=== 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: 04dd581a73886059b5d5fb12035b1998b2b62f5a ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, Cray-FFTW 3.3.8.10, # COSMA 2.6.2, ELPA 2022.11.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.1.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.1, # SIRIUS 7.4.3, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Author: Matthias Krack (03.02.2023) # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.1.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.1 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta_prefixed_scalapack.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/04dd581a73886059b5d5fb12035b1998b2b62f5a_performance_tests/01 job id: 46005531 --- 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/04dd581a73886059b5d5fb12035b1998b2b62f5a_performance_tests/02 job id: 46005533 --- 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/04dd581a73886059b5d5fb12035b1998b2b62f5a_performance_tests/03 job id: 46005535 --- 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/04dd581a73886059b5d5fb12035b1998b2b62f5a_performance_tests/04 job id: 46005537 --- 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/04dd581a73886059b5d5fb12035b1998b2b62f5a_performance_tests/05 job id: 46005538 --- 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/04dd581a73886059b5d5fb12035b1998b2b62f5a_performance_tests/06 job id: 46005539 --- 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/04dd581a73886059b5d5fb12035b1998b2b62f5a_performance_tests/07 job id: 46005540 --- 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/04dd581a73886059b5d5fb12035b1998b2b62f5a_performance_tests/08 job id: 46005541 --- 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/04dd581a73886059b5d5fb12035b1998b2b62f5a_performance_tests/09 job id: 46005542 --- 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/04dd581a73886059b5d5fb12035b1998b2b62f5a_performance_tests/10 job id: 46005543 --- 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/04dd581a73886059b5d5fb12035b1998b2b62f5a_performance_tests/11 job id: 46005544 --- 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/04dd581a73886059b5d5fb12035b1998b2b62f5a_performance_tests/12 job id: 46005545 --- 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/04dd581a73886059b5d5fb12035b1998b2b62f5a_performance_tests/13 job id: 46005546 --- 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/04dd581a73886059b5d5fb12035b1998b2b62f5a_performance_tests/14 job id: 46005547 --- 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/04dd581a73886059b5d5fb12035b1998b2b62f5a_performance_tests/15 job id: 46005548 --- 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/04dd581a73886059b5d5fb12035b1998b2b62f5a_performance_tests/16 job id: 46005549 --- 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/04dd581a73886059b5d5fb12035b1998b2b62f5a_performance_tests/17 job id: 46005550 --- 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/04dd581a73886059b5d5fb12035b1998b2b62f5a_performance_tests/18 job id: 46005551 --- 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/04dd581a73886059b5d5fb12035b1998b2b62f5a_performance_tests/19 job id: 46005552 --- 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/04dd581a73886059b5d5fb12035b1998b2b62f5a_performance_tests/20 job id: 46005553 --- 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/04dd581a73886059b5d5fb12035b1998b2b62f5a_performance_tests/21 job id: 46005554 --- 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/04dd581a73886059b5d5fb12035b1998b2b62f5a_performance_tests/22 job id: 46005555 --- 