0.141 14.666 14.693 integrate_v_rspace 127 11.3 0.004 0.004 14.535 14.563 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.343 14.364 calculate_rho_elec 127 8.7 0.131 0.146 14.343 14.364 cp_fm_diag_elpa 82 13.4 0.000 0.001 11.978 11.979 cp_fm_redistribute_end 82 14.4 4.540 11.896 4.555 11.898 cp_fm_diag_elpa_base 82 14.4 6.906 11.259 7.323 11.739 init_scf_run 11 5.9 0.000 0.001 10.400 10.401 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.400 10.400 make_images_data 4970 15.5 0.060 0.068 8.848 10.200 hybrid_alltoall_any 5155 16.4 0.429 1.975 7.765 9.460 density_rs2pw 127 9.7 0.006 0.006 7.298 9.310 pw_transfer 1535 11.6 0.086 0.107 8.966 9.089 fft_wrap_pw1pw2 1281 12.7 0.010 0.012 8.743 8.873 multiply_cannon_metrocomm4 17395 15.5 0.060 0.070 3.387 8.803 mp_irecv_dv 49801 16.2 3.268 8.560 3.268 8.560 dbcsr_mm_accdrv_process 41158 16.2 4.434 5.110 8.074 8.221 grid_integrate_task_list 127 12.3 7.267 7.799 7.267 7.799 fft_wrap_pw1pw2_140 519 13.2 0.474 0.512 7.637 7.782 cp_fm_cholesky_decompose 22 10.9 7.598 7.632 7.598 7.632 cp_fm_upper_to_full 104 14.5 5.779 7.512 5.779 7.512 wfi_extrapolate 11 7.9 0.001 0.001 7.393 7.393 rs_pw_transfer 1038 11.9 0.013 0.014 5.254 7.272 fft3d_ps 1281 14.7 2.647 2.873 7.019 7.097 dbcsr_complete_redistribute 393 12.7 1.172 1.222 4.502 6.277 calculate_dm_sparse 127 9.5 0.001 0.001 5.698 5.787 grid_collocate_task_list 127 9.7 4.861 5.758 4.861 5.758 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.410 5.416 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.520 5.177 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.321 5.099 potential_pw2rs 127 12.3 0.020 0.022 4.847 4.860 mp_alltoall_d11v 2401 14.1 4.083 4.827 4.083 4.827 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.499 4.630 mp_allgather_i34 2485 14.5 1.755 4.407 1.755 4.407 mp_waitany 11660 13.9 2.319 4.375 2.319 4.375 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=211.765000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=888.000000, yerr=9.525086 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8f18a4e0cc0e0e7c2e20248e70b07b7bca2af436_performance_tests/18/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 4.320339E+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 5927808 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.2 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.139585E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1133160 MPI messages size (bytes): total size 2.008142E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.772161E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 315952 35695099904 131072 < size <= 4194304 709496 778939400192 4194304 < size <= 16777216 69840 660837789680 16777216 < size 30480 532676608000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4083 57219. MP_Allreduce 11246 1068. MP_Sync 168 MP_Alltoall 1700 12496381. MP_SendRecv 5842 75008. MP_ISendRecv 5842 75008. MP_Wait 22272 MP_comm_split 82 MP_ISend 14840 244848. MP_IRecv 14840 244848. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.032 0.062 218.743 218.744 qs_mol_dyn_low 1 2.0 0.004 0.005 218.310 218.323 qs_forces 11 3.9 0.003 0.003 201.258 201.264 qs_energies 11 4.9 0.003 0.010 194.096 194.107 scf_env_do_scf 11 5.9 0.001 0.002 177.142 177.150 velocity_verlet 10 3.0 0.001 0.001 154.668 155.377 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 130.034 130.035 qs_scf_new_mos 116 7.6 0.001 0.001 78.751 79.049 qs_scf_loop_do_ot 116 8.6 0.001 0.001 78.750 79.048 dbcsr_multiply_generic 2485 12.5 0.194 0.199 78.211 78.854 ot_scf_mini 116 9.6 0.003 0.004 74.351 74.677 multiply_cannon 2485 13.5 0.552 0.598 53.956 56.745 multiply_cannon_loop 2485 14.5 1.171 1.196 50.287 51.749 init_scf_loop 11 6.9 0.001 0.006 46.987 46.988 rebuild_ks_matrix 127 8.3 0.001 0.001 43.824 44.106 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.020 43.823 44.106 prepare_preconditioner 11 7.9 0.000 0.001 42.840 42.857 make_preconditioner 11 8.9 0.000 0.002 