=== 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: 1668dc7f1dfc939e0dca246397b4db0b42a8bbc0 ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, Cray-FFTW 3.3.8.10, # COSMA 2.6.6, ELPA 2022.11.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.2.2, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.2, # SIRIUS 7.4.3, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Last update: 21.06.2023 # # \ 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 ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.2 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(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.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/1668dc7f1dfc939e0dca246397b4db0b42a8bbc0_performance_tests/01 job id: 48242403 --- 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/1668dc7f1dfc939e0dca246397b4db0b42a8bbc0_performance_tests/02 job id: 48242404 --- 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/1668dc7f1dfc939e0dca246397b4db0b42a8bbc0_performance_tests/03 job id: 48242405 --- 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/1668dc7f1dfc939e0dca246397b4db0b42a8bbc0_performance_tests/04 job id: 48242406 --- 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/1668dc7f1dfc939e0dca246397b4db0b42a8bbc0_performance_tests/05 job id: 48242407 --- 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/1668dc7f1dfc939e0dca246397b4db0b42a8bbc0_performance_tests/06 job id: 48242408 --- 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/1668dc7f1dfc939e0dca246397b4db0b42a8bbc0_performance_tests/07 job id: 48242409 --- 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/1668dc7f1dfc939e0dca246397b4db0b42a8bbc0_performance_tests/08 job id: 48242410 --- 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/1668dc7f1dfc939e0dca246397b4db0b42a8bbc0_performance_tests/09 job id: 48242411 --- 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/1668dc7f1dfc939e0dca246397b4db0b42a8bbc0_performance_tests/10 job id: 48242413 --- 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/1668dc7f1dfc939e0dca246397b4db0b42a8bbc0_performance_tests/11 job id: 48242418 --- 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/1668dc7f1dfc939e0dca246397b4db0b42a8bbc0_performance_tests/12 job id: 48242420 --- 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/1668dc7f1dfc939e0dca246397b4db0b42a8bbc0_performance_tests/13 job id: 48242421 --- 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/1668dc7f1dfc939e0dca246397b4db0b42a8bbc0_performance_tests/14 job id: 48242423 --- 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/1668dc7f1dfc939e0dca246397b4db0b42a8bbc0_performance_tests/15 job id: 48242427 --- 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/1668dc7f1dfc939e0dca246397b4db0b42a8bbc0_performance_tests/16 job id: 48242428 --- 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/1668dc7f1dfc939e0dca246397b4db0b42a8bbc0_performance_tests/17 job id: 48242429 --- 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/1668dc7f1dfc939e0dca246397b4db0b42a8bbc0_performance_tests/18 job id: 48242430 --- 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/1668dc7f1dfc939e0dca246397b4db0b42a8bbc0_performance_tests/19 job id: 48242431 --- 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/1668dc7f1dfc939e0dca246397b4db0b42a8bbc0_performance_tests/20 job id: 48242432 --- 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/1668dc7f1dfc939e0dca246397b4db0b42a8bbc0_performance_tests/21 job id: 48242433 --- 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/1668dc7f1dfc939e0dca246397b4db0b42a8bbc0_performance_tests/22 job id: 48242434 --- 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/1668dc7f1dfc939e0dca246397b4db0b42a8bbc0_performance_tests/23 job id: 48242435 --- 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/1668dc7f1dfc939e0dca246397b4db0b42a8bbc0_performance_tests/24 job id: 48242436 --- 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/1668dc7f1dfc939e0dca246397b4db0b42a8bbc0_performance_tests/25 job id: 48242437 --- 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/1668dc7f1dfc939e0dca246397b4db0b42a8bbc0_performance_tests/26 job id: 48242438 --- Point --- name: 510 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/1r/12t) --- Point --- name: 511 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: 512 H2O (4 NVE MD steps on 64 nodes) input file: benchmarks/QS/00512_H2O/H2O-512_md.inp required files: [] output file: result.log # nodes = 64 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1668dc7f1dfc939e0dca246397b4db0b42a8bbc0_performance_tests/27 job id: 48242439 --- Point --- name: 601 plot: h2o_512_md regex: CP2K label: (64n/12r/1t) --- Point --- name: 602 plot: h2o_512_md_mem regex: Estimated peak process memory label: (64n/12r/1t) ~~~~~~~ END TEST ~~~~~~~ === END TESTS (description) === ===== PLOTS (description) ===== ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2_mem", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md_mem", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md_mem", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md_mem", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls_mem", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_512_md", title="512 H2O (4 NVE MD steps on 64 nodes)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_512_md_mem", title="512 H2O (4 NVE MD steps on 64 nodes)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" === END PLOTS (description) === ============ RESULTS ============ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1668dc7f1dfc939e0dca246397b4db0b42a8bbc0_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 424 8. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.022 0.036 134.213 134.214 farming_run 1 2.0 133.651 133.652 134.184 134.189 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.490887E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 230 1134128. MP_Allreduce 571 1938539. MP_Sync 25 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 10 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.006 0.026 115.112 115.113 qs_energies 1 2.0 0.000 0.000 114.892 114.898 mp2_main 1 3.0 0.000 0.000 112.810 112.815 mp2_gpw_main 1 4.0 0.028 0.034 111.782 111.788 mp2_ri_gpw_compute_in 1 5.0 0.173 0.180 92.832 93.033 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 54.841 55.039 mp2_eri_3c_integrate_gpw 272 7.0 0.151 0.161 41.211 46.028 get_2c_integrals 1 6.0 0.008 0.009 37.287 37.810 integrate_v_rspace 273 8.0 0.438 0.454 24.615 29.061 pw_transfer 6555 10.6 0.377 0.398 27.368 27.974 fft_wrap_pw1pw2 5465 11.4 0.046 0.048 25.960 26.401 grid_integrate_task_list 273 9.0 20.445 25.307 20.445 25.307 fft_wrap_pw1pw2_100 2178 12.4 1.203 1.438 23.476 23.954 compute_2c_integrals 1 7.0 0.002 0.003 19.820 19.820 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.957 19.552 mp2_eri_2c_integrate_gpw 1 9.0 2.360 2.420 18.954 19.551 rpa_ri_compute_en 1 5.0 0.027 0.029 18.833 18.989 cp_fm_cholesky_decompose 12 8.2 17.394 17.904 17.394 17.904 cholesky_decomp 1 7.0 0.000 0.000 16.309 16.818 fft3d_s 5443 13.4 16.113 16.383 16.135 16.405 ao_to_mo_and_store_B_mult_1 272 7.0 10.777 15.310 10.777 15.310 calculate_wavefunction 272 8.0 5.428 5.516 12.501 13.172 rpa_num_int 1 6.0 0.000 0.005 10.648 10.657 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.625 10.633 calc_mat_Q 8 8.0 0.000 0.000 9.494 9.571 contract_S_to_Q 8 9.0 0.000 0.000 8.913 8.989 calc_potential_gpw 544 9.5 0.005 0.006 8.272 8.768 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.260 8.621 parallel_gemm_fm 14 9.1 0.000 0.000 8.485 8.597 parallel_gemm_fm_cosma 14 10.1 8.485 8.597 8.485 8.597 potential_pw2rs 545 10.0 0.107 0.110 7.714 8.369 create_integ_mat 1 6.0 0.013 0.027 7.883 7.893 collocate_single_gaussian 272 10.0 0.039 0.041 7.524 7.849 array2fm 1 7.0 0.000 0.000 6.737 7.341 pw_scatter_s 2720 13.7 4.387 4.567 4.387 4.567 pw_gather_s 2722 13.2 3.852 4.198 3.852 4.198 array2fm_buffer_send 1 8.0 2.907 3.149 2.907 3.149 pw_poisson_solve 545 10.5 1.105 1.150 2.220 2.422 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.787723, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2800.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1668dc7f1dfc939e0dca246397b4db0b42a8bbc0_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 424 9. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.029 0.040 421.366 421.368 farming_run 1 2.0 420.471 420.479 421.323 421.326 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827141120 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788822 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.238495E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 757 478553. MP_Allreduce 2021 21391. MP_Sync 37 MP_Alltoall 77 MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.032 217.619 217.620 qs_energies 1 2.0 0.002 0.011 217.318 217.337 scf_env_do_scf 1 3.0 0.000 0.000 115.052 115.052 qs_ks_update_qs_env 5 5.0 0.000 0.000 114.157 114.162 rebuild_ks_matrix 4 6.0 0.000 0.000 114.156 114.161 qs_ks_build_kohn_sham_matrix 4 7.0 0.055 0.060 114.156 114.161 hfx_ks_matrix 4 8.0 0.001 0.001 113.774 113.778 integrate_four_center 4 9.0 0.153 0.470 113.773 113.777 integrate_four_center_main 4 10.0 0.116 0.589 101.714 105.692 integrate_four_center_bin 262 11.0 101.597 105.664 101.597 105.664 mp2_main 1 3.0 0.001 0.016 101.966 101.979 mp2_gpw_main 1 4.0 0.047 0.068 101.112 101.131 init_scf_loop 1 4.0 0.000 0.000 96.848 96.849 mp2_ri_gpw_compute_in 1 5.0 0.064 0.064 74.479 75.440 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 53.935 54.895 mp2_eri_3c_integrate_gpw 91 7.0 0.142 0.157 41.635 46.433 integrate_v_rspace 95 8.0 0.396 0.566 28.014 32.714 pw_transfer 2240 10.6 0.144 0.191 29.962 30.403 fft_wrap_pw1pw2 1868 11.4 0.018 0.023 28.955 29.431 ao_to_mo_and_store_B_mult_1 91 7.0 10.614 28.553 10.614 28.553 mp2_ri_gpw_compute_en 1 5.0 0.058 0.072 26.475 28.530 grid_integrate_task_list 95 9.0 23.319 28.148 23.319 28.148 fft_wrap_pw1pw2_100 730 12.4 1.304 1.513 26.668 27.183 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.838 1.911 24.863 24.873 get_2c_integrals 1 6.0 0.001 0.008 20.447 20.479 compute_2c_integrals 1 7.0 0.002 0.003 19.431 19.433 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 18.945 19.310 mp2_eri_2c_integrate_gpw 1 9.0 1.725 1.897 18.944 19.310 fft3d_s 1823 13.4 18.454 18.932 18.467 18.945 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 18.202 18.202 calculate_wavefunction 91 8.0 2.016 2.042 9.740 9.961 potential_pw2rs 186 10.0 0.033 0.036 8.645 9.257 mp2_ri_gpw_compute_en_expansio 172 7.0 0.555 0.575 8.741 9.170 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.298 8.628 local_gemm 172 8.0 8.186 8.618 8.186 8.618 mp2_ri_gpw_compute_en_comm 22 7.0 0.500 0.518 7.892 8.390 collocate_single_gaussian 91 10.0 0.017 0.022 7.923 8.183 calc_potential_gpw 182 9.5 0.002 0.003 7.946 8.182 mp_sync 37 10.5 4.027 7.449 4.027 7.449 integrate_four_center_load 4 10.0 0.000 0.000 6.774 6.778 hfx_load_balance 1 11.0 0.000 0.000 6.774 6.778 mp_sendrecv_dm3 2068 8.0 5.921 6.411 5.921 6.411 mp2_ri_gpw_compute_en_ener 172 7.0 6.346 6.402 6.346 6.402 pw_gather_s 912 13.2 4.879 5.408 4.879 5.408 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.112633, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1498.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1668dc7f1dfc939e0dca246397b4db0b42a8bbc0_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 451.969024E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62379. MP_Allreduce 10329 270. MP_Sync 530 MP_Alltoall 2083 588782. 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 0.012 0.030 53.189 53.190 qs_mol_dyn_low 1 2.0 0.003 0.003 52.918 52.935 qs_forces 11 3.9 0.002 0.002 52.816 52.817 qs_energies 11 4.9 0.001 0.002 51.305 51.318 scf_env_do_scf 11 5.9 0.001 0.002 45.123 45.123 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 43.097 43.097 dbcsr_multiply_generic 2286 12.5 0.095 0.099 32.941 33.410 qs_scf_new_mos 108 7.5 0.000 0.001 32.947 33.243 qs_scf_loop_do_ot 108 8.5 0.000 0.001 32.947 33.243 ot_scf_mini 108 9.5 0.002 0.002 31.288 31.456 multiply_cannon 2286 13.5 0.178 0.188 26.067 27.722 multiply_cannon_loop 2286 14.5 1.809 1.915 25.409 27.074 velocity_verlet 10 3.0 0.001 0.002 25.920 25.921 ot_mini 108 10.5 0.001 0.001 18.879 19.119 qs_ot_get_derivative 108 11.5 0.001 0.001 15.959 16.122 mp_waitall_1 245248 16.5 8.007 14.241 8.007 14.241 multiply_cannon_metrocomm3 54864 15.5 0.072 0.079 5.811 12.562 multiply_cannon_multrec 54864 15.5 3.689 5.891 7.872 11.336 qs_ot_get_p 119 10.4 0.001 0.001 7.830 8.121 rebuild_ks_matrix 119 8.3 0.000 0.000 7.855 7.996 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 7.854 7.996 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.901 7.032 mp_sum_l 7287 12.8 4.895 6.477 4.895 6.477 multiply_cannon_sync_h2d 54864 15.5 5.201 6.216 5.201 6.216 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.527 6.004 dbcsr_mm_accdrv_process 76910 16.1 1.799 2.831 4.098 5.860 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.165 5.197 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.018 5.119 init_scf_run 11 5.9 0.000 0.001 4.914 4.914 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.914 4.914 sum_up_and_integrate 119 10.3 0.012 0.014 4.482 4.488 integrate_v_rspace 119 11.3 0.002 0.003 4.470 4.478 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.416 4.417 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.272 4.401 calculate_rho_elec 119 8.7 0.012 0.017 4.272 4.401 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.210 4.211 cp_fm_redistribute_end 50 14.0 2.146 4.183 2.154 4.187 cp_fm_diag_elpa_base 50 14.0 2.026 4.078 2.031 4.087 calculate_dm_sparse 119 9.5 0.000 0.001 2.991 3.133 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.895 3.109 apply_single 119 13.6 0.000 0.000 2.895 3.109 jit_kernel_multiply 13 15.8 2.236 2.949 2.236 2.949 acc_transpose_blocks 54864 15.5 0.230 0.256 2.237 2.803 rs_pw_transfer 974 11.9 0.012 0.013 2.703 2.797 calculate_first_density_matrix 1 7.0 0.000 0.000 2.694 2.695 ot_diis_step 108 11.5 0.006 0.006 2.648 2.648 multiply_cannon_metrocomm1 54864 15.5 0.056 0.065 1.555 2.479 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.379 2.441 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.390 2.393 density_rs2pw 119 9.7 0.004 0.004 2.200 2.322 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.096 2.159 wfi_extrapolate 11 7.9 0.001 0.001 2.152 2.152 grid_integrate_task_list 119 12.3 2.025 2.125 2.025 2.125 init_scf_loop 11 6.9 0.000 0.000 2.009 2.009 mp_sum_d 4135 12.0 1.232 1.878 1.232 1.878 potential_pw2rs 119 12.3 0.004 0.004 1.832 1.846 pw_transfer 1439 11.6 0.052 0.058 1.721 1.790 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 1.645 1.716 make_m2s 4572 13.5 0.053 0.056 1.658 1.698 make_images 4572 14.5 0.133 0.138 1.576 1.618 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.001 1.491 1.516 mp_alltoall_d11v 2130 13.8 1.290 1.438 1.290 1.438 grid_collocate_task_list 119 9.7 1.354 1.417 1.354 1.417 acc_transpose_blocks_sync 164592 16.5 1.177 1.417 1.177 1.417 mp_waitany 12084 13.8 1.235 1.405 1.235 1.405 fft3d_ps 1201 14.6 0.374 0.481 1.287 1.352 fft_wrap_pw1pw2_140 487 13.2 0.186 0.203 1.271 1.344 dbcsr_dot_sd 1205 11.9 0.049 0.060 0.682 1.097 acc_transpose_blocks_kernels 54864 16.5 0.244 0.376 0.785 1.091 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=53.190000, 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/1668dc7f1dfc939e0dca246397b4db0b42a8bbc0_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 488.333312E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2194560 MPI messages size (bytes): total size 310.646604E+09 min size 0.000000E+00 max size 1.145520E+06 average size 141.553031E+03 MPI breakdown and total messages size (bytes): size <= 128 724648 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 281952 4619501568 32768 < size <= 131072 494448 39143342080 131072 < size <= 4194304 440000 264807943488 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62658. MP_Allreduce 10306 303. MP_Sync 54 MP_Alltoall 2060 1021473. MP_SendRecv 16779 37093. MP_ISendRecv 16779 37093. MP_Wait 23539 MP_comm_split 50 MP_ISend 5720 128509. MP_IRecv 5720 128509. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.032 39.388 39.390 qs_mol_dyn_low 1 2.0 0.003 0.004 39.193 39.202 qs_forces 11 3.9 0.002 0.003 38.853 38.861 qs_energies 11 4.9 0.002 0.002 37.137 37.149 scf_env_do_scf 11 5.9 0.001 0.002 31.961 31.962 scf_env_do_scf_inner_loop 108 6.5 0.002 0.008 29.528 29.528 dbcsr_multiply_generic 2286 12.5 0.111 0.122 22.038 22.368 qs_scf_new_mos 108 7.5 0.001 0.001 20.413 20.641 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.412 20.641 ot_scf_mini 108 9.5 0.002 0.003 19.491 19.660 velocity_verlet 10 3.0 0.001 0.002 18.838 18.845 multiply_cannon 2286 13.5 0.207 0.216 16.994 18.455 multiply_cannon_loop 2286 14.5 1.188 1.240 15.851 17.315 ot_mini 108 10.5 0.001 0.001 12.044 12.275 mp_waitall_1 200699 16.5 5.562 10.762 5.562 10.762 qs_ot_get_derivative 108 11.5 0.001 0.001 9.697 9.867 multiply_cannon_metrocomm3 27432 15.5 0.072 0.075 4.105 9.567 multiply_cannon_multrec 27432 15.5 1.811 4.124 6.109 8.973 rebuild_ks_matrix 119 8.3 0.000 0.000 7.196 7.341 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.016 7.195 7.340 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.331 6.464 dbcsr_mm_accdrv_process 47894 16.0 3.520 5.601 4.220 6.294 qs_ot_get_p 119 10.4 0.001 0.002 4.539 4.760 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.691 4.527 sum_up_and_integrate 119 10.3 0.025 0.027 4.103 4.109 integrate_v_rspace 119 11.3 0.002 0.003 4.079 4.087 mp_sum_l 7287 12.8 1.973 4.036 1.973 4.036 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.982 4.028 apply_single 119 13.6 0.000 0.000 2.982 4.028 init_scf_run 11 5.9 0.000 0.001 3.986 3.986 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.985 3.986 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.785 3.815 calculate_rho_elec 119 8.7 0.021 0.024 3.785 3.814 rs_pw_transfer 974 11.9 0.010 0.011 2.478 3.016 make_m2s 4572 13.5 0.052 0.054 2.605 2.953 qs_ot_p2m_diag 50 11.0 0.008 0.012 2.926 2.943 make_images 4572 14.5 0.206 0.246 2.515 2.866 density_rs2pw 119 9.7 0.004 0.004 2.051 2.603 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.532 2.533 calculate_first_density_matrix 1 7.0 0.000 0.000 2.461 2.464 init_scf_loop 11 6.9 0.000 0.000 2.413 2.414 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.294 2.385 ot_diis_step 108 11.5 0.011 0.011 2.300 2.300 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.223 2.223 cp_fm_redistribute_end 50 14.0 1.127 2.199 1.130 2.202 multiply_cannon_sync_h2d 27432 15.5 1.706 2.180 1.706 2.180 cp_fm_diag_elpa_base 50 14.0 1.041 2.111 1.068 2.149 calculate_dm_sparse 119 9.5 0.000 0.001 2.065 2.143 jit_kernel_multiply 11 16.4 0.642 2.124 0.642 2.124 acc_transpose_blocks 27432 15.5 0.116 0.123 1.708 2.091 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.025 2.027 pw_transfer 1439 11.6 0.065 0.070 1.897 1.926 grid_integrate_task_list 119 12.3 1.836 1.912 1.836 1.912 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.831 1.870 potential_pw2rs 119 12.3 0.006 0.006 1.831 1.840 fft_wrap_pw1pw2 1201 12.6 0.008 0.009 1.805 1.836 make_images_data 4572 15.5 0.048 0.054 1.231 1.580 prepare_preconditioner 11 7.9 0.000 0.000 1.496 1.522 make_preconditioner 11 8.9 0.000 0.000 1.496 1.522 fft_wrap_pw1pw2_140 487 13.2 0.204 0.215 1.440 1.472 wfi_extrapolate 11 7.9 0.001 0.001 1.465 1.465 hybrid_alltoall_any 4725 16.4 0.054 0.116 1.086 1.453 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.389 1.446 grid_collocate_task_list 119 9.7 1.277 1.429 1.277 1.429 fft3d_ps 1201 14.6 0.522 0.578 1.354 1.379 mp_alltoall_d11v 2130 13.8 1.161 1.285 1.161 1.285 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.251 1.259 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.194 1.244 rs_pw_transfer_RS2PW_140 130 11.5 0.139 0.147 0.536 1.075 mp_waitany 5720 13.7 0.516 1.052 0.516 1.052 acc_transpose_blocks_kernels 27432 16.5 0.188 0.277 0.756 1.007 mp_sum_d 4135 12.0 0.525 1.003 0.525 1.003 mp_allgather_i34 2286 14.5 0.567 0.968 0.567 0.968 qs_energies_init_hamiltonians 11 5.9 0.000 0.002 0.931 0.939 acc_transpose_blocks_sync 82296 16.5 0.810 0.936 0.810 0.936 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.920 0.933 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=39.390000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=464.454545, yerr=1.671343 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1668dc7f1dfc939e0dca246397b4db0b42a8bbc0_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 519.516160E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63490. MP_Allreduce 10155 305. MP_Sync 54 MP_Alltoall 1821 1607820. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.030 32.542 32.543 qs_mol_dyn_low 1 2.0 0.003 0.003 32.317 32.324 qs_forces 11 3.9 0.002 0.002 32.256 32.259 qs_energies 11 4.9 0.001 0.002 30.660 30.664 scf_env_do_scf 11 5.9 0.000 0.001 25.873 25.873 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 23.436 23.437 dbcsr_multiply_generic 2286 12.5 0.097 0.098 16.965 17.055 velocity_verlet 10 3.0 0.002 0.006 15.430 15.431 qs_scf_new_mos 108 7.5 0.001 0.001 15.305 15.320 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.305 15.319 ot_scf_mini 108 9.5 0.002 0.002 14.550 14.568 multiply_cannon 2286 13.5 0.193 0.199 13.719 14.502 multiply_cannon_loop 2286 14.5 0.851 0.884 12.947 13.754 ot_mini 108 10.5 0.001 0.001 9.026 9.050 qs_ot_get_derivative 108 11.5 0.001 0.001 7.523 7.541 multiply_cannon_multrec 18288 15.5 1.883 2.907 7.103 7.531 rebuild_ks_matrix 119 8.3 0.000 0.000 6.350 6.368 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.349 6.368 dbcsr_mm_accdrv_process 38222 16.0 5.068 5.890 5.127 5.956 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.592 5.609 sum_up_and_integrate 119 10.3 0.030 0.031 3.927 3.931 integrate_v_rspace 119 11.3 0.002 0.003 3.897 3.904 mp_waitall_1 158411 16.6 2.532 3.689 2.532 3.689 init_scf_run 11 5.9 0.000 0.001 3.597 3.597 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.597 3.597 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.433 3.437 calculate_rho_elec 119 8.7 0.031 0.031 3.432 3.436 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.795 3.413 qs_ot_get_p 119 10.4 0.001 0.002 3.293 3.312 rs_pw_transfer 974 11.9 0.009 0.010 2.211 2.473 init_scf_loop 11 6.9 0.000 0.000 2.419 2.420 multiply_cannon_metrocomm3 18288 15.5 0.047 0.049 1.446 2.413 calculate_first_density_matrix 1 7.0 0.000 0.000 2.370 2.371 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.036 2.352 apply_single 119 13.6 0.000 0.000 2.036 2.352 qs_ot_p2m_diag 50 11.0 0.012 0.012 2.198 2.203 density_rs2pw 119 9.7 0.004 0.004 1.923 2.177 make_m2s 4572 13.5 0.044 0.045 1.795 1.946 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.906 1.907 pw_transfer 1439 11.6 0.065 0.069 1.888 1.897 grid_integrate_task_list 119 12.3 1.804 1.885 1.804 1.885 calculate_dm_sparse 119 9.5 0.000 0.000 1.855 1.864 make_images 4572 14.5 0.193 0.206 1.711 1.861 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.807 1.809 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.795 1.807 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.696 1.707 potential_pw2rs 119 12.3 0.007 0.008 1.684 1.689 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.664 1.676 cp_fm_diag_elpa_base 50 14.0 1.641 1.655 1.663 1.674 acc_transpose_blocks 18288 15.5 0.079 0.081 1.607 1.670 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.657 1.664 prepare_preconditioner 11 7.9 0.000 0.000 1.641 1.645 make_preconditioner 11 8.9 0.000 0.000 1.641 1.645 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.505 1.583 ot_diis_step 108 11.5 0.011 0.012 1.486 1.486 mp_sum_l 7287 12.8 1.054 1.472 1.054 1.472 fft_wrap_pw1pw2_140 487 13.2 0.256 0.264 1.448 1.458 grid_collocate_task_list 119 9.7 1.245 1.396 1.245 1.396 fft3d_ps 1201 14.6 0.531 0.548 1.279 1.293 multiply_cannon_sync_h2d 18288 15.5 1.047 1.239 1.047 1.239 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.196 1.200 wfi_extrapolate 11 7.9 0.001 0.001 1.174 1.174 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 0.964 0.966 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.941 0.957 make_images_data 4572 15.5 0.047 0.051 0.766 0.922 hybrid_alltoall_any 4725 16.4 0.058 0.116 0.667 0.851 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.827 0.829 acc_transpose_blocks_kernels 18288 16.5 0.217 0.225 0.782 0.798 mp_waitany 9880 13.7 0.522 0.780 0.522 0.780 acc_transpose_blocks_sync 54864 16.5 0.725 0.777 0.725 0.777 rs_pw_transfer_RS2PW_140 130 11.5 0.117 0.121 0.517 0.769 mp_alltoall_d11v 2130 13.8 0.643 0.762 0.643 0.762 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.675 0.740 cp_fm_cholesky_invert 11 10.9 0.705 0.708 0.705 0.708 mp_alltoall_z22v 1201 16.6 0.606 0.675 0.606 0.675 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=32.543000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=494.909091, yerr=0.995859 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1668dc7f1dfc939e0dca246397b4db0b42a8bbc0_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 555.577344E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63489. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.036 36.868 36.869 qs_mol_dyn_low 1 2.0 0.003 0.004 36.533 36.555 qs_forces 11 3.9 0.004 0.013 36.446 36.452 qs_energies 11 4.9 0.001 0.001 34.707 34.718 scf_env_do_scf 11 5.9 0.000 0.001 29.643 29.644 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 26.242 26.243 dbcsr_multiply_generic 2286 12.5 0.101 0.105 19.799 19.890 velocity_verlet 10 3.0 0.002 0.003 18.624 18.627 qs_scf_new_mos 108 7.5 0.001 0.001 17.750 17.811 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.749 17.810 ot_scf_mini 108 9.5 0.002 0.003 16.726 16.784 multiply_cannon 2286 13.5 0.216 0.223 16.121 16.499 multiply_cannon_loop 2286 14.5 1.527 1.608 15.167 15.482 ot_mini 108 10.5 0.001 0.001 10.275 10.338 multiply_cannon_multrec 27432 15.5 2.486 3.183 9.031 9.338 qs_ot_get_derivative 108 11.5 0.001 0.001 8.409 8.467 dbcsr_mm_accdrv_process 47916 15.9 6.042 7.556 6.444 7.792 rebuild_ks_matrix 119 8.3 0.000 0.000 6.621 6.673 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.620 6.672 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.867 5.911 sum_up_and_integrate 119 10.3 0.035 0.037 3.791 3.800 integrate_v_rspace 119 11.3 0.003 0.003 3.756 3.765 init_scf_run 11 5.9 0.000 0.001 3.722 3.723 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.722 3.723 qs_ot_get_p 119 10.4 0.002 0.003 3.626 3.704 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.536 3.568 calculate_rho_elec 119 8.7 0.040 0.046 3.536 3.567 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.103 3.495 init_scf_loop 11 6.9 0.000 0.000 3.381 3.381 acc_transpose_blocks 27432 15.5 0.119 0.122 2.391 2.501 prepare_preconditioner 11 7.9 0.000 0.000 2.483 2.491 make_preconditioner 11 8.9 0.000 0.000 2.483 2.491 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.097 2.420 make_m2s 4572 13.5 0.054 0.056 2.299 2.413 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.312 2.343 calculate_first_density_matrix 1 7.0 0.000 0.000 2.322 2.325 make_images 4572 14.5 0.271 0.329 2.192 2.304 mp_waitall_1 137007 16.6 1.743 2.277 1.743 2.277 qs_ot_p2m_diag 50 11.0 0.015 0.022 2.209 2.217 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.090 2.210 apply_single 119 13.6 0.000 0.000 2.090 2.210 calculate_dm_sparse 119 9.5 0.000 0.001 2.130 2.177 rs_pw_transfer 974 11.9 0.009 0.009 1.958 2.146 density_rs2pw 119 9.7 0.004 0.004 1.929 2.125 pw_transfer 1439 11.6 0.065 0.070 2.003 2.037 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.910 1.949 grid_integrate_task_list 119 12.3 1.825 1.919 1.825 1.919 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.836 1.837 ot_diis_step 108 11.5 0.012 0.013 1.825 1.826 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.795 1.795 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.654 1.667 fft_wrap_pw1pw2_140 487 13.2 0.291 0.305 1.575 1.614 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.546 1.556 cp_fm_diag_elpa_base 50 14.0 1.512 1.529 1.544 1.554 potential_pw2rs 119 12.3 0.008 0.009 1.535 1.539 acc_transpose_blocks_sync 82296 16.5 1.417 1.510 1.417 1.510 grid_collocate_task_list 119 9.7 1.248 1.373 1.248 1.373 fft3d_ps 1201 14.6 0.560 0.608 1.344 1.371 wfi_extrapolate 11 7.9 0.001 0.001 1.346 1.346 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.279 1.294 multiply_cannon_metrocomm3 27432 15.5 0.041 0.043 0.763 1.271 mp_sum_l 7287 12.8 0.947 1.252 0.947 1.252 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.234 1.246 cp_fm_upper_to_full 72 14.2 0.809 1.153 0.809 1.153 jit_kernel_multiply 6 16.1 0.336 1.090 0.336 1.090 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.080 1.087 dbcsr_complete_redistribute 329 12.2 0.128 0.158 0.771 1.049 make_images_data 4572 15.5 0.048 0.052 0.862 0.992 hybrid_alltoall_any 4725 16.4 0.066 0.157 0.716 0.917 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.801 0.878 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.843 0.849 acc_transpose_blocks_kernels 27432 16.5 0.271 0.279 0.826 0.845 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.571 0.837 mp_alltoall_d11v 2130 13.8 0.697 0.835 0.697 0.835 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=36.869000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=527.636364, yerr=2.739367 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1668dc7f1dfc939e0dca246397b4db0b42a8bbc0_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 616.644608E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63488. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.030 29.474 29.475 qs_mol_dyn_low 1 2.0 0.003 0.004 29.202 29.219 qs_forces 11 3.9 0.002 0.003 29.062 29.066 qs_energies 11 4.9 0.004 0.004 27.315 27.321 scf_env_do_scf 11 5.9 0.000 0.001 22.336 22.336 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 19.851 19.853 velocity_verlet 10 3.0 0.014 0.031 15.088 15.096 dbcsr_multiply_generic 2286 12.5 0.093 0.097 13.227 13.296 qs_scf_new_mos 108 7.5 0.001 0.001 11.805 11.829 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.804 11.828 ot_scf_mini 108 9.5 0.002 0.002 11.096 11.120 multiply_cannon 2286 13.5 0.236 0.247 10.482 11.021 multiply_cannon_loop 2286 14.5 0.642 0.670 9.535 9.685 ot_mini 108 10.5 0.001 0.001 6.325 6.354 multiply_cannon_multrec 9144 15.5 1.683 1.892 5.992 6.217 rebuild_ks_matrix 119 8.3 0.000 0.000 6.054 6.074 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.054 6.073 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.380 5.398 qs_ot_get_derivative 108 11.5 0.001 0.001 5.020 5.044 dbcsr_mm_accdrv_process 12550 15.8 3.543 4.224 4.200 4.270 sum_up_and_integrate 119 10.3 0.037 0.041 3.693 3.705 integrate_v_rspace 119 11.3 0.003 0.003 3.656 3.668 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.559 3.565 calculate_rho_elec 119 8.7 0.060 0.061 3.558 3.565 init_scf_run 11 5.9 0.000 0.001 3.529 3.529 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.529 3.529 qs_ot_get_p 119 10.4 0.001 0.001 2.866 2.904 init_scf_loop 11 6.9 0.000 0.000 2.464 2.466 calculate_first_density_matrix 1 7.0 0.000 0.000 2.361 2.362 make_m2s 4572 13.5 0.034 0.035 1.941 2.091 pw_transfer 1439 11.6 0.066 0.069 2.052 2.063 make_images 4572 14.5 0.270 0.301 1.852 2.000 density_rs2pw 119 9.7 0.004 0.004 1.850 1.997 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.958 1.969 grid_integrate_task_list 119 12.3 1.853 1.950 1.853 1.950 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.878 1.880 mp_waitall_1 115863 16.7 1.415 1.865 1.415 1.865 rs_pw_transfer 974 11.9 0.008 0.009 1.707 1.845 calculate_dm_sparse 119 9.5 0.000 0.000 1.779 1.798 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.720 1.739 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.717 1.718 prepare_preconditioner 11 7.9 0.000 0.000 1.704 1.707 make_preconditioner 11 8.9 0.000 0.000 1.704 1.707 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.636 1.636 fft_wrap_pw1pw2_140 487 13.2 0.367 0.376 1.610 1.620 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.590 1.613 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.510 1.520 acc_transpose_blocks 9144 15.5 0.043 0.046 1.448 1.487 potential_pw2rs 119 12.3 0.010 0.011 1.413 1.416 grid_collocate_task_list 119 9.7 1.296 1.412 1.296 1.412 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.389 1.401 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.338 1.345 cp_fm_diag_elpa_base 50 14.0 1.311 1.327 1.336 1.344 fft3d_ps 1201 14.6 0.561 0.571 1.286 1.297 ot_diis_step 108 11.5 0.013 0.013 1.292 1.292 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.236 1.271 apply_single 119 13.6 0.000 0.000 1.236 1.270 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.233 1.238 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.168 1.174 wfi_extrapolate 11 7.9 0.001 0.001 1.114 1.114 hybrid_alltoall_any 4725 16.4 0.066 0.177 0.828 1.054 jit_kernel_multiply 6 15.6 0.617 1.054 0.617 1.054 make_images_data 4572 15.5 0.042 0.046 0.828 1.005 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.875 0.932 mp_alltoall_d11v 2130 13.8 0.816 0.930 0.816 0.930 cp_fm_cholesky_invert 11 10.9 0.857 0.860 0.857 0.860 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.827 0.833 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.795 0.798 acc_transpose_blocks_sync 27432 16.5 0.716 0.757 0.716 0.757 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.677 0.732 mp_allgather_i34 2286 14.5 0.250 0.723 0.250 0.723 multiply_cannon_metrocomm3 9144 15.5 0.021 0.021 0.333 0.705 acc_transpose_blocks_kernels 9144 16.5 0.117 0.120 0.673 0.686 mp_alltoall_z22v 1201 16.6 0.600 0.626 0.600 0.626 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=29.475000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=584.545455, yerr=6.080724 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1668dc7f1dfc939e0dca246397b4db0b42a8bbc0_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 774.164480E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63723. MP_Allreduce 10154 429. MP_Sync 54 MP_Alltoall 1582 7383731. MP_SendRecv 2499 189067. MP_ISendRecv 2499 189067. MP_Wait 6399 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.024 0.080 43.714 43.716 qs_mol_dyn_low 1 2.0 0.004 0.016 43.415 43.425 qs_forces 11 3.9 0.015 0.017 43.333 43.336 qs_energies 11 4.9 0.001 0.001 41.290 41.294 scf_env_do_scf 11 5.9 0.001 0.001 35.347 35.347 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 27.428 27.429 velocity_verlet 10 3.0 0.002 0.004 24.407 24.414 dbcsr_multiply_generic 2286 12.5 0.101 0.105 19.358 19.571 qs_scf_new_mos 108 7.5 0.001 0.001 17.532 17.631 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.531 17.630 ot_scf_mini 108 9.5 0.002 0.002 16.371 16.475 multiply_cannon 2286 13.5 0.303 0.316 15.274 16.150 multiply_cannon_loop 2286 14.5 0.856 0.877 13.845 14.746 ot_mini 108 10.5 0.001 0.001 9.934 10.052 multiply_cannon_multrec 9144 15.5 3.344 4.612 8.801 8.853 qs_ot_get_derivative 108 11.5 0.001 0.001 7.849 7.952 init_scf_loop 11 6.9 0.000 0.000 7.887 7.888 rebuild_ks_matrix 119 8.3 0.000 0.000 7.385 7.531 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.385 7.530 prepare_preconditioner 11 7.9 0.000 0.000 6.886 6.899 make_preconditioner 11 8.9 0.000 0.000 6.886 6.899 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.665 6.796 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.475 6.780 dbcsr_mm_accdrv_process 12550 15.8 4.484 5.803 5.321 6.531 cp_fm_upper_to_full 72 14.2 3.149 4.506 3.149 4.506 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.332 4.339 calculate_rho_elec 119 8.7 0.118 0.121 4.331 4.338 sum_up_and_integrate 119 10.3 0.064 0.065 4.041 4.047 integrate_v_rspace 119 11.3 0.003 0.004 3.976 3.982 init_scf_run 11 5.9 0.000 0.001 3.884 3.884 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.884 3.884 qs_ot_get_p 119 10.4 0.001 0.001 3.476 3.613 mp_waitall_1 94719 16.7 2.420 3.457 2.420 3.457 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.848 3.247 dbcsr_complete_redistribute 329 12.2 0.299 0.305 1.990 2.780 pw_transfer 1439 11.6 0.069 0.070 2.687 2.693 make_m2s 4572 13.5 0.037 0.037 2.465 2.636 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.589 2.594 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.232 2.544 apply_single 119 13.6 0.000 0.000 2.231 2.543 make_images 4572 14.5 0.352 0.385 2.344 2.515 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.661 2.442 calculate_first_density_matrix 1 7.0 0.000 0.000 2.389 2.390 calculate_dm_sparse 119 9.5 0.000 0.000 2.285 2.301 multiply_cannon_metrocomm3 9144 15.5 0.021 0.022 1.377 2.272 density_rs2pw 119 9.7 0.004 0.004 2.218 2.241 mp_alltoall_i22 627 13.8 1.396 2.200 1.396 2.200 fft_wrap_pw1pw2_140 487 13.2 0.619 0.624 2.177 2.184 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.405 2.179 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.078 2.130 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.106 2.109 grid_integrate_task_list 119 12.3 2.071 2.088 2.071 2.088 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.070 2.071 ot_diis_step 108 11.5 0.014 0.015 2.059 2.059 acc_transpose_blocks 9144 15.5 0.045 0.045 1.820 1.883 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.788 1.815 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.771 1.774 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.750 1.751 mp_sum_l 7287 12.8 1.068 1.750 1.068 1.750 fft3d_ps 1201 14.6 0.594 0.609 1.603 1.608 rs_pw_transfer 974 11.9 0.009 0.009 1.517 1.551 grid_collocate_task_list 119 9.7 1.517 1.542 1.517 1.542 cp_fm_cholesky_invert 11 10.9 1.519 1.522 1.519 1.522 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.451 1.451 cp_fm_diag_elpa_base 50 14.0 1.305 1.357 1.450 1.450 wfi_extrapolate 11 7.9 0.001 0.001 1.429 1.429 potential_pw2rs 119 12.3 0.014 0.014 1.411 1.413 hybrid_alltoall_any 4725 16.4 0.090 0.150 1.110 1.387 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.320 1.341 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.291 1.297 make_images_data 4572 15.5 0.046 0.049 1.051 1.275 mp_alltoall_d11v 2130 13.8 1.125 1.145 1.125 1.145 acc_transpose_blocks_sync 27432 16.5 1.089 1.119 1.089 1.119 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.099 1.114 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.002 1.046 jit_kernel_multiply 6 15.5 0.808 1.018 0.808 1.018 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.932 0.946 qs_create_task_list 11 7.9 0.000 0.001 0.933 0.944 generate_qs_task_list 11 8.9 0.369 0.388 0.933 0.943 mp_alltoall_z22v 1201 16.6 0.873 0.891 0.873 0.891 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=43.716000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=727.363636, yerr=16.154215 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1668dc7f1dfc939e0dca246397b4db0b42a8bbc0_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 501.346304E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66212. MP_Allreduce 9776 488. MP_Sync 52 MP_Alltoall 1938 2365395. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.032 115.063 115.064 qs_mol_dyn_low 1 2.0 0.003 0.009 114.786 114.796 qs_forces 11 3.9 0.011 0.030 114.701 114.705 qs_energies 11 4.9 0.001 0.001 111.516 111.546 scf_env_do_scf 11 5.9 0.000 0.001 101.847 101.849 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 94.766 94.766 qs_scf_new_mos 99 7.5 0.000 0.001 69.250 69.460 qs_scf_loop_do_ot 99 8.5 0.000 0.001 69.250 69.460 dbcsr_multiply_generic 2055 12.4 0.114 0.120 66.582 67.091 ot_scf_mini 99 9.5 0.002 0.002 65.521 65.791 multiply_cannon 2055 13.4 0.175 0.181 49.524 51.265 multiply_cannon_loop 2055 14.4 1.781 1.824 47.556 49.534 velocity_verlet 10 3.0 0.002 0.005 48.612 48.613 ot_mini 99 10.5 0.001 0.001 36.563 36.861 qs_ot_get_derivative 99 11.5 0.001 0.001 27.853 28.075 mp_waitall_1 220248 16.4 19.710 21.434 19.710 21.434 rebuild_ks_matrix 110 8.3 0.000 0.000 19.886 20.055 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.011 19.885 20.055 qs_ks_update_qs_env 110 7.6 0.001 0.001 17.831 17.992 qs_ot_get_p 110 10.4 0.001 0.002 17.682 17.934 multiply_cannon_multrec 49320 15.4 10.838 11.792 16.819 17.535 qs_ot_p2m_diag 48 11.0 0.012 0.018 13.923 14.047 cp_dbcsr_syevd 48 12.0 0.003 0.003 11.896 11.896 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 11.132 11.328 sum_up_and_integrate 110 10.3 0.036 0.042 11.048 11.063 integrate_v_rspace 110 11.3 0.003 0.004 11.012 11.034 cp_fm_diag_elpa 48 13.0 0.000 0.000 10.915 10.946 cp_fm_diag_elpa_base 48 14.0 10.902 10.934 10.913 10.944 multiply_cannon_metrocomm1 49320 15.4 0.066 0.069 8.552 10.770 qs_rho_update_rho_low 110 7.6 0.001 0.001 9.951 10.081 calculate_rho_elec 110 8.6 0.021 0.026 9.951 10.080 mp_sum_l 6594 12.7 8.153 9.904 8.153 9.904 multiply_cannon_sync_h2d 49320 15.4 9.027 9.647 9.027 9.647 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 8.710 9.247 apply_single 110 13.6 0.000 0.000 8.709 9.247 multiply_cannon_metrocomm3 49320 15.4 0.081 0.086 7.512 8.996 make_m2s 4110 13.4 0.061 0.065 8.446 8.618 make_images 4110 14.4 0.177 0.193 8.349 8.521 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.866 8.429 ot_diis_step 99 11.5 0.006 0.006 8.205 8.224 init_scf_run 11 5.9 0.000 0.001 7.373 7.374 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.373 7.374 init_scf_loop 11 6.9 0.000 0.000 7.032 7.032 rs_pw_transfer 902 11.9 0.011 0.013 6.076 6.636 density_rs2pw 110 9.6 0.004 0.005 5.864 6.432 dbcsr_mm_accdrv_process 87628 16.1 3.003 3.108 5.856 6.143 pw_transfer 1331 11.6 0.056 0.064 6.033 6.132 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 5.944 6.045 calculate_dm_sparse 110 9.5 0.001 0.001 5.126 5.360 fft_wrap_pw1pw2_140 451 13.1 0.452 0.500 5.222 5.327 mp_alltoall_d11v 2046 13.8 4.876 5.232 4.876 5.232 fft3d_ps 1111 14.6 0.796 0.884 5.138 5.210 qs_ot_get_orbitals 99 10.5 0.000 0.001 5.021 5.178 potential_pw2rs 110 12.3 0.006 0.007 5.067 5.102 make_images_sizes 4110 15.4 0.004 0.005 3.701 4.704 mp_alltoall_i44 4110 16.4 3.697 4.700 3.697 4.700 prepare_preconditioner 11 7.9 0.000 0.000 4.644 4.669 make_preconditioner 11 8.9 0.000 0.000 4.644 4.669 wfi_extrapolate 11 7.9 0.001 0.001 4.648 4.648 make_full_inverse_cholesky 11 9.9 0.000 0.000 4.379 4.440 mp_alltoall_z22v 1111 16.6 3.967 4.131 3.967 4.131 make_images_data 4110 15.4 0.044 0.048 3.140 3.738 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.707 3.712 mp_waitany 14300 13.8 2.969 3.575 2.969 3.575 grid_integrate_task_list 110 12.3 3.263 3.435 3.263 3.435 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.229 3.289 hybrid_alltoall_any 4261 16.3 0.084 0.480 2.597 3.205 yz_to_x 341 15.4 0.055 0.063 2.820 2.951 rs_gather_matrices 110 12.3 0.158 0.178 2.549 2.908 mp_sum_d 3889 11.9 1.960 2.697 1.960 2.697 calculate_first_density_matrix 1 7.0 0.000 0.000 2.616 2.625 acc_transpose_blocks 49320 15.4 0.214 0.223 2.532 2.605 jit_kernel_multiply 13 15.9 2.574 2.601 2.574 2.601 cp_fm_cholesky_invert 11 10.9 2.544 2.550 2.544 2.550 grid_collocate_task_list 110 9.6 2.158 2.363 2.158 2.363 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=115.064000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.181818, yerr=2.757409 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1668dc7f1dfc939e0dca246397b4db0b42a8bbc0_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 592.044032E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1972800 MPI messages size (bytes): total size 1.077520E+12 min size 0.000000E+00 max size 4.537280E+06 average size 546.188250E+03 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 222984 1826684928 8192 < size <= 32768 520356 13399818240 32768 < size <= 131072 372336 35386294272 131072 < size <= 4194304 787758 788321309808 4194304 < size <= 16777216 54450 238588003280 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66430. MP_Allreduce 9775 566. MP_Sync 52 MP_Alltoall 1717 3176667. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.283 0.315 70.563 70.564 qs_mol_dyn_low 1 2.0 0.003 0.004 69.785 69.796 qs_forces 11 3.9 0.004 0.005 69.612 69.614 qs_energies 11 4.9 0.001 0.001 66.223 66.227 scf_env_do_scf 11 5.9 0.001 0.003 57.583 57.587 scf_env_do_scf_inner_loop 99 6.5 0.002 0.008 49.955 49.956 dbcsr_multiply_generic 2055 12.4 0.118 0.121 38.182 38.349 velocity_verlet 10 3.0 0.002 0.008 36.635 36.639 qs_scf_new_mos 99 7.5 0.001 0.001 33.448 33.570 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.447 33.570 multiply_cannon 2055 13.4 0.226 0.248 31.409 32.568 ot_scf_mini 99 9.5 0.003 0.003 31.772 31.905 multiply_cannon_loop 2055 14.4 1.156 1.184 30.084 30.868 ot_mini 99 10.5 0.001 0.001 18.522 18.662 multiply_cannon_multrec 24660 15.4 6.981 8.283 14.046 15.479 rebuild_ks_matrix 110 8.3 0.000 0.000 13.931 14.029 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 13.930 14.029 qs_ot_get_derivative 99 11.5 0.001 0.001 12.698 12.833 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.248 12.336 mp_waitall_1 176588 16.5 7.634 10.210 7.634 10.210 multiply_cannon_metrocomm3 24660 15.4 0.073 0.077 5.181 7.976 multiply_cannon_sync_h2d 24660 15.4 6.368 7.687 6.368 7.687 init_scf_loop 11 6.9 0.000 0.000 7.592 7.593 dbcsr_mm_accdrv_process 52282 16.1 5.428 6.446 6.897 7.194 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.513 7.070 apply_single 110 13.6 0.000 0.001 6.513 7.070 sum_up_and_integrate 110 10.3 0.052 0.059 6.711 6.721 qs_ot_get_p 110 10.4 0.001 0.001 6.547 6.691 integrate_v_rspace 110 11.3 0.002 0.003 6.659 6.670 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.259 6.268 calculate_rho_elec 110 8.6 0.040 0.048 6.258 6.268 init_scf_run 11 5.9 0.000 0.001 6.250 6.251 scf_env_initial_rho_setup 11 6.9 0.004 0.005 6.250 6.250 ot_diis_step 99 11.5 0.010 0.010 5.776 5.776 prepare_preconditioner 11 7.9 0.000 0.000 5.542 5.560 make_preconditioner 11 8.9 0.000 0.000 5.542 5.560 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.742 5.457 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.085 5.239 make_m2s 4110 13.4 0.057 0.060 4.350 4.855 make_images 4110 14.4 0.409 0.466 4.240 4.741 qs_ot_p2m_diag 48 11.0 0.028 0.043 4.627 4.648 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.158 4.159 density_rs2pw 110 9.6 0.004 0.005 3.333 3.939 pw_transfer 1331 11.6 0.066 0.075 3.694 3.833 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 3.586 3.726 rs_pw_transfer 902 11.9 0.012 0.014 3.008 3.663 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.582 3.593 cp_fm_diag_elpa_base 48 14.0 3.535 3.551 3.579 3.590 wfi_extrapolate 11 7.9 0.001 0.001 3.552 3.552 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.306 3.381 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.378 3.380 grid_integrate_task_list 110 12.3 3.158 3.317 3.158 3.317 fft_wrap_pw1pw2_140 451 13.1 0.523 0.540 3.050 3.190 calculate_dm_sparse 110 9.5 0.001 0.001 2.977 3.006 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.974 3.003 hybrid_alltoall_any 4261 16.3 0.107 0.456 2.067 2.885 make_images_data 4110 15.4 0.049 0.053 2.352 2.863 fft3d_ps 1111 14.6 1.110 1.330 2.555 2.711 calculate_first_density_matrix 1 7.0 0.000 0.000 2.595 2.599 cp_fm_cholesky_invert 11 10.9 2.558 2.565 2.558 2.565 mp_sum_l 6594 12.7 1.786 2.551 1.786 2.551 grid_collocate_task_list 110 9.6 2.093 2.538 2.093 2.538 potential_pw2rs 110 12.3 0.008 0.008 2.463 2.478 jit_kernel_multiply 11 16.3 1.104 2.164 1.104 2.164 mp_alltoall_d11v 2046 13.8 1.841 2.160 1.841 2.160 acc_transpose_blocks 24660 15.4 0.118 0.122 1.966 2.003 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.981 2.000 mp_waitany 10164 13.8 1.256 1.953 1.256 1.953 qs_energies_init_hamiltonians 11 5.9 0.007 0.014 1.878 1.881 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.838 1.851 mp_allgather_i34 2055 14.4 0.630 1.710 0.630 1.710 multiply_cannon_metrocomm4 22605 15.4 0.078 0.083 0.781 1.636 rs_pw_transfer_RS2PW_140 121 11.5 0.206 0.215 0.986 1.627 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.536 1.546 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.390 1.494 dbcsr_complete_redistribute 325 12.2 0.270 0.365 1.225 1.494 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=70.564000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=560.454545, yerr=6.213821 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1668dc7f1dfc939e0dca246397b4db0b42a8bbc0_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 667.553792E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66421. MP_Allreduce 9774 562. MP_Sync 52 MP_Alltoall 1496 4511006. MP_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.107 0.424 61.031 61.056 qs_mol_dyn_low 1 2.0 0.003 0.003 59.942 59.955 qs_forces 11 3.9 0.004 0.006 59.845 59.849 qs_energies 11 4.9 0.002 0.009 56.602 56.610 scf_env_do_scf 11 5.9 0.000 0.001 48.608 48.609 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 40.130 40.131 velocity_verlet 10 3.0 0.003 0.009 32.714 32.716 dbcsr_multiply_generic 2055 12.4 0.108 0.113 28.506 28.775 qs_scf_new_mos 99 7.5 0.001 0.001 25.013 25.108 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.013 25.107 ot_scf_mini 99 9.5 0.002 0.003 23.763 23.877 multiply_cannon 2055 13.4 0.211 0.223 22.429 23.540 multiply_cannon_loop 2055 14.4 0.816 0.866 21.276 22.399 ot_mini 99 10.5 0.001 0.001 13.599 13.714 rebuild_ks_matrix 110 8.3 0.000 0.000 12.395 12.532 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.395 12.532 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.892 11.016 multiply_cannon_multrec 16440 15.4 3.738 4.903 9.915 10.970 mp_waitall_1 139946 16.5 6.661 9.577 6.661 9.577 qs_ot_get_derivative 99 11.5 0.001 0.001 9.195 9.307 init_scf_loop 11 6.9 0.001 0.005 8.444 8.445 multiply_cannon_metrocomm3 16440 15.4 0.046 0.049 4.225 6.943 prepare_preconditioner 11 7.9 0.000 0.000 6.666 6.688 make_preconditioner 11 8.9 0.000 0.002 6.666 6.687 sum_up_and_integrate 110 10.3 0.061 0.061 6.576 6.592 integrate_v_rspace 110 11.3 0.003 0.004 6.515 6.530 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.020 6.366 dbcsr_mm_accdrv_process 34862 16.1 5.331 5.774 6.023 6.197 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.065 6.076 calculate_rho_elec 110 8.6 0.059 0.059 6.064 6.075 init_scf_run 11 5.9 0.000 0.001 5.522 5.522 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.522 5.522 qs_ot_get_p 110 10.4 0.001 0.002 5.366 5.498 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.876 5.379 apply_single 110 13.6 0.000 0.000 4.876 5.379 make_m2s 4110 13.4 0.049 0.051 4.059 4.426 density_rs2pw 110 9.6 0.004 0.005 3.127 4.380 ot_diis_step 99 11.5 0.011 0.011 4.376 4.376 make_images 4110 14.4 0.400 0.518 3.944 4.311 rs_pw_transfer 902 11.9 0.010 0.011 2.748 4.009 multiply_cannon_sync_h2d 16440 15.4 3.259 3.881 3.259 3.881 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.193 3.785 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.751 3.754 pw_transfer 1331 11.6 0.066 0.074 3.657 3.664 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.550 3.559 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.382 3.382 grid_integrate_task_list 110 12.3 3.176 3.347 3.176 3.347 fft_wrap_pw1pw2_140 451 13.1 0.642 0.649 3.066 3.076 wfi_extrapolate 11 7.9 0.001 0.001 2.969 2.969 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.853 2.862 cp_fm_diag_elpa_base 48 14.0 2.787 2.816 2.852 2.861 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.777 2.779 make_images_data 4110 15.4 0.046 0.050 2.236 2.700 hybrid_alltoall_any 4261 16.3 0.110 0.386 2.051 2.670 cp_fm_cholesky_invert 11 10.9 2.600 2.606 2.600 2.606 mp_waitany 17072 13.8 1.243 2.554 1.243 2.554 calculate_dm_sparse 110 9.5 0.001 0.001 2.522 2.551 grid_collocate_task_list 110 9.6 2.123 2.526 2.123 2.526 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.449 2.497 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.412 2.479 calculate_first_density_matrix 1 7.0 0.000 0.001 2.454 2.456 multiply_cannon_metrocomm4 14385 15.4 0.047 0.051 0.870 2.410 fft3d_ps 1111 14.6 1.091 1.102 2.367 2.377 mp_sum_l 6594 12.7 1.683 2.365 1.683 2.365 potential_pw2rs 110 12.3 0.011 0.011 2.293 2.304 mp_irecv_dv 48980 15.7 0.795 2.274 0.795 2.274 rs_pw_transfer_RS2PW_140 121 11.5 0.175 0.179 0.948 2.194 mp_alltoall_d11v 2046 13.8 1.766 2.159 1.766 2.159 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.032 2.037 dbcsr_complete_redistribute 325 12.2 0.329 0.361 1.416 1.874 cp_fm_upper_to_full 70 14.2 1.395 1.770 1.395 1.770 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.700 1.716 acc_transpose_blocks 16440 15.4 0.080 0.083 1.586 1.689 cp_fm_cholesky_decompose 22 10.9 1.546 1.563 1.546 1.563 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.003 1.372 1.480 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.459 1.474 mp_allgather_i34 2055 14.4 0.467 1.447 0.467 1.447 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.956 1.409 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.285 1.291 rs_gather_matrices 110 12.3 0.234 0.262 0.967 1.291 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=61.056000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=632.636364, yerr=7.946099 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1668dc7f1dfc939e0dca246397b4db0b42a8bbc0_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 738.549760E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66419. MP_Allreduce 9774 603. MP_Sync 52 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.055 0.071 66.188 66.189 qs_mol_dyn_low 1 2.0 0.003 0.010 65.907 65.917 qs_forces 11 3.9 0.003 0.004 65.806 65.811 qs_energies 11 4.9 0.001 0.001 62.341 62.349 scf_env_do_scf 11 5.9 0.000 0.001 54.101 54.104 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 42.593 42.594 velocity_verlet 10 3.0 0.005 0.020 37.445 37.448 dbcsr_multiply_generic 2055 12.4 0.115 0.123 30.743 30.994 qs_scf_new_mos 99 7.5 0.001 0.001 27.496 27.608 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.495 27.608 ot_scf_mini 99 9.5 0.002 0.003 25.810 25.916 multiply_cannon 2055 13.4 0.249 0.281 23.821 25.011 multiply_cannon_loop 2055 14.4 1.425 1.504 22.276 22.843 ot_mini 99 10.5 0.001 0.001 14.955 15.090 multiply_cannon_multrec 24660 15.4 4.110 6.864 13.004 14.134 rebuild_ks_matrix 110 8.3 0.000 0.000 12.263 12.361 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.262 12.360 init_scf_loop 11 6.9 0.000 0.000 11.468 11.469 qs_ot_get_derivative 99 11.5 0.001 0.001 10.790 10.901 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.792 10.883 prepare_preconditioner 11 7.9 0.000 0.000 9.714 9.732 make_preconditioner 11 8.9 0.000 0.000 9.714 9.732 dbcsr_mm_accdrv_process 52304 16.0 7.477 8.931 8.737 9.694 make_full_inverse_cholesky 11 9.9 0.000 0.000 7.950 9.410 sum_up_and_integrate 110 10.3 0.067 0.071 6.504 6.519 integrate_v_rspace 110 11.3 0.003 0.003 6.437 6.453 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.151 6.158 calculate_rho_elec 110 8.6 0.078 0.081 6.151 6.158 mp_waitall_1 121746 16.5 4.203 6.024 4.203 6.024 qs_ot_get_p 110 10.4 0.001 0.001 5.719 5.873 make_m2s 4110 13.4 0.059 0.060 5.323 5.793 make_images 4110 14.4 0.579 0.696 5.184 5.652 init_scf_run 11 5.9 0.000 0.001 5.476 5.477 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.476 5.477 cp_fm_upper_to_full 70 14.2 3.339 4.846 3.339 4.846 ot_diis_step 99 11.5 0.011 0.011 4.127 4.128 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.046 4.113 apply_single 110 13.6 0.000 0.000 4.046 4.113 qs_ot_p2m_diag 48 11.0 0.054 0.063 3.869 3.883 dbcsr_complete_redistribute 325 12.2 0.425 0.475 2.632 3.776 pw_transfer 1331 11.6 0.065 0.076 3.738 3.768 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.632 3.666 density_rs2pw 110 9.6 0.004 0.004 3.015 3.540 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.485 3.537 grid_integrate_task_list 110 12.3 3.279 3.425 3.279 3.425 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.335 3.398 multiply_cannon_metrocomm3 24660 15.4 0.039 0.040 1.488 3.356 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.309 3.309 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.104 3.234 fft_wrap_pw1pw2_140 451 13.1 0.672 0.693 3.126 3.164 hybrid_alltoall_any 4261 16.3 0.124 0.462 2.223 3.057 make_images_data 4110 15.4 0.048 0.052 2.597 3.042 calculate_dm_sparse 110 9.5 0.001 0.001 2.977 3.010 wfi_extrapolate 11 7.9 0.001 0.001 2.986 2.986 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.756 2.876 rs_pw_transfer 902 11.9 0.010 0.011 2.340 2.872 mp_alltoall_i22 605 13.7 1.650 2.807 1.650 2.807 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.785 2.795 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.793 2.795 cp_fm_diag_elpa_base 48 14.0 2.638 2.699 2.783 2.794 cp_fm_cholesky_invert 11 10.9 2.591 2.599 2.591 2.599 acc_transpose_blocks 24660 15.4 0.112 0.114 2.407 2.585 multiply_cannon_sync_h2d 24660 15.4 2.386 2.556 2.386 2.556 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.438 2.477 grid_collocate_task_list 110 9.6 2.222 2.454 2.222 2.454 fft3d_ps 1111 14.6 1.084 1.116 2.388 2.403 calculate_first_density_matrix 1 7.0 0.000 0.000 2.388 2.392 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.273 2.278 potential_pw2rs 110 12.3 0.012 0.013 2.132 2.141 mp_alltoall_d11v 2046 13.8 1.726 2.109 1.726 2.109 jit_kernel_multiply 10 15.8 0.912 1.820 0.912 1.820 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.783 1.820 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.712 1.722 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.614 1.711 cp_fm_cholesky_decompose 22 10.9 1.613 1.666 1.613 1.666 mp_allgather_i34 2055 14.4 0.623 1.594 0.623 1.594 acc_transpose_blocks_sync 73980 16.4 1.399 1.565 1.399 1.565 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.535 1.550 mp_waitany 13376 13.8 1.056 1.534 1.056 1.534 multiply_cannon_metrocomm4 20550 15.4 0.061 0.065 0.850 1.491 mp_sum_l 6594 12.7 0.904 1.486 0.904 1.486 mp_irecv_dv 62702 16.1 0.745 1.411 0.745 1.411 rs_pw_transfer_RS2PW_140 121 11.5 0.167 0.173 0.816 1.356 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=66.189000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=698.181818, yerr=11.495597 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1668dc7f1dfc939e0dca246397b4db0b42a8bbc0_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 871.477248E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66417. MP_Allreduce 9774 644. MP_Sync 52 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.031 73.653 73.654 qs_mol_dyn_low 1 2.0 0.010 0.112 73.264 73.276 qs_forces 11 3.9 0.099 0.106 73.062 73.070 qs_energies 11 4.9 0.004 0.005 68.912 68.918 scf_env_do_scf 11 5.9 0.000 0.001 59.261 59.262 scf_env_do_scf_inner_loop 99 6.5 0.004 0.007 49.028 49.048 velocity_verlet 10 3.0 0.011 0.030 41.151 41.160 dbcsr_multiply_generic 2055 12.4 0.105 0.109 29.055 29.436 qs_scf_new_mos 99 7.5 0.001 0.001 28.517 28.594 qs_scf_loop_do_ot 99 8.5 0.001 0.001 28.516 28.594 ot_scf_mini 99 9.5 0.002 0.002 26.921 26.972 multiply_cannon 2055 13.4 0.247 0.257 20.310 22.137 multiply_cannon_loop 2055 14.4 0.603 0.623 18.350 18.964 rebuild_ks_matrix 110 8.3 0.000 0.000 15.792 15.871 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 15.792 15.870 ot_mini 99 10.5 0.001 0.001 14.592 14.639 qs_ks_update_qs_env 110 7.6 0.001 0.001 14.296 14.346 mp_waitall_1 103326 16.6 9.469 12.100 9.469 12.100 init_scf_loop 11 6.9 0.000 0.000 10.186 10.195 qs_ot_get_derivative 99 11.5 0.001 0.001 9.567 9.615 sum_up_and_integrate 110 10.3 0.080 0.082 8.977 8.991 integrate_v_rspace 110 11.3 0.003 0.003 8.897 8.910 multiply_cannon_multrec 8220 15.4 3.140 4.601 7.558 8.859 qs_rho_update_rho_low 110 7.6 0.001 0.001 8.425 8.447 calculate_rho_elec 110 8.6 0.114 0.115 8.424 8.446 prepare_preconditioner 11 7.9 0.000 0.000 8.066 8.071 make_preconditioner 11 8.9 0.000 0.000 8.066 8.071 make_full_inverse_cholesky 11 9.9 0.000 0.000 7.475 7.640 qs_ot_get_p 110 10.4 0.003 0.003 7.384 7.448 make_m2s 4110 13.4 0.038 0.039 6.356 6.989 make_images 4110 14.4 0.654 0.716 6.226 6.858 init_scf_run 11 5.9 0.000 0.001 6.041 6.041 scf_env_initial_rho_setup 11 6.9 0.001 0.002 6.041 6.041 pw_transfer 1331 11.6 0.066 0.072 5.992 6.017 multiply_cannon_metrocomm3 8220 15.4 0.019 0.020 3.613 5.932 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 5.884 5.913 qs_ot_p2m_diag 48 11.0 0.081 0.085 5.247 5.280 density_rs2pw 110 9.6 0.004 0.004 4.861 5.231 dbcsr_mm_accdrv_process 17442 15.9 3.153 4.381 4.279 5.230 ot_diis_step 99 11.5 0.013 0.014 5.004 5.008 fft_wrap_pw1pw2_140 451 13.1 0.836 0.849 4.762 4.795 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.762 4.791 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.326 4.538 apply_single 110 13.6 0.000 0.000 4.326 4.537 make_images_data 4110 15.4 0.041 0.046 