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/04dd581a73886059b5d5fb12035b1998b2b62f5a_performance_tests/23 job id: 46005556 --- 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/04dd581a73886059b5d5fb12035b1998b2b62f5a_performance_tests/24 job id: 46005557 --- 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/04dd581a73886059b5d5fb12035b1998b2b62f5a_performance_tests/25 job id: 46005558 --- 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/04dd581a73886059b5d5fb12035b1998b2b62f5a_performance_tests/26 job id: 46005560 --- 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/04dd581a73886059b5d5fb12035b1998b2b62f5a_performance_tests/27 job id: 46005562 --- 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/04dd581a73886059b5d5fb12035b1998b2b62f5a_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 344 9. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.022 0.034 135.811 135.812 farming_run 1 2.0 135.220 135.222 135.779 135.783 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.460105E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 230 1103589. MP_Allreduce 491 2254389. MP_Sync 25 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 10 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.006 0.025 116.219 116.219 qs_energies 1 2.0 0.000 0.000 116.018 116.020 mp2_main 1 3.0 0.000 0.000 113.688 113.690 mp2_gpw_main 1 4.0 0.019 0.023 112.558 112.560 mp2_ri_gpw_compute_in 1 5.0 0.172 0.174 93.368 93.872 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 55.474 55.978 mp2_eri_3c_integrate_gpw 272 7.0 0.154 0.170 41.811 47.251 get_2c_integrals 1 6.0 0.000 0.000 36.966 37.720 integrate_v_rspace 273 8.0 0.433 0.447 25.176 30.421 pw_transfer 6555 10.6 0.375 0.395 27.373 27.879 grid_integrate_task_list 273 9.0 20.971 26.722 20.971 26.722 fft_wrap_pw1pw2 5465 11.4 0.045 0.048 26.046 26.549 fft_wrap_pw1pw2_100 2178 12.4 1.175 1.415 23.569 24.070 compute_2c_integrals 1 7.0 0.002 0.003 19.625 19.625 rpa_ri_compute_en 1 5.0 0.001 0.002 19.082 19.322 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.824 19.316 mp2_eri_2c_integrate_gpw 1 9.0 2.382 2.436 18.821 19.314 cp_fm_cholesky_decompose 12 8.2 17.351 18.186 17.351 18.186 cholesky_decomp 1 7.0 0.000 0.000 16.185 16.942 fft3d_s 5443 13.4 16.141 16.590 16.164 16.613 ao_to_mo_and_store_B_mult_1 272 7.0 10.844 15.555 10.844 15.555 calculate_wavefunction 272 8.0 5.442 5.607 12.570 13.172 rpa_num_int 1 6.0 0.000 0.001 10.645 10.646 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.556 10.643 calc_mat_Q 8 8.0 0.000 0.000 9.350 9.446 contract_S_to_Q 8 9.0 0.000 0.000 8.771 8.865 calc_potential_gpw 544 9.5 0.005 0.006 8.246 8.611 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.183 8.493 parallel_gemm_fm 14 9.1 0.000 0.000 8.359 8.445 parallel_gemm_fm_cosma 14 10.1 8.359 8.445 8.359 8.445 potential_pw2rs 545 10.0 0.108 0.110 7.678 8.284 create_integ_mat 1 6.0 0.022 0.027 7.836 7.837 collocate_single_gaussian 272 10.0 0.040 0.042 7.438 7.722 array2fm 1 7.0 0.000 0.000 6.640 7.117 pw_scatter_s 2720 13.7 4.438 4.592 4.438 4.592 pw_gather_s 2722 13.2 3.884 4.216 3.884 4.216 array2fm_buffer_send 1 8.0 2.900 3.067 2.900 3.067 pw_poisson_solve 545 10.5 1.095 1.135 2.127 2.329 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.559890, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2731.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/04dd581a73886059b5d5fb12035b1998b2b62f5a_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 205321. MP_Allreduce 344 10. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.029 0.037 396.471 396.472 farming_run 1 2.0 395.649 395.658 396.433 396.436 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827128576 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788821 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.229431E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 743 386399. MP_Allreduce 1941 22272. MP_Sync 37 MP_Alltoall 77 8570545. 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 0.010 0.032 210.076 210.077 qs_energies 1 2.0 0.005 0.019 209.771 209.798 scf_env_do_scf 1 3.0 0.000 0.000 106.537 106.538 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.655 105.664 rebuild_ks_matrix 4 6.0 0.000 0.000 105.654 105.663 qs_ks_build_kohn_sham_matrix 4 7.0 0.055 0.065 105.654 105.663 hfx_ks_matrix 4 8.0 0.001 0.001 105.269 105.274 integrate_four_center 4 9.0 0.145 0.480 105.269 105.273 mp2_main 1 3.0 0.003 0.025 102.937 102.952 mp2_gpw_main 1 4.0 0.040 0.099 102.078 102.091 integrate_four_center_main 4 10.0 0.101 0.495 96.706 99.450 integrate_four_center_bin 266 11.0 96.605 99.439 96.605 99.439 init_scf_loop 1 4.0 0.000 0.000 92.229 92.229 mp2_ri_gpw_compute_in 1 5.0 0.072 0.101 74.946 76.018 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.012 54.491 55.565 mp2_eri_3c_integrate_gpw 91 7.0 0.143 0.159 42.201 47.216 integrate_v_rspace 95 8.0 0.395 0.567 28.578 33.408 pw_transfer 2240 10.6 0.144 0.164 29.934 30.483 fft_wrap_pw1pw2 1868 11.4 0.018 0.020 28.933 29.426 ao_to_mo_and_store_B_mult_1 91 7.0 10.604 29.261 10.604 29.261 grid_integrate_task_list 95 9.0 23.866 28.895 23.866 28.895 mp2_ri_gpw_compute_en 1 5.0 0.063 0.100 26.960 28.717 fft_wrap_pw1pw2_100 730 12.4 1.277 1.467 26.643 27.042 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.839 1.917 25.202 25.214 get_2c_integrals 1 6.0 0.007 0.060 20.339 20.377 compute_2c_integrals 1 7.0 0.006 0.015 19.309 19.325 compute_2c_integrals_loop_lm 1 8.0 0.002 0.009 18.873 19.189 mp2_eri_2c_integrate_gpw 1 9.0 1.740 1.880 18.871 19.188 fft3d_s 1823 13.4 18.441 18.793 18.454 18.806 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.307 14.307 calculate_wavefunction 91 8.0 2.015 2.050 9.744 9.935 mp2_ri_gpw_compute_en_expansio 172 7.0 0.558 0.584 8.870 9.348 potential_pw2rs 186 10.0 0.034 0.036 8.636 9.218 local_gemm 172 8.0 8.313 8.786 8.313 8.786 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.237 8.627 mp2_ri_gpw_compute_en_comm 22 7.0 0.503 0.519 8.112 8.585 calc_potential_gpw 182 9.5 0.002 0.002 7.910 8.200 collocate_single_gaussian 91 10.0 0.017 0.023 7.894 8.174 mp_sendrecv_dm3 2068 8.0 6.142 6.592 6.142 6.592 mp2_ri_gpw_compute_en_ener 172 7.0 6.337 6.436 6.337 6.436 mp_sync 37 10.5 3.229 6.096 3.229 6.096 pw_gather_s 912 13.2 4.919 5.443 4.919 5.443 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=102.079885, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1516.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/04dd581a73886059b5d5fb12035b1998b2b62f5a_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 452.882432E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62385. MP_Allreduce 10249 272. MP_Sync 530 MP_Alltoall 2083 31991. 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 0.015 0.032 52.676 52.678 qs_mol_dyn_low 1 2.0 0.003 0.005 52.397 52.406 qs_forces 11 3.9 0.003 0.008 52.257 52.259 qs_energies 11 4.9 0.002 0.007 50.803 50.822 scf_env_do_scf 11 5.9 0.000 0.001 44.840 44.841 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 42.874 42.874 qs_scf_new_mos 108 7.5 0.000 0.001 32.991 33.294 qs_scf_loop_do_ot 108 8.5 0.000 0.001 32.991 33.293 dbcsr_multiply_generic 2286 12.5 0.093 0.098 32.790 33.223 ot_scf_mini 108 9.5 0.002 0.003 31.372 31.568 multiply_cannon 2286 13.5 