42.840 42.856 ot_mini 116 10.6 0.001 0.001 41.679 41.987 make_full_inverse_cholesky 11 9.9 0.000 0.000 36.468 41.510 qs_ks_update_qs_env 127 7.6 0.001 0.001 40.879 41.133 mp_waitall_1 146592 16.7 30.656 40.532 30.656 40.532 multiply_cannon_multrec 29820 15.5 14.193 19.287 26.018 30.569 sum_up_and_integrate 127 10.3 0.139 0.152 28.778 28.805 integrate_v_rspace 127 11.3 0.004 0.004 28.638 28.670 qs_ot_get_derivative 116 11.6 0.001 0.002 22.266 22.586 make_m2s 4970 13.5 0.094 0.098 19.971 21.016 make_images 4970 14.5 1.934 2.248 19.668 20.710 rs_pw_transfer 1038 11.9 0.013 0.014 18.941 20.270 qs_ot_get_p 127 10.4 0.001 0.001 19.119 19.402 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 18.881 19.356 apply_single 127 13.6 0.001 0.001 18.881 19.356 ot_diis_step 116 11.6 0.017 0.018 19.283 19.284 potential_pw2rs 127 12.3 0.023 0.023 18.686 18.698 mp_sum_dm 438 4.9 16.375 17.115 16.375 17.115 update_particle_set 20 4.0 0.000 0.000 16.307 17.015 md_write_output 11 3.9 0.414 9.915 0.710 17.006 cp_fm_cholesky_invert 11 10.9 16.624 16.634 16.624 16.634 cp_fm_upper_to_full 104 14.7 11.123 16.388 11.123 16.388 md_output 10 3.0 0.000 0.000 0.667 15.885 multiply_cannon_metrocomm3 29820 15.5 0.047 0.050 6.016 15.264 qs_ot_p2m_diag 82 11.4 0.338 0.385 15.147 15.199 rs_pw_transfer_PW2RS_20 127 14.3 0.309 0.318 14.251 14.563 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.464 14.516 calculate_rho_elec 127 8.7 0.174 0.189 14.463 14.516 cp_dbcsr_syevd 82 12.4 0.005 0.005 13.889 13.890 dbcsr_complete_redistribute 393 12.7 1.499 1.624 8.900 12.733 multiply_cannon_sync_h2d 29820 15.5 11.619 12.619 11.619 12.619 make_images_data 4970 15.5 0.063 0.068 10.621 12.352 dbcsr_mm_accdrv_process 61748 16.2 7.341 8.308 11.399 11.894 copy_fm_to_dbcsr 208 11.6 0.002 0.002 7.561 11.380 hybrid_alltoall_any 5155 16.4 0.519 2.197 9.369 11.170 init_scf_run 11 5.9 0.000 0.001 10.904 10.905 scf_env_initial_rho_setup 11 6.9 0.001 0.002 10.904 10.905 cp_fm_diag_elpa 82 13.4 0.000 0.000 10.820 10.822 cp_fm_redistribute_end 82 14.4 1.868 10.741 1.883 10.747 cp_fm_diag_elpa_base 82 14.4 8.249 10.124 8.826 10.610 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.349 10.078 mp_alltoall_i22 712 14.1 5.642 9.623 5.642 9.623 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.119 9.359 pw_transfer 1535 11.6 0.087 0.106 9.150 9.217 fft_wrap_pw1pw2 1281 12.7 0.010 0.012 8.924 8.996 density_rs2pw 127 9.7 0.006 0.006 6.996 8.358 fft_wrap_pw1pw2_140 519 13.2 0.478 0.489 7.893 7.975 grid_integrate_task_list 127 12.3 7.446 7.884 7.446 7.884 cp_fm_cholesky_decompose 22 10.9 7.628 7.717 7.628 7.717 wfi_extrapolate 11 7.9 0.001 0.001 7.652 7.652 fft3d_ps 1281 14.7 2.753 2.821 7.162 7.225 write_trajectory 44 4.9 0.295 7.067 0.296 7.091 multiply_cannon_metrocomm4 24850 15.5 0.075 0.084 2.733 6.752 mp_irecv_dv 75445 16.2 2.591 6.498 2.591 6.498 calculate_dm_sparse 127 9.5 0.001 0.001 6.078 6.181 grid_collocate_task_list 127 9.7 5.041 5.799 5.041 5.799 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.538 5.595 mp_alltoall_d11v 2401 14.1 4.693 5.544 4.693 5.544 qs_energies_init_hamiltonians 11 5.9 0.006 0.041 4.604 4.605 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.501 4.597 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.333 4.419 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=218.744000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1077.363636, yerr=17.540979 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8f18a4e0cc0e0e7c2e20248e70b07b7bca2af436_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 1.524650E+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 4003 58193. MP_Allreduce 11005 1175. MP_Sync 86 MP_Alltoall 1700 18828160. 