3.661 4.527 fft3d_ps 1111 14.6 1.137 1.167 4.426 4.454 rs_pw_transfer 902 11.9 0.010 0.010 3.959 4.347 hybrid_alltoall_any 4261 16.3 0.201 0.858 3.408 4.110 cp_fm_cholesky_invert 11 10.9 4.041 4.065 4.041 4.065 mp_alltoall_d11v 2046 13.8 3.744 4.008 3.744 4.008 potential_pw2rs 110 12.3 0.015 0.015 3.889 3.893 wfi_extrapolate 11 7.9 0.001 0.001 3.573 3.573 grid_integrate_task_list 110 12.3 3.378 3.561 3.378 3.561 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.360 3.368 cp_fm_diag_elpa_base 48 14.0 3.261 3.326 3.358 3.366 mp_alltoall_z22v 1111 16.6 3.058 3.081 3.058 3.081 multiply_cannon_metrocomm1 8220 15.4 0.022 0.023 2.186 3.076 mp_waitany 9240 13.8 2.557 2.961 2.557 2.961 calculate_dm_sparse 110 9.5 0.001 0.001 2.892 2.955 mp_allgather_i34 2055 14.4 1.120 2.929 1.120 2.929 qs_energies_init_hamiltonians 11 5.9 0.025 0.028 2.899 2.900 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.836 2.859 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.846 2.848 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.767 2.818 grid_collocate_task_list 110 9.6 2.327 2.594 2.327 2.594 dbcsr_complete_redistribute 325 12.2 0.572 0.598 2.381 2.528 multiply_cannon_sync_h2d 8220 15.4 2.312 2.494 2.312 2.494 rs_pw_transfer_RS2PW_140 121 11.5 0.172 0.178 1.968 2.369 calculate_first_density_matrix 1 7.0 0.000 0.000 2.335 2.336 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.219 2.240 cp_fm_cholesky_decompose 22 10.9 2.206 2.229 2.206 2.229 make_images_sizes 4110 15.4 0.005 0.005 1.074 2.047 mp_sum_d 3887 11.9 1.680 2.046 1.680 2.046 mp_alltoall_i44 4110 16.4 1.069 2.042 1.069 2.042 mp_sum_l 6594 12.7 1.645 2.025 1.645 2.025 yz_to_x 561 15.1 0.101 0.103 1.989 2.004 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.784 1.996 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.894 1.920 dbcsr_dot_sd 1091 11.9 0.532 0.538 1.597 1.831 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.685 1.806 rs_gather_matrices 110 12.3 0.327 0.370 1.541 1.793 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.756 1.788 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.613 1.762 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.712 1.723 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.577 1.595 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=73.654000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=817.272727, yerr=16.293283 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1668dc7f1dfc939e0dca246397b4db0b42a8bbc0_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 1.379906E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67098. MP_Allreduce 9752 812. MP_Sync 52 MP_Alltoall 1474 16505187. MP_SendRecv 2310 360267. MP_ISendRecv 2310 360267. MP_Wait 5214 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.043 0.052 88.205 88.206 qs_mol_dyn_low 1 2.0 0.006 0.032 87.507 87.516 qs_forces 11 3.9 0.037 0.043 87.434 87.439 qs_energies 11 4.9 0.004 0.005 83.185 83.189 scf_env_do_scf 11 5.9 0.000 0.001 72.942 72.943 velocity_verlet 10 3.0 0.002 0.003 56.005 56.011 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 44.661 44.662 dbcsr_multiply_generic 2055 12.4 0.120 0.125 29.291 29.388 init_scf_loop 11 6.9 0.000 0.000 28.210 28.213 qs_scf_new_mos 99 7.5 0.001 0.001 26.476 26.510 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.475 26.509 prepare_preconditioner 11 7.9 0.000 0.000 26.164 26.172 make_preconditioner 11 8.9 0.000 0.000 26.164 26.172 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.411 25.642 ot_scf_mini 99 9.5 0.002 0.002 24.673 24.699 multiply_cannon 2055 13.4 0.337 0.362 22.307 23.083 multiply_cannon_loop 2055 14.4 0.813 0.834 20.520 20.826 cp_fm_upper_to_full 70 14.2 12.811 18.288 12.811 18.288 rebuild_ks_matrix 110 8.3 0.000 0.001 14.119 14.155 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 14.118 14.154 ot_mini 99 10.5 0.001 0.001 13.942 13.969 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.760 12.793 dbcsr_complete_redistribute 325 12.2 1.058 1.081 7.337 10.537 multiply_cannon_multrec 8220 15.4 4.075 4.237 9.757 9.884 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.325 9.524 qs_ot_get_derivative 99 11.5 0.001 0.001 9.385 9.411 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.737 8.900 mp_alltoall_i22 605 13.7 5.359 8.574 5.359 8.574 mp_waitall_1 84994 16.7 7.332 8.128 7.332 8.128 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.684 7.725 calculate_rho_elec 110 8.6 0.224 0.225 7.683 7.724 sum_up_and_integrate 110 10.3 0.150 0.151 7.282 7.294 integrate_v_rspace 110 11.3 0.004 0.004 7.132 7.144 init_scf_run 11 5.9 0.000 0.001 5.905 5.905 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.905 5.905 make_m2s 4110 13.4 0.043 0.044 5.305 5.802 dbcsr_mm_accdrv_process 11614 15.7 3.846 4.146 5.534 5.763 make_images 4110 14.4 0.887 0.944 5.116 5.612 qs_ot_get_p 110 10.4 0.002 0.002 5.459 5.490 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.702 5.168 apply_single 110 13.6 0.000 0.000 4.702 5.168 cp_fm_cholesky_invert 11 10.9 5.145 5.149 5.145 5.149 multiply_cannon_metrocomm3 8220 15.4 0.019 0.020 4.744 5.078 pw_transfer 1331 11.6 0.075 0.076 4.967 4.973 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 4.849 4.855 ot_diis_step 99 11.5 0.015 0.015 4.540 4.540 fft_wrap_pw1pw2_140 451 13.1 1.338 1.341 4.243 4.250 qs_ot_p2m_diag 48 11.0 0.151 0.156 3.938 3.946 grid_integrate_task_list 110 12.3 3.697 3.757 3.697 3.757 qs_energies_init_hamiltonians 11 5.9 0.047 0.052 3.729 3.731 density_rs2pw 110 9.6 0.004 0.004 3.677 3.704 hybrid_alltoall_any 4261 16.3 0.264 0.565 2.763 3.491 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.987 3.441 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.425 3.426 wfi_extrapolate 11 7.9 0.001 0.001 3.341 3.341 make_images_data 4110 15.4 0.045 0.048 2.734 3.329 calculate_dm_sparse 110 9.5 0.001 0.001 3.169 3.198 multiply_cannon_sync_h2d 8220 15.4 3.147 3.180 3.147 3.180 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.965 2.966 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.852 2.852 cp_fm_diag_elpa_base 48 14.0 2.310 2.511 2.850 2.851 fft3d_ps 1111 14.6 1.302 1.317 2.775 2.780 grid_collocate_task_list 110 9.6 2.688 2.714 2.688 2.714 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.613 2.629 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.434 2.445 calculate_first_density_matrix 1 7.0 0.000 0.000 2.442 2.444 potential_pw2rs 110 12.3 0.021 0.021 2.357 2.360 rs_pw_transfer 902 11.9 0.010 0.011 2.237 2.281 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.178 2.242 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.098 2.187 mp_alltoall_d11v 2046 13.8 1.899 1.990 1.899 1.990 cp_fm_cholesky_decompose 22 10.9 1.941 1.962 1.941 1.962 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.955 1.959 qs_create_task_list 11 7.9 0.001 0.001 1.883 1.933 generate_qs_task_list 11 8.9 0.732 0.786 1.882 1.932 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.736 1.770 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=88.206000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1235.818182, yerr=57.386755 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1668dc7f1dfc939e0dca246397b4db0b42a8bbc0_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 1.103326E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 12044928 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.1 marketing flops 145.647559E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 629.207040E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10431456 MPI messages size (bytes): total size 4.526012E+12 min size 0.000000E+00 max size 4.537280E+06 average size 433.881156E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3605352 96403587072 32768 < size <= 131072 1305088 74666999808 131072 < size <= 4194304 5190114 3200148350408 4194304 < size <= 16777216 263934 1154745839544 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4024 57450. MP_Allreduce 11139 794. MP_Sync 87 MP_Alltoall 2242 1858232. MP_SendRecv 24510 18752. MP_ISendRecv 24510 18752. MP_Wait 42802 MP_ISend 16140 108019. MP_IRecv 16140 108019. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.020 0.061 208.321 208.324 qs_mol_dyn_low 1 2.0 0.006 0.027 207.734 207.748 qs_forces 11 3.9 0.005 0.005 207.101 207.103 qs_energies 11 4.9 0.001 0.002 201.450 201.467 scf_env_do_scf 11 5.9 0.001 0.002 184.858 184.862 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 164.597 164.599 dbcsr_multiply_generic 2527 12.6 0.184 0.189 126.152 126.953 velocity_verlet 10 3.0 0.001 0.002 124.958 124.964 qs_scf_new_mos 118 7.6 0.001 0.001 124.191 124.480 qs_scf_loop_do_ot 118 8.6 0.001 0.001 124.191 124.479 ot_scf_mini 118 9.6 0.003 0.003 117.387 117.678 multiply_cannon 2527 13.6 0.238 0.247 102.246 104.181 multiply_cannon_loop 2527 14.6 2.440 2.491 100.084 101.904 ot_mini 118 10.6 0.001 0.001 66.184 66.443 multiply_cannon_multrec 60648 15.6 32.042 34.004 41.885 43.173 qs_ot_get_derivative 118 11.6 0.001 0.001 41.298 41.554 rebuild_ks_matrix 129 8.3 0.001 0.001 33.889 34.239 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.027 33.889 34.239 mp_waitall_1 269254 16.5 28.797 31.974 28.797 31.974 qs_ks_update_qs_env 129 7.6 0.001 0.001 30.396 30.712 qs_ot_get_p 129 10.4 0.001 0.002 28.801 29.127 multiply_cannon_sync_h2d 60648 15.6 26.473 28.026 26.473 28.026 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 24.397 25.121 apply_single 129 13.6 0.001 0.001 24.396 25.121 ot_diis_step 118 11.6 0.008 0.008 24.630 24.631 qs_ot_p2m_diag 83 11.4 0.079 0.091 22.251 22.317 init_scf_loop 11 6.9 0.000 0.001 20.187 20.189 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.658 19.890 cp_dbcsr_syevd 83 12.4 0.005 0.005 19.391 19.391 multiply_cannon_metrocomm3 60648 15.6 0.118 0.126 16.086 18.309 cp_fm_diag_elpa 83 13.4 0.000 0.000 16.476 16.513 cp_fm_diag_elpa_base 83 14.4 16.408 16.446 16.472 16.507 prepare_preconditioner 11 7.9 0.000 0.000 15.606 15.659 make_preconditioner 11 8.9 0.000 0.000 15.606 15.659 make_full_inverse_cholesky 11 9.9 0.000 0.000 14.840 14.982 make_m2s 5054 13.6 0.105 0.112 13.961 14.332 sum_up_and_integrate 129 10.3 0.089 0.106 14.256 14.271 integrate_v_rspace 129 11.3 0.004 0.005 14.167 14.184 make_images 5054 14.6 0.398 0.418 13.776 14.159 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.941 14.125 calculate_rho_elec 129 8.7 0.046 0.065 13.941 14.124 init_scf_run 11 5.9 0.000 0.001 12.456 12.457 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.456 12.457 density_rs2pw 129 9.7 0.006 0.007 7.085 10.383 dbcsr_mm_accdrv_process 125468 16.2 4.741 4.897 9.413 9.935 mp_sum_l 8010 12.9 8.707 9.701 8.707 9.701 wfi_extrapolate 11 7.9 0.001 0.001 9.067 9.067 rs_pw_transfer 1054 12.0 0.017 0.018 5.690 9.028 cp_fm_cholesky_invert 11 10.9 8.820 8.828 8.820 8.828 calculate_dm_sparse 129 9.5 0.001 0.001 8.609 8.730 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 8.238 8.359 qs_ot_get_orbitals 118 10.6 0.001 0.001 8.193 8.278 pw_transfer 1559 11.6 0.075 0.092 7.794 7.967 multiply_cannon_metrocomm1 60648 15.6 0.096 0.100 6.198 7.928 make_images_data 5054 15.6 0.068 0.074 6.855 7.793 fft_wrap_pw1pw2 1301 12.7 0.011 0.013 7.590 7.764 grid_integrate_task_list 129 12.3 7.159 7.592 7.159 7.592 hybrid_alltoall_any 5240 16.5 0.296 2.279 6.007 7.222 fft_wrap_pw1pw2_140 527 13.2 1.282 1.329 6.676 6.870 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.004 6.820 6.831 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.957 6.128 grid_collocate_task_list 129 9.7 4.696 6.016 4.696 6.016 mp_waitany 16140 13.9 2.687 5.998 2.687 5.998 fft3d_ps 1301 14.7 2.161 2.804 5.342 5.664 mp_alltoall_d11v 2423 14.1 4.314 5.515 4.314 5.515 rs_pw_transfer_RS2PW_140 140 11.5 0.280 0.298 2.166 5.493 cp_fm_cholesky_decompose 22 10.9 4.597 4.609 4.597 4.609 potential_pw2rs 129 12.3 0.009 0.010 4.538 4.558 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=208.324000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=596.727273, yerr=5.801111 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1668dc7f1dfc939e0dca246397b4db0b42a8bbc0_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 838.578176E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100943E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834845264 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4020 57958. MP_Allreduce 11128 958. MP_Sync 87 MP_Alltoall 1969 5157912. MP_SendRecv 12032 47072. MP_ISendRecv 12032 47072. MP_Wait 25916 MP_ISend 11748 212467. MP_IRecv 11748 212467. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.118 0.228 191.320 191.321 qs_mol_dyn_low 1 2.0 0.003 0.004 190.031 190.601 qs_forces 11 3.9 0.004 0.005 189.936 189.938 qs_energies 11 4.9 0.003 0.009 183.105 183.119 scf_env_do_scf 11 5.9 0.001 0.001 166.468 166.478 scf_env_do_scf_inner_loop 117 6.6 0.005 0.008 133.617 133.620 velocity_verlet 10 3.0 0.001 0.002 119.407 119.409 dbcsr_multiply_generic 2507 12.6 0.191 0.196 97.001 98.294 qs_scf_new_mos 117 7.6 0.001 0.001 93.999 94.515 qs_scf_loop_do_ot 117 8.6 0.001 0.002 93.999 94.514 ot_scf_mini 117 9.6 0.004 0.005 89.196 89.780 multiply_cannon 2507 13.6 0.502 0.557 76.716 81.463 multiply_cannon_loop 2507 14.6 1.559 1.633 73.793 76.592 ot_mini 117 10.6 0.001 0.001 50.188 50.698 mp_waitall_1 214728 16.6 24.014 37.888 24.014 37.888 multiply_cannon_multrec 30084 15.6 20.845 25.145 31.478 36.055 rebuild_ks_matrix 128 8.3 0.001 0.001 32.775 33.490 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.023 32.774 33.489 init_scf_loop 11 6.9 0.001 0.006 32.763 32.764 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.439 30.081 qs_ot_get_derivative 117 11.6 0.001 0.002 28.183 28.757 prepare_preconditioner 11 7.9 0.000 0.000 28.410 28.484 make_preconditioner 11 8.9 0.000 0.001 28.410 28.483 multiply_cannon_metrocomm3 30084 15.6 0.097 0.102 15.734 28.190 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.109 27.649 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.149 23.315 apply_single 128 13.6 0.001 0.001 22.149 23.315 qs_ot_get_p 128 10.4 0.003 0.012 21.261 21.941 ot_diis_step 117 11.6 0.014 0.015 21.831 21.833 multiply_cannon_sync_h2d 30084 15.6 17.900 20.340 17.900 20.340 cp_fm_cholesky_invert 11 10.9 16.605 16.617 16.605 16.617 qs_ot_p2m_diag 83 11.4 0.189 0.218 16.339 16.375 make_m2s 5014 13.6 0.090 0.095 14.279 15.436 make_images 5014 14.6 1.180 1.373 14.064 15.222 cp_dbcsr_syevd 83 12.4 0.006 0.006 15.128 15.129 sum_up_and_integrate 128 10.3 0.116 0.133 14.676 14.706 integrate_v_rspace 128 11.3 0.003 0.004 14.560 14.595 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.128 14.160 calculate_rho_elec 128 8.7 0.088 0.104 14.127 14.159 cp_fm_diag_elpa 83 13.4 0.000 0.001 11.970 12.011 cp_fm_diag_elpa_base 83 14.4 11.714 11.803 11.965 12.000 init_scf_run 11 5.9 0.000 0.001 11.781 11.783 scf_env_initial_rho_setup 11 6.9 0.001 0.002 11.781 11.782 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 11.141 11.573 multiply_cannon_metrocomm4 27577 15.6 0.108 0.122 3.923 10.980 dbcsr_mm_accdrv_process 62242 16.2 5.515 6.087 10.084 10.676 mp_irecv_dv 69486 16.3 3.711 10.574 3.711 10.574 make_images_data 5014 15.6 0.068 0.079 8.300 9.830 density_rs2pw 128 9.7 0.007 0.007 7.430 9.805 hybrid_alltoall_any 5200 16.5 0.354 1.594 6.987 9.106 pw_transfer 1547 11.6 0.086 0.094 8.777 8.831 fft_wrap_pw1pw2 1291 12.7 0.011 0.011 8.552 8.610 wfi_extrapolate 11 7.9 0.001 0.001 8.328 8.328 rs_pw_transfer 1046 11.9 0.014 0.018 5.698 8.089 fft_wrap_pw1pw2_140 523 13.2 1.341 1.361 7.573 7.642 grid_integrate_task_list 128 12.3 7.190 7.600 7.190 7.600 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.273 7.043 cp_fm_cholesky_decompose 22 10.9 6.935 7.017 6.935 7.017 calculate_dm_sparse 128 9.5 0.001 0.001 6.456 6.601 mp_sum_l 7950 12.9 4.209 6.496 4.209 6.496 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.258 6.267 grid_collocate_task_list 128 9.7 4.781 5.948 4.781 5.948 fft3d_ps 1291 14.7 2.816 2.989 5.852 5.893 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.380 5.597 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.356 5.425 mp_waitany 11748 13.9 2.525 4.992 2.525 4.992 potential_pw2rs 128 12.3 0.015 0.018 4.912 4.929 mp_alltoall_d11v 2415 14.1 4.122 4.726 4.122 4.726 rs_pw_transfer_RS2PW_140 139 11.5 0.349 0.375 2.136 4.513 mp_allgather_i34 2507 14.6 1.434 4.445 1.434 4.445 mp_sum_d 4479 12.1 2.648 4.075 2.648 4.075 dbcsr_complete_redistribute 395 12.7 0.770 0.851 3.087 3.948 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=191.321000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=798.363636, yerr=3.937529 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1668dc7f1dfc939e0dca246397b4db0b42a8bbc0_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 2.928533E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 3984192 0.0% 0.0% 100.0% average stack size 0.0 0.0 1695.