0.189 0.195 25.843 27.447 multiply_cannon_loop 2286 14.5 1.478 1.565 25.180 26.789 velocity_verlet 10 3.0 0.001 0.002 25.418 25.419 ot_mini 108 10.5 0.001 0.004 18.800 19.052 qs_ot_get_derivative 108 11.5 0.001 0.002 15.906 16.074 mp_waitall_1 245248 16.5 8.134 14.188 8.134 14.188 multiply_cannon_metrocomm3 54864 15.5 0.069 0.075 5.866 12.658 multiply_cannon_multrec 54864 15.5 4.266 6.669 7.751 11.334 qs_ot_get_p 119 10.4 0.001 0.001 7.986 8.263 rebuild_ks_matrix 119 8.3 0.000 0.000 7.754 7.885 qs_ks_build_kohn_sham_matrix 119 9.3 0.011 0.019 7.754 7.885 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.835 6.953 multiply_cannon_sync_h2d 54864 15.5 6.004 6.915 6.004 6.915 mp_sum_l 7207 12.9 5.130 6.661 5.130 6.661 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.667 6.101 qs_ot_p2m_diag 50 11.0 0.004 0.008 5.341 5.382 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.066 5.175 dbcsr_mm_accdrv_process 76910 16.1 1.175 1.866 3.405 4.802 init_scf_run 11 5.9 0.000 0.001 4.713 4.713 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.712 4.713 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.517 4.518 sum_up_and_integrate 119 10.3 0.012 0.015 4.437 4.445 integrate_v_rspace 119 11.3 0.003 0.003 4.424 4.434 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.329 4.330 cp_fm_redistribute_end 50 14.0 2.209 4.307 2.216 4.310 qs_rho_update_rho_low 119 7.7 0.001 0.002 4.062 4.214 calculate_rho_elec 119 8.7 0.012 0.018 4.061 4.214 cp_fm_diag_elpa_base 50 14.0 2.089 4.201 2.093 4.211 apply_preconditioner_dbcsr 119 12.6 0.000 0.001 2.832 3.055 apply_single 119 13.6 0.000 0.000 2.832 3.055 calculate_dm_sparse 119 9.5 0.000 0.001 2.888 3.044 jit_kernel_multiply 13 15.8 2.168 2.849 2.168 2.849 rs_pw_transfer 974 11.9 0.012 0.013 2.688 2.793 multiply_cannon_metrocomm1 54864 15.5 0.053 0.058 1.660 2.754 ot_diis_step 108 11.5 0.006 0.009 2.627 2.628 calculate_first_density_matrix 1 7.0 0.001 0.002 2.611 2.612 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.399 2.474 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.311 2.313 density_rs2pw 119 9.7 0.004 0.005 2.129 2.260 acc_transpose_blocks 54864 15.5 0.228 0.256 1.708 2.156 grid_integrate_task_list 119 12.3 2.020 2.115 2.020 2.115 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.064 2.111 wfi_extrapolate 11 7.9 0.001 0.001 2.045 2.045 init_scf_loop 11 6.9 0.001 0.005 1.951 1.951 mp_sum_d 4135 12.0 1.285 1.914 1.285 1.914 potential_pw2rs 119 12.3 0.004 0.004 1.801 1.809 pw_transfer 1439 11.6 0.053 0.058 1.652 1.719 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.576 1.644 make_m2s 4572 13.5 0.054 0.056 1.506 1.550 make_images 4572 14.5 0.133 0.138 1.424 1.469 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.430 1.454 mp_waitany 12084 13.8 1.257 1.436 1.257 1.436 mp_alltoall_d11v 2130 13.8 1.178 1.432 1.178 1.432 grid_collocate_task_list 119 9.7 1.293 1.366 1.293 1.366 fft_wrap_pw1pw2_140 487 13.2 0.186 0.202 1.217 1.287 fft3d_ps 1201 14.6 0.367 0.472 1.217 1.284 dbcsr_dot_sd 1205 11.9 0.049 0.061 0.692 1.095 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=52.678000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=431.181818, yerr=1.192262 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/04dd581a73886059b5d5fb12035b1998b2b62f5a_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 489.525248E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2194560 MPI messages size (bytes): total size 310.646604E+09 min size 0.000000E+00 max size 1.145520E+06 average size 141.553031E+03 MPI breakdown and total messages size (bytes): size <= 128 724648 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 281952 4619501568 32768 < size <= 131072 494448 39143342080 131072 < size <= 4194304 440000 264807943488 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62664. MP_Allreduce 10226 305. MP_Sync 54 MP_Alltoall 2060 1391281. 