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 0.021 0.037 207.151 207.153 qs_mol_dyn_low 1 2.0 0.003 0.004 206.733 206.747 qs_forces 11 3.9 0.003 0.003 191.950 191.957 qs_energies 11 4.9 0.035 0.039 184.363 184.375 scf_env_do_scf 11 5.9 0.001 0.001 166.981 166.984 velocity_verlet 10 3.0 0.001 0.001 145.345 146.266 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 116.306 116.308 qs_scf_new_mos 116 7.6 0.001 0.001 75.435 75.566 qs_scf_loop_do_ot 116 8.6 0.001 0.001 75.434 75.565 dbcsr_multiply_generic 2485 12.5 0.181 0.185 74.613 75.202 ot_scf_mini 116 9.6 0.003 0.004 71.015 71.150 multiply_cannon 2485 13.5 0.582 0.617 54.924 59.572 multiply_cannon_loop 2485 14.5 0.444 0.456 49.891 50.994 init_scf_loop 11 6.9 0.000 0.000 50.524 50.525 prepare_preconditioner 11 7.9 0.000 0.000 46.514 46.539 make_preconditioner 11 8.9 0.000 0.000 46.514 46.539 make_full_inverse_cholesky 11 9.9 0.000 0.000 44.417 44.689 ot_mini 116 10.6 0.001 0.001 39.940 40.080 mp_waitall_1 124680 16.7 26.326 34.422 26.326 34.422 rebuild_ks_matrix 127 8.3 0.001 0.001 30.076 30.177 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.019 30.076 30.176 qs_ks_update_qs_env 127 7.6 0.001 0.001 27.293 27.383 multiply_cannon_multrec 9940 15.5 10.292 15.975 17.534 22.562 cp_fm_cholesky_decompose 22 10.9 22.076 22.171 22.076 22.171 multiply_cannon_metrocomm3 9940 15.5 0.024 0.025 12.821 20.160 ot_diis_step 116 11.6 0.019 0.020 20.009 20.009 qs_ot_get_derivative 116 11.6 0.001 0.002 19.856 19.989 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.597 19.914 apply_single 127 13.6 0.001 0.001 19.597 19.913 make_m2s 4970 13.5 0.065 0.069 16.026 18.845 cp_fm_cholesky_invert 11 10.9 18.805 18.811 18.805 18.811 make_images 4970 14.5 2.262 2.611 15.723 18.541 qs_ot_get_p 127 10.4 0.001 0.001 18.097 18.285 qs_rho_update_rho_low 127 7.7 0.001 0.001 15.536 15.570 calculate_rho_elec 127 8.7 0.257 0.267 15.535 15.569 sum_up_and_integrate 127 10.3 0.178 0.187 15.149 15.199 integrate_v_rspace 127 11.3 0.004 0.004 14.970 15.028 mp_sum_dm 438 4.9 13.883 14.818 13.883 14.818 update_particle_set 20 4.0 0.000 0.000 13.830 14.751 md_output 10 3.0 0.000 0.000 0.927 14.749 md_write_output 11 3.9 0.019 0.293 0.923 14.747 write_trajectory 44 4.9 0.902 14.429 0.904 14.453 qs_ot_p2m_diag 82 11.4 0.489 0.495 14.283 14.300 cp_dbcsr_syevd 82 12.4 0.005 0.005 13.152 13.153 make_images_data 4970 15.5 0.051 0.060 9.743 12.666 multiply_cannon_sync_h2d 9940 15.5 11.420 12.297 11.420 12.297 hybrid_alltoall_any 5155 16.4 0.835 3.777 9.544 12.219 init_scf_run 11 5.9 0.000 0.001 10.545 10.545 scf_env_initial_rho_setup 11 6.9 0.003 0.004 10.544 10.545 cp_fm_diag_elpa 82 13.4 0.000 0.000 10.124 10.135 cp_fm_diag_elpa_base 82 14.4 9.877 9.962 10.117 10.129 pw_transfer 1535 11.6 0.085 0.093 9.948 9.982 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 9.726 9.763 fft_wrap_pw1pw2_140 519 13.2 0.492 0.513 8.549 8.596 grid_integrate_task_list 127 12.3 7.705 8.138 7.705 8.138 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 7.919 8.021 density_rs2pw 127 9.7 0.005 0.006 7.131 7.986 fft3d_ps 1281 14.7 2.684 2.756 7.903 7.921 multiply_cannon_metrocomm1 9940 15.5 0.029 0.029 4.487 7.916 mp_allgather_i34 2485 14.5 3.248 7.608 3.248 7.608 wfi_extrapolate 11 7.9 0.001 0.001 7.565 7.566 dbcsr_mm_accdrv_process 20590 16.1 2.609 3.432 6.876 7.497 calculate_dm_sparse 127 9.5 0.001 0.001 6.088 6.167 grid_collocate_task_list 127 9.7 5.354 6.010 5.354 6.010 mp_alltoall_d11v 2401 14.1 5.116 5.996 5.116 5.996 multiply_cannon_metrocomm4 7455 15.5 0.023 0.027 1.850 5.635 dbcsr_complete_redistribute 393 12.7 2.106 2.172 5.206 5.607 mp_irecv_dv 28618 15.9 1.814 5.547 1.814 5.547 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 5.303 5.305 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.239 5.251 rs_pw_transfer 1038 11.9 0.012 0.013 4.154 5.040 potential_pw2rs 127 12.3 0.026 0.026 4.618 4.631 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.164 4.217 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=207.153000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1430.090909, yerr=55.395199 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8f18a4e0cc0e0e7c2e20248e70b07b7bca2af436_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456039424 