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 953.880576E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1042912 MPI messages size (bytes): total size 2.716210E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.604448E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 281856 36943429632 131072 < size <= 4194304 660064 996105256960 4194304 < size <= 16777216 65632 931531265168 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 58208. MP_Allreduce 11082 999. MP_Sync 87 MP_Alltoall 1712 9388896. MP_SendRecv 7936 75008. MP_ISendRecv 7936 75008. MP_Wait 21820 MP_ISend 11748 275205. MP_IRecv 11748 275205. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.074 0.227 176.738 176.739 qs_mol_dyn_low 1 2.0 0.004 0.008 175.885 175.908 qs_forces 11 3.9 0.005 0.007 175.779 175.781 qs_energies 11 4.9 0.002 0.008 169.112 169.120 scf_env_do_scf 11 5.9 0.002 0.009 153.341 153.343 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 118.538 118.556 velocity_verlet 10 3.0 0.002 0.004 112.706 112.708 dbcsr_multiply_generic 2507 12.6 0.182 0.188 81.820 82.793 qs_scf_new_mos 117 7.6 0.001 0.001 80.093 80.511 qs_scf_loop_do_ot 117 8.6 0.001 0.004 80.092 80.510 ot_scf_mini 117 9.6 0.004 0.006 75.900 76.336 multiply_cannon 2507 13.6 0.501 0.518 61.924 66.630 multiply_cannon_loop 2507 14.6 1.125 1.182 59.345 62.103 ot_mini 117 10.6 0.001 0.002 42.400 42.841 init_scf_loop 11 6.9 0.002 0.011 34.702 34.720 mp_waitall_1 170520 16.6 25.032 34.643 25.032 34.643 rebuild_ks_matrix 128 8.3 0.001 0.001 30.919 31.445 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.023 30.918 31.444 prepare_preconditioner 11 7.9 0.000 0.004 30.617 30.666 make_preconditioner 11 8.9 0.000 0.002 30.617 30.666 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.299 29.648 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.787 28.275 multiply_cannon_multrec 20056 15.6 13.003 16.529 22.359 25.961 multiply_cannon_metrocomm3 20056 15.6 0.062 0.067 15.664 25.006 qs_ot_get_derivative 117 11.6 0.002 0.002 22.860 23.286 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.633 20.699 apply_single 128 13.6 0.001 0.001 19.633 20.699 qs_ot_get_p 128 10.4 0.001 0.002 18.889 19.444 ot_diis_step 117 11.6 0.019 0.022 19.427 19.427 make_m2s 5014 13.6 0.079 0.087 14.884 15.570 make_images 5014 14.6 1.152 1.238 14.650 15.326 multiply_cannon_sync_h2d 20056 15.6 13.534 15.168 13.534 15.168 sum_up_and_integrate 128 10.3 0.132 0.145 14.674 14.713 qs_ot_p2m_diag 83 11.4 0.266 0.272 14.625 14.635 integrate_v_rspace 128 11.3 0.004 0.005 14.542 14.573 qs_rho_update_rho_low 128 7.7 0.001 0.002 14.489 14.527 calculate_rho_elec 128 8.7 0.130 0.146 14.488 14.525 cp_fm_cholesky_invert 11 10.9 14.319 14.328 14.319 14.328 cp_dbcsr_syevd 83 12.4 0.005 0.006 13.570 13.570 init_scf_run 11 5.9 0.000 0.001 10.713 10.713 scf_env_initial_rho_setup 11 6.9 0.001 0.003 10.713 10.713 cp_fm_diag_elpa 83 13.4 0.000 0.000 10.402 10.425 cp_fm_diag_elpa_base 83 14.4 9.992 10.147 10.399 10.421 make_images_data 5014 15.6 0.063 0.071 8.933 10.049 multiply_cannon_metrocomm4 17549 15.6 0.065 0.075 3.530 9.556 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.010 9.312 mp_irecv_dv 50230 16.2 3.400 9.287 3.400 9.287 hybrid_alltoall_any 5200 16.5 0.451 2.052 7.770 9.279 density_rs2pw 128 9.7 0.006 0.006 7.170 9.013 pw_transfer 1547 11.6 0.086 0.104 8.902 9.004 dbcsr_mm_accdrv_process 41502 16.2 5.577 5.944 8.823 8.965 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.678 8.786 fft_wrap_pw1pw2_140 523 13.2 1.418 1.440 7.703 7.823 grid_integrate_task_list 128 12.3 7.294 7.726 7.294 7.726 wfi_extrapolate 11 7.9 0.001 0.002 7.479 7.479 cp_fm_cholesky_decompose 22 10.9 7.260 7.291 7.260 7.291 cp_fm_upper_to_full 105 14.8 5.671 7.127 5.671 7.127 rs_pw_transfer 1046 11.9 0.014 0.015 5.174 7.009 dbcsr_complete_redistribute 395 12.7 1.174 1.211 4.547 6.238 grid_collocate_task_list 128 9.7 4.979 5.889 4.979 5.889 fft3d_ps 1291 14.7 2.739 2.969 5.793 5.876 calculate_dm_sparse 128 9.5 0.001 0.001 5.779 5.863 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.630 5.636 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.622 5.260 mp_alltoall_d11v 2415 14.1 4.421 5.132 4.421 5.132 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.392 5.083 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.730 4.881 mp_sum_l 7950 12.9 3.228 4.872 3.228 4.872 potential_pw2rs 128 12.3 0.020 0.022 4.757 4.780 mp_waitany 11748 13.9 2.388 4.297 2.388 4.297 mp_allgather_i34 2507 14.6 1.080 4.184 1.080 4.184 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.958 3.992 transfer_fm_to_dbcsr 11 9.9 0.017 0.022 2.298 3.985 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 3.778 3.780 rs_pw_transfer_RS2PW_140 139 11.5 0.328 0.348 1.940 3.775 mp_alltoall_i22 716 14.1 1.871 3.687 1.871 3.687 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.633 3.671 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=176.739000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=905.818182, yerr=7.107207 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1668dc7f1dfc939e0dca246397b4db0b42a8bbc0_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430457200640 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 1986252263424 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992001093632 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 11613077360640 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239162695680 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239162695680 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 94.232994E+12 0.0% 0.0% 100.0% flops max/rank 4.387242E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806382528 0.0% 0.0% 100.0% number of processed stacks 6026880 0.0% 0.0% 100.0% average stack size 0.0 0.0 1129.3 marketing flops 145.651870E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 1.145803E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1153224 MPI messages size (bytes): total size 2.039489E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.768511E+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 322096 36390305792 131072 < size <= 4194304 721976 792118951936 4194304 < size <= 16777216 70800 669921808880 16777216 < size 30960 541065216000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 57626. MP_Allreduce 11184 1079. MP_Sync 88 MP_Alltoall 1724 12509632. MP_SendRecv 5934 75008. MP_ISendRecv 5934 75008. MP_Wait 22612 MP_ISend 15064 244788. MP_IRecv 15064 244788. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.050 0.163 204.455 204.460 qs_mol_dyn_low 1 2.0 0.003 0.003 203.671 203.685 qs_forces 11 3.9 0.004 0.004 203.552 203.564 qs_energies 11 4.9 0.002 0.007 196.246 196.260 scf_env_do_scf 11 5.9 0.001 0.001 178.947 178.955 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 131.101 131.102 velocity_verlet 10 3.0 0.002 0.002 129.470 129.473 qs_scf_new_mos 118 7.6 0.001 0.001 90.282 90.585 qs_scf_loop_do_ot 118 8.6 0.001 0.001 90.281 90.585 dbcsr_multiply_generic 2529 12.6 0.190 0.196 85.540 86.414 ot_scf_mini 118 9.6 0.003 0.004 85.276 85.584 multiply_cannon 2529 13.6 0.554 0.591 57.361 60.910 multiply_cannon_loop 2529 14.6 1.850 1.946 53.357 55.441 init_scf_loop 11 6.9 0.002 0.012 47.715 47.716 ot_mini 118 10.6 0.001 0.001 46.591 46.868 prepare_preconditioner 11 7.9 0.000 0.001 43.496 43.519 make_preconditioner 11 8.9 0.000 0.002 43.496 43.519 make_full_inverse_cholesky 11 9.9 0.012 0.023 37.006 42.075 rebuild_ks_matrix 129 8.3 0.001 0.001 32.207 32.532 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.019 32.207 32.532 multiply_cannon_multrec 30348 15.6 13.809 18.243 26.827 30.944 qs_ks_update_qs_env 129 7.6 0.001 0.001 29.183 29.472 mp_waitall_1 149172 16.7 20.077 29.282 20.077 29.282 qs_ot_get_derivative 118 11.6 0.001 0.002 26.411 26.714 qs_ot_get_p 129 10.4 0.001 0.002 24.173 24.510 make_m2s 5058 13.6 0.095 0.102 22.895 23.967 make_images 5058 14.6 1.980 2.242 22.595 23.669 ot_diis_step 118 11.6 0.018 0.019 20.031 20.033 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 19.382 20.014 apply_single 129 13.6 0.001 0.001 19.381 20.014 qs_ot_p2m_diag 84 11.4 0.347 0.394 19.109 19.156 cp_dbcsr_syevd 84 12.4 0.005 0.006 17.519 17.522 cp_fm_upper_to_full 106 14.8 11.559 17.061 11.559 17.061 cp_fm_cholesky_invert 11 10.9 16.843 16.855 16.843 16.855 sum_up_and_integrate 129 10.3 0.141 0.152 16.087 16.114 integrate_v_rspace 129 11.3 0.004 0.004 15.946 15.976 qs_rho_update_rho_low 129 7.7 0.001 0.001 15.736 15.775 calculate_rho_elec 129 8.7 0.174 0.191 15.736 15.774 multiply_cannon_metrocomm3 30348 15.6 0.051 0.054 6.434 15.466 make_images_data 5058 15.6 0.065 0.071 12.400 14.427 cp_fm_diag_elpa 84 13.4 0.000 0.000 13.699 13.713 cp_fm_diag_elpa_base 84 14.4 12.658 13.028 13.691 13.704 dbcsr_complete_redistribute 397 12.7 1.553 1.666 9.590 13.407 dbcsr_mm_accdrv_process 62780 16.2 8.559 9.364 12.593 13.087 hybrid_alltoall_any 5245 16.5 0.536 2.241 10.709 12.849 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 11.811 12.039 copy_fm_to_dbcsr 210 11.7 0.001 0.002 8.102 11.931 multiply_cannon_sync_h2d 30348 15.6 10.547 11.396 10.547 11.396 init_scf_run 11 5.9 0.000 0.001 11.309 11.310 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.309 11.310 pw_transfer 1559 11.6 0.087 0.100 10.329 10.406 transfer_fm_to_dbcsr 11 9.9 0.001 0.007 6.467 10.200 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 10.102 10.185 mp_alltoall_i22 720 14.1 5.797 9.742 5.797 9.742 density_rs2pw 129 9.7 0.006 0.007 7.838 9.135 fft_wrap_pw1pw2_140 527 13.2 1.574 1.601 8.811 8.887 cp_fm_cholesky_decompose 22 10.9 7.939 8.025 7.939 8.025 grid_integrate_task_list 129 12.3 7.549 7.974 7.549 7.974 wfi_extrapolate 11 7.9 0.001 0.001 7.954 7.954 fft3d_ps 1301 14.7 2.858 2.942 6.999 7.074 rs_pw_transfer 1054 12.0 0.013 0.015 5.756 7.072 mp_alltoall_d11v 2429 14.1 6.043 6.840 6.043 6.840 calculate_dm_sparse 129 9.5 0.001 0.001 6.701 6.802 multiply_cannon_metrocomm4 25290 15.6 0.084 0.092 2.786 6.049 grid_collocate_task_list 129 9.7 5.192 5.902 5.192 5.902 mp_irecv_dv 76751 16.2 2.633 5.788 2.633 5.788 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.711 5.742 potential_pw2rs 129 12.3 0.024 0.024 5.475 5.502 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.663 4.753 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.497 4.612 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.510 4.595 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 4.506 4.510 mp_sum_l 8016 12.9 3.351 4.304 3.351 4.304 mp_allgather_i34 2529 14.6 2.074 4.154 2.074 4.154 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=204.460000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1085.909091, yerr=15.523909 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1668dc7f1dfc939e0dca246397b4db0b42a8bbc0_performance_tests/19/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 5.820057E+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 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 1.506542E+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 716108490000 ------------------------------------------------------------------------------- - - - 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 3981 58503. MP_Allreduce 11030 1168. MP_Sync 86 MP_Alltoall 1700 18828162. 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 0.025 0.054 196.251 196.256 qs_mol_dyn_low 1 2.0 0.003 0.003 195.802 195.826 qs_forces 11 3.9 0.004 0.004 195.666 195.671 qs_energies 11 4.9 0.007 0.008 188.030 188.037 scf_env_do_scf 11 5.9 0.001 0.001 170.182 170.187 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 132.394 132.395 velocity_verlet 10 3.0 0.002 0.002 115.517 115.522 qs_scf_new_mos 116 7.6 0.001 0.001 87.642 87.805 qs_scf_loop_do_ot 116 8.6 0.001 0.001 87.641 87.805 ot_scf_mini 116 9.6 0.003 0.003 82.741 83.047 dbcsr_multiply_generic 2485 12.5 0.185 0.198 80.938 81.504 multiply_cannon 2485 13.5 0.556 0.605 56.717 61.535 multiply_cannon_loop 2485 14.5 0.810 0.848 53.143 54.350 ot_mini 116 10.6 0.001 0.001 43.144 43.395 init_scf_loop 11 6.9 0.000 0.000 37.633 37.635 mp_waitall_1 124680 16.7 30.449 37.452 30.449 37.452 prepare_preconditioner 11 7.9 0.000 0.000 33.577 33.601 make_preconditioner 11 8.9 0.000 0.000 33.577 33.601 rebuild_ks_matrix 127 8.3 0.001 0.001 32.671 32.847 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.018 32.670 32.847 make_full_inverse_cholesky 11 9.9 0.017 0.026 31.375 31.737 qs_ks_update_qs_env 127 7.6 0.001 0.001 29.873 30.041 qs_ot_get_p 127 10.4 0.002 0.002 25.408 25.751 qs_ot_get_derivative 116 11.6 0.001 0.002 25.268 25.523 multiply_cannon_multrec 9940 15.5 10.196 14.194 17.789 20.561 multiply_cannon_metrocomm3 9940 15.5 0.025 0.027 12.287 20.222 qs_ot_p2m_diag 82 11.4 0.490 0.496 20.139 20.190 make_m2s 4970 13.5 0.065 0.070 18.640 19.875 make_images 4970 14.5 2.159 2.581 18.337 19.568 cp_fm_cholesky_invert 11 10.9 19.465 19.471 19.465 19.471 qs_rho_update_rho_low 127 7.7 0.001 0.001 18.915 18.977 calculate_rho_elec 127 8.7 0.254 0.264 18.914 18.977 cp_dbcsr_syevd 82 12.4 0.005 0.006 18.482 18.485 ot_diis_step 116 11.6 0.020 0.020 17.766 17.769 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 17.460 17.650 apply_single 127 13.6 0.001 0.001 17.460 17.650 sum_up_and_integrate 127 10.3 0.180 0.189 16.255 16.301 integrate_v_rspace 127 11.3 0.004 0.004 16.075 16.129 cp_fm_diag_elpa 82 13.4 0.000 0.000 14.461 14.472 cp_fm_diag_elpa_base 82 14.4 14.197 14.293 14.452 14.464 make_images_data 4970 15.5 0.054 0.063 10.906 13.142 pw_transfer 1535 11.6 0.085 0.094 12.600 12.617 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 12.379 12.403 