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613072052224 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176077312 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176077312 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.233020E+12 0.0% 0.0% 100.0% flops max/rank 11.786061E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383904 0.0% 0.0% 100.0% number of processed stacks 1980288 0.0% 0.0% 100.0% average stack size 0.0 0.0 3437.1 marketing flops 145.650931E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 2.984157E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101160 MPI messages size (bytes): total size 1.144970E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.318403E+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 45648 35433480192 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592466352 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 58564. MP_Allreduce 11104 1511. MP_Sync 88 MP_Alltoall 1724 36993632. MP_SendRecv 1806 218624. MP_ISendRecv 1806 218624. MP_Wait 9876 MP_ISend 6456 1080169. MP_IRecv 6456 1080169. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.062 0.125 352.349 352.367 qs_mol_dyn_low 1 2.0 0.003 0.004 351.740 351.755 qs_forces 11 3.9 0.003 0.004 315.966 315.969 qs_energies 11 4.9 0.002 0.003 306.547 306.553 scf_env_do_scf 11 5.9 0.001 0.001 284.136 284.153 velocity_verlet 10 3.0 0.001 0.001 249.098 253.567 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 158.912 158.914 init_scf_loop 11 6.9 0.000 0.000 124.967 124.970 prepare_preconditioner 11 7.9 0.000 0.000 119.778 119.815 make_preconditioner 11 8.9 0.000 0.000 119.778 119.815 make_full_inverse_cholesky 11 9.9 0.000 0.000 95.750 116.902 qs_scf_new_mos 118 7.6 0.001 0.001 106.870 107.039 qs_scf_loop_do_ot 118 8.6 0.001 0.001 106.869 107.038 ot_scf_mini 118 9.6 0.004 0.004 101.994 102.130 dbcsr_multiply_generic 2529 12.6 0.215 0.223 98.942 99.899 cp_fm_upper_to_full 106 14.8 52.814 75.802 52.814 75.802 multiply_cannon 2529 13.6 0.721 0.757 71.911 74.746 multiply_cannon_loop 2529 14.6 0.475 0.491 68.133 72.019 ot_mini 118 10.6 0.001 0.001 59.760 59.921 mp_waitall_1 104580 16.8 41.491 47.139 41.491 47.139 dbcsr_complete_redistribute 397 12.7 3.968 4.009 29.687 42.512 rebuild_ks_matrix 129 8.3 0.001 0.001 39.198 39.371 qs_ks_build_kohn_sham_matrix 129 9.3 0.020 0.022 39.197 39.370 copy_fm_to_dbcsr 210 11.7 0.002 0.002 26.272 39.027 qs_ot_get_derivative 118 11.6 0.002 0.002 38.750 38.884 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 23.982 36.640 qs_ks_update_qs_env 129 7.6 0.001 0.001 36.130 36.295 mp_sum_dm 438 4.9 31.305 35.784 31.305 35.784 md_write_output 11 3.9 4.465 35.715 4.470 35.751 update_particle_set 20 4.0 0.000 0.000 31.279 35.748 multiply_cannon_metrocomm3 10116 15.6 0.024 0.024 31.769 35.458 mp_alltoall_i22 720 14.1 21.800 34.769 21.800 34.769 cp_fm_cholesky_invert 11 10.9 33.711 33.717 33.711 33.717 qs_ot_get_p 129 10.4 0.001 0.001 26.506 26.669 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 23.707 23.815 qs_ot_p2m_diag 84 11.4 0.889 0.896 22.299 22.328 make_m2s 5058 13.6 0.076 0.078 20.295 21.302 qs_rho_update_rho_low 129 7.7 0.001 0.001 21.184 21.207 calculate_rho_elec 129 8.7 0.487 0.488 21.184 21.206 ot_diis_step 118 11.6 0.023 0.025 20.960 20.961 make_images 5058 14.6 3.797 4.013 19.812 20.820 cp_dbcsr_syevd 84 12.4 0.006 0.006 20.530 20.532 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.992 20.442 apply_single 129 13.6 0.001 0.001 19.991 20.441 sum_up_and_integrate 129 10.3 0.323 0.325 20.052 20.122 integrate_v_rspace 129 11.3 0.004 0.004 19.728 19.799 multiply_cannon_multrec 10116 15.6 10.538 12.286 18.023 18.129 cp_fm_diag_elpa 84 13.4 0.000 0.000 17.273 17.275 cp_fm_diag_elpa_base 84 14.4 12.874 14.467 17.269 17.269 multiply_cannon_sync_h2d 10116 15.6 15.773 15.784 15.773 15.784 mp_sum_l 7936 13.1 4.042 15.234 4.042 15.234 hybrid_alltoall_any 5245 16.5 1.311 3.047 11.135 13.073 make_images_data 