hybrid_alltoall_any 5155 16.4 0.837 3.759 9.967 11.766 multiply_cannon_metrocomm1 9940 15.5 0.030 0.031 7.765 11.524 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 11.199 11.438 multiply_cannon_sync_h2d 9940 15.5 10.678 10.967 10.678 10.967 init_scf_run 11 5.9 0.000 0.001 10.856 10.856 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.856 10.856 fft_wrap_pw1pw2_140 519 13.2 1.900 1.939 10.756 10.785 density_rs2pw 127 9.7 0.006 0.006 9.167 10.074 fft3d_ps 1281 14.7 2.745 2.812 8.879 8.906 cp_fm_cholesky_decompose 22 10.9 8.259 8.360 8.259 8.360 grid_integrate_task_list 127 12.3 7.659 8.154 7.659 8.154 dbcsr_mm_accdrv_process 20590 16.1 3.204 4.323 7.249 7.923 mp_alltoall_d11v 2401 14.1 6.877 7.891 6.877 7.891 wfi_extrapolate 11 7.9 0.001 0.001 7.758 7.758 calculate_dm_sparse 127 9.5 0.001 0.001 6.608 6.792 dbcsr_complete_redistribute 393 12.7 2.126 2.213 6.097 6.470 mp_allgather_i34 2485 14.5 1.771 6.365 1.771 6.365 grid_collocate_task_list 127 9.7 5.440 6.136 5.440 6.136 rs_pw_transfer 1038 11.9 0.012 0.013 4.976 5.878 mp_alltoall_z22v 1281 16.7 5.526 5.662 5.526 5.662 potential_pw2rs 127 12.3 0.026 0.026 5.593 5.611 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.370 5.370 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.288 5.316 make_images_sizes 4970 15.5 0.007 0.007 2.388 4.742 mp_alltoall_i44 4970 16.5 2.382 4.736 2.382 4.736 mp_sum_l 7884 12.9 3.349 4.498 3.349 4.498 copy_fm_to_dbcsr 208 11.6 0.001 0.002 4.080 4.457 rs_scatter_matrices 138 9.7 1.562 1.937 4.284 4.424 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.297 4.393 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.113 4.162 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=196.256000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1425.272727, yerr=16.679279 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1668dc7f1dfc939e0dca246397b4db0b42a8bbc0_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963542011904 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019182452736 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019182452736 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796564E+12 0.0% 0.0% 100.0% flops max/rank 11.606412E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499488 0.0% 0.0% 100.0% number of processed stacks 1947808 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.6 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 3.138208E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 99400 MPI messages size (bytes): total size 1.127422E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.342272E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 44768 34745614336 4194304 < size <= 16777216 43984 376564613120 16777216 < size 10032 716108490000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3995 59236. MP_Allreduce 11065 1504. MP_Sync 86 MP_Alltoall 1700 36954382. MP_SendRecv 1778 218624. MP_ISendRecv 1778 218624. MP_Wait 9728 MP_ISend 6360 1080477. MP_IRecv 6360 1080477. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.055 0.084 297.626 297.633 qs_mol_dyn_low 1 2.0 0.004 0.014 296.944 296.966 qs_forces 11 3.9 0.008 0.011 295.973 295.978 qs_energies 11 4.9 0.002 0.002 286.464 286.477 scf_env_do_scf 11 5.9 0.001 0.001 263.609 263.616 velocity_verlet 10 3.0 0.004 0.004 214.240 214.356 scf_env_do_scf_inner_loop 116 6.6 0.004 0.009 138.384 138.386 init_scf_loop 11 6.9 0.000 0.000 124.953 124.954 prepare_preconditioner 11 7.9 0.000 0.000 119.879 119.903 make_preconditioner 11 8.9 0.000 0.000 119.879 119.903 make_full_inverse_cholesky 11 9.9 0.038 0.039 95.831 117.039 qs_scf_new_mos 116 7.6 0.001 0.001 89.145 89.203 qs_scf_loop_do_ot 116 8.6 0.001 0.001 89.144 89.202 ot_scf_mini 116 9.6 0.003 0.004 84.288 84.324 dbcsr_multiply_generic 2485 12.5 0.211 0.223 81.587 82.109 cp_fm_upper_to_full 104 14.8 53.263 76.300 53.263 76.300 multiply_cannon 2485 13.5 0.675 0.722 58.794 59.664 multiply_cannon_loop 2485 14.5 1.032 1.039 54.862 56.263 ot_mini 116 10.6 0.001 0.001 44.018 44.056 dbcsr_complete_redistribute 393 12.7 4.009 4.058 30.001 42.833 copy_fm_to_dbcsr 208 11.6 0.001 0.001 26.518 39.330 rebuild_ks_matrix 127 8.3 0.001 0.001 37.193 37.224 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.017 37.192 37.224 transfer_fm_to_dbcsr 11 9.9 0.030 0.030 24.005 36.730 mp_alltoall_i22 712 14.1 21.784 34.730 21.784 34.730 qs_ks_update_qs_env 127 7.6 0.001 0.001 34.101 34.141 cp_fm_cholesky_invert 11 10.9 33.342 33.349 33.342 33.349 mp_waitall_1 102768 16.8 28.053 32.411 28.053 32.411 qs_ot_get_p 127 10.4 0.002 0.002 25.336 25.393 qs_ot_get_derivative 116 11.6 0.002 0.002 24.502 24.535 qs_ot_p2m_diag 82 11.4 0.869 0.875 21.365 21.396 multiply_cannon_metrocomm3 9940 15.5 0.025 0.027 19.130 20.527 qs_rho_update_rho_low 127 7.7 0.001 0.001 20.296 20.306 calculate_rho_elec 127 8.7 0.477 0.478 20.295 20.305 make_m2s 4970 13.5 0.074 0.077 18.232 19.648 cp_dbcsr_syevd 82 12.4 0.006 0.006 19.577 19.579 ot_diis_step 116 11.6 0.022 0.022 19.486 19.487 sum_up_and_integrate 127 10.3 0.319 0.320 19.312 19.394 make_images 4970 14.5 3.042 3.237 17.757 19.177 integrate_v_rspace 127 11.3 0.004 0.004 18.993 19.075 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 18.837 18.942 apply_single 127 13.6 0.001 0.001 18.837 18.942 multiply_cannon_multrec 9940 15.5 10.241 12.050 17.963 18.047 cp_fm_diag_elpa 82 13.4 0.000 0.000 16.427 16.428 cp_fm_diag_elpa_base 82 14.4 12.052 13.652 16.424 16.424 multiply_cannon_sync_h2d 9940 15.5 14.205 14.221 14.205 14.221 init_scf_run 11 5.9 0.000 0.001 12.476 12.477 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.476 12.477 hybrid_alltoall_any 5155 16.4 1.300 3.023 10.331 12.440 make_images_data 4970 15.5 0.062 0.067 10.168 12.288 pw_transfer 1535 11.6 0.092 0.093 11.376 11.381 fft_wrap_pw1pw2 1281 12.7 0.011 0.011 11.143 11.151 fft_wrap_pw1pw2_140 519 13.2 3.091 3.146 9.928 9.938 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.674 9.708 dbcsr_mm_accdrv_process 20590 16.0 4.181 5.935 7.475 9.354 wfi_extrapolate 11 7.9 0.001 0.001 9.265 9.265 cp_fm_cholesky_decompose 22 10.9 9.041 9.062 9.041 9.062 grid_integrate_task_list 127 12.3 8.525 8.736 8.525 8.736 mp_alltoall_d11v 2401 14.1 7.888 8.580 7.888 8.580 qs_energies_init_hamiltonians 11 5.9 0.008 0.012 8.203 8.206 density_rs2pw 127 9.7 0.005 0.006 7.904 8.057 calculate_dm_sparse 127 9.5 0.001 0.001 6.739 6.801 grid_collocate_task_list 127 9.7 6.391 6.477 6.391 6.477 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.327 6.368 fft3d_ps 1281 14.7 2.831 2.844 6.209 6.247 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.142 6.236 rs_scatter_matrices 138 9.7 3.916 4.782 5.985 6.205 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=297.633000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2846.818182, yerr=160.543292 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1668dc7f1dfc939e0dca246397b4db0b42a8bbc0_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 1.259618E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255646. MP_Allreduce 3139 6114. MP_Sync 4 MP_Alltoall 54 15448269. 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 0.020 0.036 85.182 85.183 qs_energies 1 2.0 0.000 0.000 84.712 84.722 ls_scf 1 3.0 0.000 0.000 83.818 83.828 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.686 72.832 multiply_cannon 111 7.7 0.017 0.020 56.008 57.122 multiply_cannon_loop 111 8.7 0.224 0.241 52.598 53.804 ls_scf_main 1 4.0 0.000 0.000 52.351 52.352 density_matrix_trs4 2 5.0 0.002 0.003 46.843 46.926 ls_scf_init_scf 1 4.0 0.000 0.000 28.439 28.440 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.316 27.355 mp_waitall_1 11031 10.9 22.476 25.890 22.476 25.890 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 25.131 25.152 multiply_cannon_multrec 2664 9.7 8.148 8.866 15.623 17.371 multiply_cannon_sync_h2d 2664 9.7 13.563 14.729 13.563 14.729 make_m2s 222 7.7 0.008 0.011 13.090 13.568 make_images 222 8.7 0.098 0.108 13.068 13.549 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.712 12.157 make_images_data 222 9.7 0.004 0.005 7.662 8.320 dbcsr_mm_accdrv_process 4760 10.4 0.591 0.731 7.089 8.095 hybrid_alltoall_any 227 10.6 0.216 1.850 6.591 8.065 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.437 7.985 dbcsr_mm_accdrv_process_sort 4760 11.4 6.299 7.175 6.299 7.175 calculate_norms 4752 9.8 5.524 6.122 5.524 6.122 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.993 5.138 mp_sum_l 887 5.1 3.049 4.232 3.049 4.232 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.086 3.778 mp_irecv_dv 6231 10.9 2.068 3.754 2.068 3.754 make_images_sizes 222 9.7 0.000 0.000 0.763 3.633 mp_alltoall_i44 222 10.7 0.763 3.633 0.763 3.633 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.287 3.287 arnoldi_extremal 4 6.8 0.000 0.000 3.200 3.221 arnoldi_normal_ev 4 7.8 0.001 0.003 3.200 3.221 build_subspace 16 8.4 0.009 0.012 3.109 3.111 ls_scf_post 1 4.0 0.000 0.000 3.027 3.037 ls_scf_store_result 1 5.0 0.000 0.000 2.853 2.889 dbcsr_special_finalize 555 9.7 0.005 0.006 2.313 2.797 dbcsr_merge_single_wm 555 10.7 0.469 0.610 2.305 2.788 make_images_pack 222 9.7 2.205 2.617 2.207 2.619 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.319 2.558 dbcsr_sort_data 658 11.4 2.094 2.511 2.094 2.511 dbcsr_matrix_vector_mult_local 304 10.0 2.068 2.467 2.070 2.469 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.307 2.385 buffer_matrices_ensure_size 222 8.7 1.765 2.106 1.765 2.106 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.771 1.772 rebuild_ks_matrix 3 7.3 0.000 0.000 1.761 1.762 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.761 1.762 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.183000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1140.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1668dc7f1dfc939e0dca246397b4db0b42a8bbc0_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 2.175271E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266673. MP_Allreduce 3138 10075. MP_Sync 4 MP_Alltoall 47 15335933. MP_SendRecv 141 57600. MP_ISendRecv 141 57600. MP_Wait 687 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.071 0.150 90.611 90.614 qs_energies 1 2.0 0.000 0.000 89.853 89.892 ls_scf 1 3.0 0.000 0.003 88.515 88.529 dbcsr_multiply_generic 111 6.7 0.015 0.016 74.696 75.080 multiply_cannon 111 7.7 0.027 0.040 52.814 57.652 ls_scf_main 1 4.0 0.001 0.005 55.022 55.028 multiply_cannon_loop 111 8.7 0.135 0.145 49.979 53.781 density_matrix_trs4 2 5.0 0.003 0.006 49.348 49.574 mp_waitall_1 9105 10.9 20.877 31.030 20.877 31.030 ls_scf_init_scf 1 4.0 0.000 0.002 29.815 29.818 ls_scf_init_matrix_S 1 5.0 0.000 0.001 28.557 28.676 multiply_cannon_multrec 1332 9.7 13.277 16.557 22.627 26.992 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.205 26.216 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.579 21.327 make_m2s 222 7.7 0.006 0.008 15.170 15.782 make_images 222 8.7 1.370 1.699 15.140 15.752 dbcsr_mm_accdrv_process 4041 10.4 0.323 0.534 8.945 10.551 dbcsr_mm_accdrv_process_sort 4041 11.4 8.467 10.016 8.467 10.016 make_images_data 222 9.7 0.004 0.005 8.818 9.720 hybrid_alltoall_any 227 10.6 0.542 2.547 8.126 9.547 mp_sum_l 887 5.1 5.248 8.779 5.248 8.779 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.188 8.150 mp_irecv_dv 3311 11.0 3.168 8.102 3.168 8.102 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.004 7.072 calculate_norms 2376 9.8 6.052 6.605 6.052 6.605 multiply_cannon_sync_h2d 1332 9.7 4.915 6.243 4.915 6.243 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.896 5.095 arnoldi_extremal 4 6.8 0.000 0.000 4.588 4.600 arnoldi_normal_ev 4 7.8 0.001 0.005 4.588 4.600 build_subspace 16 8.4 0.014 0.021 4.338 4.341 ls_scf_post 1 4.0 0.004 0.018 3.678 3.691 ls_scf_store_result 1 5.0 0.000 0.000 3.404 3.495 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.124 3.354 dbcsr_matrix_vector_mult_local 304 10.0 2.743 3.223 2.745 3.225 mp_allgather_i34 111 8.7 0.829 3.063 0.829 3.063 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.618 2.712 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.129 2.492 dbcsr_data_new 4174 10.1 2.112 2.399 2.112 2.399 make_images_pack 222 9.7 1.819 2.137 1.822 2.139 dbcsr_sort_data 436 11.2 1.836 2.058 1.836 2.058 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.903 1.907 rebuild_ks_matrix 3 7.3 0.000 0.000 1.887 1.892 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.007 1.887 1.892 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=90.614000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1774.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1668dc7f1dfc939e0dca246397b4db0b42a8bbc0_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 2.909549E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265448. MP_Allreduce 3138 10896. MP_Sync 4 MP_Alltoall 47 23526250. MP_SendRecv 93 57600. MP_ISendRecv 93 57600. MP_Wait 639 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.045 0.087 97.789 97.790 qs_energies 1 2.0 0.001 0.007 97.156 97.168 ls_scf 1 3.0 0.000 0.001 95.733 95.740 dbcsr_multiply_generic 111 6.7 0.016 0.018 78.095 78.408 ls_scf_main 1 4.0 0.000 0.005 59.835 59.839 multiply_cannon 111 7.7 0.033 0.066 53.193 58.168 multiply_cannon_loop 111 8.7 0.116 0.129 50.228 54.200 density_matrix_trs4 2 5.0 0.003 0.015 53.454 53.643 mp_waitall_1 7281 11.0 25.556 35.437 25.556 35.437 ls_scf_init_scf 1 4.0 0.000 0.001 32.210 32.213 ls_scf_init_matrix_S 1 5.0 0.000 0.001 30.565 30.663 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 28.110 28.121 multiply_cannon_multrec 888 9.7 12.647 15.018 21.282 24.297 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.907 23.420 make_m2s 222 7.7 0.006 0.008 18.024 18.932 make_images 222 8.7 1.583 1.858 17.986 18.895 make_images_data 222 9.7 0.004 0.004 10.485 12.414 hybrid_alltoall_any 227 10.6 0.638 2.932 9.701 11.924 mp_sum_l 887 5.1 5.190 9.505 5.190 9.505 dbcsr_mm_accdrv_process 3754 10.4 0.296 0.492 8.162 9.444 dbcsr_mm_accdrv_process_sort 3754 11.4 7.730 8.952 7.730 8.952 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 4.794 8.795 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.055 7.548 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.472 7.171 multiply_cannon_sync_h2d 888 9.7 6.008 7.159 6.008 7.159 mp_irecv_dv 2335 11.1 2.457 7.127 2.457 7.127 arnoldi_extremal 4 6.8 0.000 0.000 6.424 6.437 arnoldi_normal_ev 4 7.8 0.001 0.005 6.424 6.437 build_subspace 16 8.4 0.014 0.020 6.106 6.113 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.961 5.185 dbcsr_matrix_vector_mult 304 9.0 0.010 0.020 4.551 5.017 calculate_norms 1584 9.8 4.303 4.694 4.303 4.694 mp_allgather_i34 111 8.7 1.053 4.074 1.053 4.074 ls_scf_post 1 4.0 0.001 0.006 3.688 3.696 make_images_sizes 222 9.7 0.000 0.000 1.774 3.655 mp_alltoall_i44 222 10.7 1.774 3.654 1.774 3.654 dbcsr_matrix_vector_mult_local 304 10.0 2.995 3.582 2.997 3.584 ls_scf_store_result 1 5.0 0.000 0.000 3.414 3.508 ls_scf_dm_to_ks 2 5.0 0.000 0.007 3.068 3.164 dbcsr_data_new 4116 9.9 2.112 2.443 2.112 2.443 mp_sum_dv 2129 10.8 2.130 2.403 2.130 2.403 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.218 2.222 rebuild_ks_matrix 3 7.3 0.000 0.000 2.199 2.202 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.004 2.199 2.202 dbcsr_sort_data 325 11.1 1.820 2.127 1.820 2.127 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=97.790000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2277.