5058 15.6 0.060 0.064 10.938 12.796 pw_transfer 1559 11.6 0.094 0.094 12.630 12.643 fft_wrap_pw1pw2 1301 12.7 0.011 0.011 12.392 12.405 init_scf_run 11 5.9 0.000 0.001 12.298 12.299 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.298 12.299 fft_wrap_pw1pw2_140 527 13.2 0.548 0.556 10.995 11.008 fft3d_ps 1301 14.7 2.767 2.803 10.407 10.428 mp_alltoall_d11v 2429 14.1 8.341 9.210 8.341 9.210 wfi_extrapolate 11 7.9 0.001 0.001 9.168 9.168 dbcsr_mm_accdrv_process 20934 16.1 3.869 5.672 7.244 9.106 cp_fm_cholesky_decompose 22 10.9 8.999 9.008 8.999 9.008 grid_integrate_task_list 129 12.3 8.655 8.829 8.655 8.829 density_rs2pw 129 9.7 0.005 0.005 8.652 8.803 qs_energies_init_hamiltonians 11 5.9 0.003 0.008 8.100 8.104 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=352.367000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2636.272727, yerr=176.525605 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8f18a4e0cc0e0e7c2e20248e70b07b7bca2af436_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 1.262096E+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 6805335. 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 0.042 0.088 125.060 125.062 qs_energies 1 2.0 0.000 0.000 124.458 124.478 ls_scf 1 3.0 0.000 0.002 123.533 123.540 dbcsr_multiply_generic 111 6.7 0.015 0.016 86.587 86.760 multiply_cannon 111 7.7 0.018 0.021 56.756 70.532 multiply_cannon_loop 111 8.7 0.209 0.225 53.390 67.265 ls_scf_main 1 4.0 0.000 0.002 66.308 66.309 density_matrix_trs4 2 5.0 0.002 0.003 60.762 60.846 mp_waitall_1 11031 10.9 23.359 37.915 23.359 37.915 ls_scf_post 1 4.0 0.001 0.005 28.946 28.952 ls_scf_init_scf 1 4.0 0.000 0.002 28.279 28.283 ls_scf_init_matrix_S 1 5.0 0.000 0.001 27.145 27.204 rs_pw_transfer 27 10.4 0.000 0.001 6.085 26.544 mp_waitany 642 12.2 5.905 26.368 5.905 26.368 rs_pw_transfer_RS2PW_150 5 9.2 0.035 0.049 5.876 26.339 write_mo_free_results 1 5.0 18.970 25.955 26.074 26.128 calculate_hirshfeld_normalizat 1 6.0 0.048 0.053 5.691 26.038 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.057 25.071 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 10.622 24.545 mp_sum_l 807 5.4 16.262 18.376 16.262 18.376 multiply_cannon_multrec 2664 9.7 8.157 8.965 15.474 17.346 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 15.466 17.244 multiply_cannon_sync_h2d 2664 9.7 13.632 16.039 13.632 16.039 make_m2s 222 7.7 0.008 0.011 13.047 13.442 make_images 222 8.7 0.099 0.109 13.025 13.420 qs_scf_post_transport 1 6.0 1.332 10.654 1.332 10.654 make_images_data 222 9.7 0.004 0.005 7.619 8.207 dbcsr_mm_accdrv_process 4760 10.4 0.520 0.643 6.937 7.954 hybrid_alltoall_any 227 10.6 0.215 1.856 6.548 7.724 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.447 7.643 dbcsr_mm_accdrv_process_sort 4760 11.4 6.217 7.144 6.217 7.144 calculate_norms 4752 9.8 5.518 6.210 5.518 6.210 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.042 5.178 multiply_cannon_metrocomm4 2442 9.7 0.012 0.016 2.064 3.979 mp_irecv_dv 6231 10.9 2.047 3.960 2.047 3.960 make_images_sizes 222 9.7 0.000 0.000 0.691 3.770 mp_alltoall_i44 222 10.7 0.690 3.770 0.690 3.770 arnoldi_extremal 4 6.8 0.000 0.002 3.307 3.330 arnoldi_normal_ev 4 7.8 0.001 0.007 3.306 3.330 build_subspace 16 8.4 0.009 0.013 3.200 3.206 ls_scf_store_result 1 5.0 0.000 0.000 2.870 2.913 dbcsr_special_finalize 555 9.7 0.005 0.006 2.381 2.751 dbcsr_merge_single_wm 555 10.7 0.457 0.616 2.373 2.743 make_images_pack 222 9.7 2.211 2.637 2.213 2.638 dbcsr_matrix_vector_mult 304 9.0 0.005 0.013 2.346 2.588 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=125.062000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1142.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8f18a4e0cc0e0e7c2e20248e70b07b7bca2af436_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 2.128036E+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 0.046 0.061 105.007 105.008 qs_energies 1 2.0 0.000 0.000 104.494 104.501 ls_scf 1 3.0 0.000 0.001 103.124 103.130 dbcsr_multiply_generic 111 6.7 0.015 0.016 89.088 89.442 ls_scf_main 1 4.0 0.000 0.000 68.926 68.934 density_matrix_trs4 2 5.0 0.002 0.003 63.197 63.377 multiply_cannon 111 7.7 0.028 0.037 53.128 56.403 multiply_cannon_loop 111 8.7 0.116 0.122 49.864 53.028 mp_waitall_1 9105 10.9 34.938 44.693 34.938 44.693 ls_scf_init_scf 1 4.0 0.000 0.003 30.674 30.675 make_m2s 222 7.7 0.006 0.007 29.279 29.893 make_images 222 8.7 1.576 1.927 29.249 29.863 ls_scf_init_matrix_S 1 5.0 0.000 0.002 29.445 29.566 multiply_cannon_multrec 1332 9.7 13.067 16.928 22.270 27.262 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.036 27.048 make_images_data 222 9.7 0.004 0.004 22.574 23.736 hybrid_alltoall_any 227 10.6 0.519 2.448 21.971 22.950 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.753 21.624 dbcsr_mm_accdrv_process 4041 10.4 0.274 0.455 8.805 10.373 dbcsr_mm_accdrv_process_sort 4041 11.4 8.399 9.918 8.399 9.918 mp_sum_l 807 5.4 5.365 8.671 5.365 8.671 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.233 7.893 mp_irecv_dv 3311 11.0 3.214 7.841 3.214 7.841 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.213 7.000 calculate_norms 2376 9.8 6.002 6.890 6.002 6.890 multiply_cannon_sync_h2d 1332 9.7 4.777 5.843 4.777 5.843 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.041 5.229 arnoldi_extremal 4 6.8 0.000 0.000 4.646 4.663 arnoldi_normal_ev 4 7.8 0.001 0.005 4.646 4.663 build_subspace 16 8.4 0.014 0.021 4.386 4.388 ls_scf_post 1 4.0 0.001 0.007 3.524 3.530 dbcsr_matrix_vector_mult 304 9.0 0.009 0.021 3.141 3.379 ls_scf_store_result 1 5.0 0.000 0.000 3.216 3.347 dbcsr_matrix_vector_mult_local 304 10.0 2.729 3.208 2.732 3.209 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.612 2.704 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.240 2.651 make_images_pack 222 9.7 2.028 2.396 2.030 2.398 mp_allgather_i34 111 8.7 0.964 2.331 0.964 2.331 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=105.008000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1729.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8f18a4e0cc0e0e7c2e20248e70b07b7bca2af436_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 2.699624E+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 0.034 0.046 107.921 107.922 qs_energies 1 2.0 0.000 0.000 107.275 107.278 ls_scf 1 3.0 0.000 0.000 105.861 105.864 dbcsr_multiply_generic 111 6.7 0.015 0.018 90.511 90.764 ls_scf_main 1 4.0 0.000 0.000 71.547 71.551 multiply_cannon 111 7.7 0.040 0.108 66.951 70.638 multiply_cannon_loop 111 8.7 0.099 0.104 63.313 67.166 density_matrix_trs4 2 5.0 0.002 0.003 65.484 65.620 mp_waitall_1 7281 11.0 37.982 47.993 37.982 47.993 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 18.719 36.957 ls_scf_init_scf 1 4.0 0.000 0.000 30.832 30.835 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.608 29.707 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.200 27.212 multiply_cannon_multrec 888 9.7 12.534 15.112 21.079 24.216 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 10.126 20.118 make_m2s 222 7.7 0.006 0.007 16.951 18.220 make_images 222 8.7 1.979 2.281 16.913 18.180 make_images_data 222 9.7 0.003 0.004 9.757 10.835 hybrid_alltoall_any 227 10.6 0.618 2.860 9.415 10.673 dbcsr_mm_accdrv_process 3754 10.4 0.248 0.424 8.051 9.263 dbcsr_mm_accdrv_process_sort 3754 11.4 7.678 8.840 7.678 8.840 mp_sum_l 807 5.4 5.046 8.252 5.046 8.252 multiply_cannon_sync_h2d 888 9.7 5.970 7.104 5.970 7.104 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.461 6.741 mp_irecv_dv 2335 11.1 2.445 6.682 2.445 6.682 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.875 6.402 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.964 5.149 arnoldi_extremal 4 6.8 0.000 0.000 5.062 5.080 arnoldi_normal_ev 4 7.8 0.001 0.005 5.062 5.080 build_subspace 16 8.4 0.014 0.020 4.755 4.760 calculate_norms 1584 9.8 4.316 4.704 4.316 4.704 mp_allgather_i34 111 8.7 1.469 3.850 1.469 3.850 dbcsr_matrix_vector_mult 304 9.0 0.009 0.020 3.437 3.752 dbcsr_matrix_vector_mult_local 304 10.0 3.003 3.572 3.005 3.574 ls_scf_post 1 4.0 0.000 0.000 3.481 3.484 ls_scf_store_result 1 5.0 0.000 0.000 3.246 3.302 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.909 3.037 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=107.922000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2198.