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1668dc7f1dfc939e0dca246397b4db0b42a8bbc0_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 3.322216E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266673. MP_Allreduce 3138 13030. MP_Sync 4 MP_Alltoall 47 30278988. MP_SendRecv 69 86400. MP_ISendRecv 69 86400. MP_Wait 531 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.089 0.174 97.395 97.396 qs_energies 1 2.0 0.000 0.000 96.477 96.486 ls_scf 1 3.0 0.000 0.003 94.816 94.822 dbcsr_multiply_generic 111 6.7 0.018 0.023 78.480 78.762 ls_scf_main 1 4.0 0.000 0.002 58.991 58.992 multiply_cannon 111 7.7 0.051 0.110 51.615 56.240 density_matrix_trs4 2 5.0 0.004 0.016 52.822 52.942 multiply_cannon_loop 111 8.7 0.152 0.165 46.543 49.875 ls_scf_init_scf 1 4.0 0.000 0.002 32.589 32.591 ls_scf_init_matrix_S 1 5.0 0.000 0.001 31.331 31.399 mp_waitall_1 6369 11.0 22.702 29.716 22.702 29.716 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 28.860 28.873 multiply_cannon_multrec 1332 9.7 14.115 17.218 22.079 25.512 make_m2s 222 7.7 0.007 0.008 21.243 22.634 make_images 222 8.7 3.148 3.603 21.192 22.584 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.114 17.273 make_images_data 222 9.7 0.004 0.006 11.893 13.621 hybrid_alltoall_any 227 10.6 0.798 3.780 11.194 13.097 dbcsr_mm_accdrv_process 3641 10.4 0.276 0.483 7.604 9.151 dbcsr_mm_accdrv_process_sort 3641 11.4 7.167 8.659 7.167 8.659 mp_sum_l 887 5.1 4.085 7.271 4.085 7.271 multiply_cannon_sync_h2d 1332 9.7 5.413 6.089 5.413 6.089 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.111 6.026 mp_irecv_dv 3229 10.9 2.086 5.946 2.086 5.946 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.527 5.878 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.061 5.604 arnoldi_extremal 4 6.8 0.000 0.000 5.218 5.234 arnoldi_normal_ev 4 7.8 0.001 0.005 5.218 5.233 build_subspace 16 8.4 0.014 0.021 4.881 4.890 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.542 4.768 mp_allgather_i34 111 8.7 2.190 4.593 2.190 4.593 calculate_norms 2376 9.8 4.200 4.585 4.200 4.585 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.586 3.882 dbcsr_matrix_vector_mult_local 304 10.0 3.192 3.696 3.194 3.698 dbcsr_sort_data 658 11.4 3.074 3.416 3.074 3.416 ls_scf_post 1 4.0 0.000 0.001 3.236 3.242 dbcsr_special_finalize 555 9.7 0.006 0.007 2.822 3.159 dbcsr_merge_single_wm 555 10.7 0.539 0.660 2.814 3.151 ls_scf_dm_to_ks 2 5.0 0.000 0.001 3.046 3.100 ls_scf_store_result 1 5.0 0.000 0.000 2.971 3.035 dbcsr_data_release 10477 10.7 1.583 2.488 1.583 2.488 dbcsr_finalize 304 7.8 0.049 0.061 1.803 2.024 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.954 1.956 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.396000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2724.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1668dc7f1dfc939e0dca246397b4db0b42a8bbc0_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 4.764959E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265536. MP_Allreduce 3129 15263. MP_Sync 4 MP_Alltoall 47 46208988. MP_SendRecv 45 115200. MP_ISendRecv 45 115200. MP_Wait 528 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.039 0.057 91.743 91.744 qs_energies 1 2.0 0.000 0.000 90.951 90.956 ls_scf 1 3.0 0.000 0.000 89.021 89.026 dbcsr_multiply_generic 111 6.7 0.017 0.018 70.518 70.733 ls_scf_main 1 4.0 0.000 0.000 56.405 56.406 multiply_cannon 111 7.7 0.083 0.118 52.351 56.014 multiply_cannon_loop 111 8.7 0.089 0.095 49.809 51.059 density_matrix_trs4 2 5.0 0.002 0.003 49.331 49.409 ls_scf_init_scf 1 4.0 0.000 0.000 29.228 29.230 mp_waitall_1 5436 11.0 24.157 28.634 24.157 28.634 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.945 27.973 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 25.844 25.862 multiply_cannon_multrec 444 9.7 13.822 16.304 20.972 22.545 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.578 15.350 make_m2s 222 7.7 0.005 0.005 13.567 14.465 make_images 222 8.7 2.043 2.478 13.500 14.396 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.095 13.794 hybrid_alltoall_any 227 10.6 0.801 3.830 8.107 9.704 make_images_data 222 9.7 0.003 0.004 8.298 9.575 dbcsr_mm_accdrv_process 3003 10.4 0.344 0.425 6.851 7.989 multiply_cannon_sync_h2d 444 9.7 6.667 7.932 6.667 7.932 dbcsr_mm_accdrv_process_sort 3003 11.4 6.483 7.564 6.483 7.564 arnoldi_extremal 4 6.8 0.000 0.000 5.829 5.846 arnoldi_normal_ev 4 7.8 0.002 0.005 5.828 5.845 build_subspace 16 8.4 0.015 0.020 5.429 5.441 mp_sum_l 887 5.1 2.603 4.841 2.603 4.841 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.453 4.634 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.173 4.388 dbcsr_matrix_vector_mult_local 304 10.0 3.706 4.184 3.708 4.186 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.545 4.039 mp_irecv_dv 1241 11.2 1.533 4.013 1.533 4.013 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.832 3.864 calculate_norms 792 9.8 3.612 3.752 3.612 3.752 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.589 3.662 mp_allgather_i34 111 8.7 1.130 3.615 1.130 3.615 ls_scf_post 1 4.0 0.000 0.000 3.388 3.393 make_images_sizes 222 9.7 0.000 0.000 0.875 3.330 mp_alltoall_i44 222 10.7 0.874 3.329 0.874 3.329 ls_scf_store_result 1 5.0 0.000 0.000 3.178 3.211 dbcsr_finalize 304 7.8 0.062 0.077 2.193 2.292 dbcsr_data_new 4608 9.7 1.796 2.158 1.796 2.158 dbcsr_merge_all 275 8.9 0.478 0.522 2.051 2.132 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.091 2.092 rebuild_ks_matrix 3 7.3 0.000 0.000 2.059 2.060 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.059 2.060 qs_energies_init_hamiltonians 1 3.0 0.001 0.002 1.914 1.914 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=91.744000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3734.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1668dc7f1dfc939e0dca246397b4db0b42a8bbc0_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 8.932110E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284089. MP_Allreduce 3123 21388. MP_Sync 4 MP_Alltoall 47 88727262. MP_SendRecv 42 732600. MP_ISendRecv 42 732600. MP_Wait 267 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.076 0.097 109.524 109.525 qs_energies 1 2.0 0.000 0.000 108.129 108.137 ls_scf 1 3.0 0.000 0.000 105.131 105.137 dbcsr_multiply_generic 111 6.7 0.024 0.028 76.825 76.942 ls_scf_main 1 4.0 0.000 0.000 65.520 65.521 density_matrix_trs4 2 5.0 0.002 0.003 56.335 56.402 multiply_cannon 111 7.7 0.132 0.241 49.352 51.102 multiply_cannon_loop 111 8.7 0.098 0.099 46.389 47.285 ls_scf_init_scf 1 4.0 0.000 0.000 35.792 35.792 ls_scf_init_matrix_S 1 5.0 0.000 0.000 34.031 34.060 matrix_sqrt_Newton_Schulz 2 6.5 0.009 0.021 31.181 31.190 mp_waitall_1 4527 11.1 23.410 27.072 23.410 27.072 make_m2s 222 7.7 0.005 0.005 23.525 24.550 make_images 222 8.7 3.576 3.869 23.418 24.439 multiply_cannon_multrec 444 9.7 17.811 18.469 22.499 23.047 hybrid_alltoall_any 227 10.6 1.656 3.619 13.413 16.606 make_images_data 222 9.7 0.003 0.004 13.691 15.979 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 11.110 11.724 multiply_cannon_sync_h2d 444 9.7 8.792 8.848 8.792 8.848 arnoldi_extremal 4 6.8 0.000 0.000 8.732 8.760 arnoldi_normal_ev 4 7.8 0.004 0.012 8.732 8.760 build_subspace 16 8.4 0.025 0.036 8.115 8.126 dbcsr_matrix_vector_mult 304 9.0 0.017 0.031 6.181 6.358 dbcsr_matrix_vector_mult_local 304 10.0 5.058 5.366 5.061 5.369 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.026 5.300 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.185 5.274 dbcsr_mm_accdrv_process 1814 10.4 0.327 0.372 4.516 4.677 dbcsr_mm_accdrv_process_sort 1814 11.4 4.166 4.305 4.166 4.305 make_images_sizes 222 9.7 0.000 0.000 1.698 4.098 mp_alltoall_i44 222 10.7 1.697 4.098 1.697 4.098 ls_scf_post 1 4.0 0.000 0.000 3.818 3.826 ls_scf_store_result 1 5.0 0.000 0.000 3.536 3.570 calculate_norms 792 9.8 3.231 3.278 3.231 3.278 mp_allgather_i34 111 8.7 1.040 3.178 1.040 3.178 dbcsr_finalize 304 7.8 0.082 0.089 3.064 3.123 qs_energies_init_hamiltonians 1 3.0 0.023 0.037 2.968 2.968 dbcsr_merge_all 275 8.9 0.883 0.918 2.849 2.903 dbcsr_complete_redistribute 5 7.6 1.427 1.468 2.747 2.872 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.548 2.549 matrix_ls_to_qs 2 6.0 0.000 0.000 2.392 2.528 dbcsr_sort_data 325 11.1 2.441 2.503 2.441 2.503 rebuild_ks_matrix 3 7.3 0.000 0.000 2.480 2.482 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.480 2.482 dbcsr_new_transposed 4 7.5 0.246 0.260 2.452 2.464 dbcsr_data_new 6591 9.6 1.878 2.338 1.878 2.338 mp_alltoall_d11v 48 9.2 2.232 2.311 2.232 2.311 dbcsr_frobenius_norm 74 6.6 2.059 2.140 2.208 2.239 mp_sum_l 887 5.1 1.556 2.236 1.556 2.236 dbcsr_add_d 103 6.2 0.000 0.000 2.126 2.206 dbcsr_add_anytype 103 7.2 0.858 0.890 2.125 2.205 dbcsr_redistribute 4 8.5 1.361 1.395 2.179 2.200 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=109.525000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6998.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1668dc7f1dfc939e0dca246397b4db0b42a8bbc0_performance_tests/27/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 7009386627072 0.0% 0.0% 100.0% flops 9 x 9 x 32 7335108845568 0.0% 0.0% 100.0% flops 9 x 22 x 32 9866241589248 0.0% 0.0% 100.0% flops 22 x 9 x 32 9884108906496 0.0% 0.0% 100.0% flops 22 x 22 x 32 13354440523776 0.0% 0.0% 100.0% flops 32 x 32 x 9 20607185977344 0.0% 0.0% 100.0% flops 32 x 32 x 22 25186560638976 0.0% 0.0% 100.0% flops 9 x 32 x 32 28458319085568 0.0% 0.0% 100.0% flops 22 x 32 x 32 34782389993472 0.0% 0.0% 100.0% flops 9 x 32 x 9 42881542373376 0.0% 0.0% 100.0% flops 22 x 32 x 9 55680402235392 0.0% 0.0% 100.0% flops 9 x 32 x 22 55680402235392 0.0% 0.0% 100.0% flops 22 x 32 x 22 72328573419520 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 383.054662E+12 0.0% 0.0% 100.0% flops max/rank 733.641090E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 26899403712 0.0% 0.0% 100.0% number of processed stacks 118860288 0.0% 0.0% 100.0% average stack size 0.0 0.0 226.3 marketing flops 780.439111E+12 ------------------------------------------------------------------------------- # multiplications 1445 max memory usage/rank 592.777216E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 102097920 MPI messages size (bytes): total size 37.227590E+12 min size 0.000000E+00 max size 4.551360E+06 average size 364.626312E+03 MPI breakdown and total messages size (bytes): size <= 128 731472 0 128 < size <= 8192 11922720 97670922240 8192 < size <= 32768 24718992 614677610496 32768 < size <= 131072 20000256 1970081366016 131072 < size <= 4194304 42515668 24886801223040 4194304 < size <= 16777216 2208812 9656099886720 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4640 78072. MP_Allreduce 13232 2081. MP_Sync 1064 MP_Alltoall 2588 4037197. MP_SendRecv 168740 11136. MP_ISendRecv 92040 11136. MP_Wait 102830 MP_comm_split 40 MP_ISend 26090 85106. MP_IRecv 37890 59644. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.083 0.188 229.585 229.586 qs_mol_dyn_low 1 2.0 0.004 0.027 228.244 228.266 qs_forces 5 3.8 0.007 0.077 228.116 228.128 qs_energies 5 4.8 0.016 0.030 225.126 225.146 scf_env_do_scf 5 5.8 0.000 0.001 210.000 210.003 scf_env_do_scf_inner_loop 105 6.6 0.003 0.016 183.045 183.056 qs_scf_new_mos 105 7.6 0.000 0.001 142.012 142.209 qs_scf_loop_do_ot 105 8.6 0.001 0.002 142.012 142.208 dbcsr_multiply_generic 1445 12.2 0.128 0.142 132.886 133.271 ot_scf_mini 105 9.6 0.003 0.005 132.182 132.324 multiply_cannon 1445 13.2 0.273 0.285 114.206 116.244 multiply_cannon_loop 1445 14.2 2.839 2.987 112.560 113.589 velocity_verlet 4 3.0 0.001 0.005 106.463 106.463 ot_mini 105 10.6 0.001 0.021 59.849 59.960 multiply_cannon_multrec 69360 15.2 29.728 34.998 39.715 44.598 qs_ot_get_p 112 10.4 0.001 0.001 42.513 42.801 mp_waitall_1 488190 16.1 34.059 40.507 34.059 40.507 qs_ot_get_derivative 55 11.6 0.001 0.001 38.084 38.197 multiply_cannon_sync_h2d 69360 15.2 29.253 33.146 29.253 33.146 multiply_cannon_metrocomm3 69360 15.2 0.205 0.216 25.377 31.993 qs_ot_p2m_diag 40 11.0 0.020 0.033 31.608 31.681 rebuild_ks_matrix 110 8.4 0.000 0.000 29.240 29.434 qs_ks_build_kohn_sham_matrix 110 9.4 0.012 0.024 29.240 29.434 cp_dbcsr_syevd 40 12.0 0.002 0.002 28.337 28.338 qs_ks_update_qs_env 112 7.6 0.001 0.001 26.885 27.050 init_scf_loop 7 6.6 0.001 0.011 26.925 26.927 cp_fm_syevd 40 13.0 0.000 0.000 23.235 23.387 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 23.004 23.210 apply_single 62 13.6 0.000 0.000 23.004 23.210 prepare_preconditioner 7 7.6 0.000 0.000 22.045 22.081 make_preconditioner 7 8.6 0.000 0.011 22.044 22.081 ot_new_cg_direction 55 11.6 0.003 0.070 21.024 21.025 cp_fm_redistribute_end 40 14.0 9.284 18.527 9.288 18.528 cp_fm_syevd_base 40 14.0 9.232 18.477 9.232 18.477 qs_rho_update_rho_low 110 7.6 0.001 0.003 17.538 17.887 calculate_rho_elec 110 8.6 0.030 0.036 17.537 17.886 make_full_inverse_cholesky 7 9.6 0.000 0.000 14.784 14.866 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.627 14.754 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 13.712 13.819 rs_pw_transfer 690 11.5 0.011 0.012 11.801 13.146 density_rs2pw 110 9.6 0.006 0.007 11.565 12.930 mp_sum_l 4764 12.2 11.881 12.630 11.881 12.630 pw_transfer 1645 12.4 0.083 0.102 12.157 12.339 init_scf_run 5 5.8 0.000 0.001 12.334 12.335 scf_env_initial_rho_setup 5 6.8 0.017 0.034 12.334 12.335 fft_wrap_pw1pw2 1425 13.5 0.013 0.016 12.015 12.202 calculate_dm_sparse 110 9.5 0.000 0.001 11.505 11.678 dbcsr_mm_accdrv_process 154766 15.8 6.143 6.324 9.854 10.739 qs_vxc_create 110 10.4 0.003 0.013 10.485 10.524 cp_fm_cholesky_invert 7 10.6 10.444 10.452 10.444 10.452 fft_wrap_pw1pw2_240 915 15.0 1.194 1.267 10.268 10.432 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 10.281 10.360 check_diag 80 13.5 8.594 8.886 9.423 9.575 sum_up_and_integrate 60 10.3 0.028 0.031 8.771 8.781 integrate_v_rspace 60 11.3 0.002 0.003 8.743 8.754 fft3d_pb 915 16.0 2.419 2.612 8.328 8.483 acc_transpose_blocks 69360 15.2 0.363 0.382 7.869 8.200 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.001 7.694 7.711 xc_rho_set_and_dset_create 110 12.4 0.078 0.104 7.368 7.607 calculate_first_density_matrix 1 7.0 0.001 0.019 7.517 7.544 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 7.181 7.233 xc_vxc_pw_create 60 11.3 0.039 0.049 7.001 7.040 make_full_single_inverse 7 9.6 0.002 0.025 6.977 7.013 make_m2s 2890 13.2 0.080 0.089 6.342 6.931 make_images 2890 14.2 0.238 0.260 6.233 6.823 multiply_cannon_metrocomm1 69360 15.2 0.097 0.106 4.395 6.539 xc_pw_derive 510 13.4 0.005 0.006 6.230 6.300 mp_alltoall_z22v 2340 17.7 5.353 5.678 5.353 5.678 mp_waitany 7680 13.5 4.398 5.616 4.398 5.616 acc_transpose_blocks_kernels 69360 16.2 0.842 0.886 5.130 5.336 potential_pw2rs 60 12.3 0.003 0.003 4.954 4.976 multiply_cannon_metrocomm4 67915 15.2 0.183 0.200 2.024 4.781 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=229.586000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=562.400000, yerr=3.382307 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 1668dc7f1dfc939e0dca246397b4db0b42a8bbc0 Summary: empty Status: OK