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8f18a4e0cc0e0e7c2e20248e70b07b7bca2af436_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 3.373224E+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 0.039 0.047 110.992 110.993 qs_energies 1 2.0 0.000 0.000 110.289 110.294 ls_scf 1 3.0 0.000 0.000 108.636 108.640 dbcsr_multiply_generic 111 6.7 0.016 0.017 92.246 92.521 ls_scf_main 1 4.0 0.000 0.000 72.790 72.791 density_matrix_trs4 2 5.0 0.002 0.003 66.622 66.747 multiply_cannon 111 7.7 0.058 0.125 51.588 56.059 multiply_cannon_loop 111 8.7 0.115 0.127 46.476 49.147 mp_waitall_1 6369 11.0 36.821 43.358 36.821 43.358 make_m2s 222 7.7 0.006 0.008 35.116 36.545 make_images 222 8.7 3.142 3.604 35.066 36.497 ls_scf_init_scf 1 4.0 0.000 0.000 32.604 32.605 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.389 31.448 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.893 28.907 make_images_data 222 9.7 0.004 0.004 25.832 27.471 hybrid_alltoall_any 227 10.6 0.798 3.842 25.208 27.038 multiply_cannon_multrec 1332 9.7 14.084 17.584 21.931 24.699 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.240 17.095 dbcsr_mm_accdrv_process 3641 10.4 0.220 0.406 7.482 9.010 dbcsr_mm_accdrv_process_sort 3641 11.4 7.109 8.588 7.109 8.588 mp_sum_l 807 5.4 3.959 7.305 3.959 7.305 multiply_cannon_sync_h2d 1332 9.7 5.498 6.319 5.498 6.319 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.088 6.187 mp_irecv_dv 3229 10.9 2.065 6.112 2.065 6.112 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.082 5.732 arnoldi_extremal 4 6.8 0.000 0.000 5.301 5.313 arnoldi_normal_ev 4 7.8 0.002 0.006 5.301 5.313 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.577 5.295 build_subspace 16 8.4 0.014 0.021 4.955 4.961 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.550 4.713 mp_allgather_i34 111 8.7 2.226 4.642 2.226 4.642 calculate_norms 2376 9.8 4.193 4.533 4.193 4.533 dbcsr_matrix_vector_mult 304 9.0 0.010 0.020 3.630 3.951 dbcsr_matrix_vector_mult_local 304 10.0 3.175 3.674 3.177 3.676 dbcsr_sort_data 658 11.4 3.081 3.348 3.081 3.348 ls_scf_post 1 4.0 0.000 0.000 3.242 3.247 dbcsr_special_finalize 555 9.7 0.006 0.007 2.830 3.121 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.065 3.116 dbcsr_merge_single_wm 555 10.7 0.537 0.654 2.821 3.113 ls_scf_store_result 1 5.0 0.000 0.000 2.987 3.040 dbcsr_data_release 10477 10.7 1.567 2.414 1.567 2.414 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=110.993000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2730.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8f18a4e0cc0e0e7c2e20248e70b07b7bca2af436_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 4.636709E+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 0.083 0.089 112.975 112.976 qs_energies 1 2.0 0.000 0.000 112.103 112.110 ls_scf 1 3.0 0.000 0.000 110.114 110.124 dbcsr_multiply_generic 111 6.7 0.017 0.018 91.322 91.557 multiply_cannon 111 7.7 0.094 0.183 68.983 74.851 multiply_cannon_loop 111 8.7 0.069 0.076 64.368 66.494 ls_scf_main 1 4.0 0.000 0.000 61.935 61.936 density_matrix_trs4 2 5.0 0.002 0.003 54.666 54.764 mp_waitall_1 5436 11.0 39.938 46.239 39.938 46.239 ls_scf_init_scf 1 4.0 0.000 0.000 44.604 44.608 ls_scf_init_matrix_S 1 5.0 0.000 0.000 43.281 43.323 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 41.115 41.130 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 24.976 29.978 multiply_cannon_multrec 444 9.7 13.990 16.468 21.025 23.844 make_m2s 222 7.7 0.004 0.005 17.404 19.985 make_images 222 8.7 3.725 4.434 17.342 19.925 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.034 14.361 make_images_data 222 9.7 0.003 0.004 9.725 12.141 hybrid_alltoall_any 227 10.6 0.785 3.758 9.539 11.917 multiply_cannon_sync_h2d 444 9.7 6.505 8.163 6.505 8.163 dbcsr_mm_accdrv_process 3003 10.4 0.178 0.354 6.737 7.853 dbcsr_mm_accdrv_process_sort 3003 11.4 6.420 7.509 6.420 7.509 mp_allgather_i34 111 8.7 2.847 7.019 2.847 7.019 arnoldi_extremal 4 6.8 0.000 0.000 5.898 5.916 arnoldi_normal_ev 4 7.8 0.002 0.005 5.898 5.916 build_subspace 16 8.4 0.015 0.019 5.507 5.516 mp_sum_l 807 5.4 2.979 5.309 2.979 5.309 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.562 4.754 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.637 4.549 mp_irecv_dv 1241 11.2 1.617 4.519 1.617 4.519 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 4.172 4.392 dbcsr_matrix_vector_mult_local 304 10.0 3.637 4.125 3.639 4.127 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.105 4.063 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.694 3.767 calculate_norms 792 9.8 3.521 3.609 3.521 3.609 ls_scf_post 1 4.0 0.000 0.000 3.575 3.581 ls_scf_store_result 1 5.0 0.000 0.000 3.352 3.399 make_images_sizes 222 9.7 0.000 0.000 1.029 3.266 mp_alltoall_i44 222 10.7 1.028 3.266 1.028 3.266 dbcsr_finalize 304 7.8 0.062 0.077 2.203 2.269 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=112.976000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3641.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8f18a4e0cc0e0e7c2e20248e70b07b7bca2af436_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 8.763265E+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 0.086 0.105 147.679 147.680 qs_energies 1 2.0 0.000 0.000 146.181 146.197 ls_scf 1 3.0 0.000 0.000 143.236 143.251 dbcsr_multiply_generic 111 6.7 0.023 0.027 89.743 89.866 ls_scf_init_scf 1 4.0 0.000 0.000 74.749 74.750 ls_scf_main 1 4.0 0.000 0.000 64.870 64.871 multiply_cannon 111 7.7 0.128 0.254 62.502 63.964 multiply_cannon_loop 111 8.7 0.067 0.069 59.207 60.100 density_matrix_trs4 2 5.0 0.002 0.003 55.517 55.566 ls_scf_init_matrix_S 1 5.0 0.000 0.000 46.171 46.182 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 43.381 43.395 mp_waitall_1 4527 11.1 34.621 38.319 34.621 38.319 dbcsr_dot_sd 53 6.4 1.584 1.642 8.522 28.597 ls_scf_initial_guess 1 5.0 0.000 0.000 28.578 28.589 ls_scf_qs_atomic_guess 1 6.0 0.000 0.000 28.578 28.589 mp_sum_d 1055 6.0 7.126 27.112 7.126 27.112 calculate_atomic_block_dm 1 7.0 10.016 26.936 26.958 26.970 calculate_atom 2 8.0 0.000 0.000 10.099 26.943 calculate_atom_restricted 2 9.0 10.099 26.943 10.099 26.943 make_m2s 222 7.7 0.005 0.005 23.558 24.367 make_images 222 8.7 4.579 5.002 23.452 24.258 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 23.268 23.771 multiply_cannon_multrec 444 9.7 17.892 18.608 22.487 22.988 hybrid_alltoall_any 227 10.6 1.660 3.623 12.602 15.150 make_images_data 222 9.7 0.003 0.003 12.787 15.042 multiply_cannon_sync_h2d 444 9.7 8.854 8.910 8.854 8.910 arnoldi_extremal 4 6.8 0.000 0.000 7.282 7.294 arnoldi_normal_ev 4 7.8 0.003 0.009 7.282 7.294 build_subspace 16 8.4 0.026 0.037 6.726 6.740 dbcsr_matrix_vector_mult 304 9.0 0.016 0.032 5.348 5.486 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.327 5.409 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.973 5.229 dbcsr_matrix_vector_mult_local 304 10.0 4.899 5.184 4.902 5.187 dbcsr_mm_accdrv_process 1814 10.4 0.230 0.319 4.423 4.552 dbcsr_mm_accdrv_process_sort 1814 11.4 4.124 4.257 4.124 4.257 ls_scf_post 1 4.0 0.000 0.000 3.617 3.632 make_images_sizes 222 9.7 0.000 0.000 1.442 3.546 mp_alltoall_i44 222 10.7 1.441 3.546 1.441 3.546 ls_scf_store_result 1 5.0 0.000 0.000 3.363 3.380 calculate_norms 792 9.8 3.236 3.270 3.236 3.270 dbcsr_finalize 304 7.8 0.082 0.089 3.086 3.184 dbcsr_merge_all 275 8.9 0.891 0.918 2.871 2.963 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=147.680000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6910.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 8f18a4e0cc0e0e7c2e20248e70b07b7bca2af436 Summary: empty Status: OK