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 0.018 0.037 38.807 38.809 qs_mol_dyn_low 1 2.0 0.003 0.005 38.537 38.544 qs_forces 11 3.9 0.008 0.047 38.428 38.433 qs_energies 11 4.9 0.002 0.010 36.715 36.717 scf_env_do_scf 11 5.9 0.001 0.002 31.411 31.412 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 28.972 28.972 dbcsr_multiply_generic 2286 12.5 0.100 0.103 21.407 21.758 qs_scf_new_mos 108 7.5 0.001 0.001 19.877 20.108 qs_scf_loop_do_ot 108 8.5 0.001 0.001 19.877 20.108 ot_scf_mini 108 9.5 0.003 0.003 18.983 19.158 multiply_cannon 2286 13.5 0.209 0.224 16.505 18.115 velocity_verlet 10 3.0 0.002 0.013 18.070 18.071 multiply_cannon_loop 2286 14.5 0.901 0.976 15.407 16.726 ot_mini 108 10.5 0.001 0.002 11.667 11.903 mp_waitall_1 200699 16.5 5.604 10.958 5.604 10.958 multiply_cannon_metrocomm3 27432 15.5 0.068 0.069 4.161 9.487 qs_ot_get_derivative 108 11.5 0.001 0.002 9.215 9.392 multiply_cannon_multrec 27432 15.5 1.973 4.328 6.067 8.827 rebuild_ks_matrix 119 8.3 0.000 0.000 7.148 7.287 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 7.148 7.287 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.314 6.440 dbcsr_mm_accdrv_process 47894 16.0 3.074 5.215 4.024 5.914 qs_ot_get_p 119 10.4 0.001 0.001 4.424 4.647 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.586 4.442 sum_up_and_integrate 119 10.3 0.025 0.028 4.149 4.156 integrate_v_rspace 119 11.3 0.003 0.006 4.124 4.135 apply_preconditioner_dbcsr 119 12.6 0.000 0.001 3.043 4.117 apply_single 119 13.6 0.000 0.000 3.043 4.117 init_scf_run 11 5.9 0.000 0.001 4.058 4.059 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.058 4.058 mp_sum_l 7207 12.9 2.021 3.968 2.021 3.968 qs_rho_update_rho_low 119 7.7 0.001 0.002 3.734 3.762 calculate_rho_elec 119 8.7 0.021 0.024 3.733 3.762 qs_ot_p2m_diag 50 11.0 0.009 0.013 3.014 3.032 rs_pw_transfer 974 11.9 0.010 0.012 2.530 2.942 multiply_cannon_sync_h2d 27432 15.5 2.209 2.841 2.209 2.841 make_m2s 4572 13.5 0.053 0.054 2.438 2.667 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.583 2.583 make_images 4572 14.5 0.201 0.237 2.351 2.579 calculate_first_density_matrix 1 7.0 0.001 0.002 2.576 2.577 density_rs2pw 119 9.7 0.004 0.004 2.059 2.487 init_scf_loop 11 6.9 0.001 0.004 2.419 2.420 ot_diis_step 108 11.5 0.011 0.013 2.402 2.403 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.266 2.266 cp_fm_redistribute_end 50 14.0 1.148 2.239 1.151 2.242 calculate_dm_sparse 119 9.5 0.000 0.001 2.144 2.219 cp_fm_diag_elpa_base 50 14.0 1.058 2.150 1.087 2.189 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.097 2.185 jit_kernel_multiply 10 16.1 0.898 2.090 0.898 2.090 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.018 2.020 grid_integrate_task_list 119 12.3 1.839 1.940 1.839 1.940 pw_transfer 1439 11.6 0.065 0.070 1.894 1.921 potential_pw2rs 119 12.3 0.006 0.006 1.874 1.882 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.794 1.835 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.802 1.832 make_images_data 4572 15.5 0.045 0.052 1.132 1.572 prepare_preconditioner 11 7.9 0.000 0.001 1.520 1.547 make_preconditioner 11 8.9 0.000 0.002 1.520 1.547 acc_transpose_blocks 27432 15.5 0.109 0.113 1.200 1.500 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.414 1.471 fft_wrap_pw1pw2_140 487 13.2 0.205 0.215 1.430 1.461 hybrid_alltoall_any 4725 16.4 0.051 0.112 0.975 1.453 wfi_extrapolate 11 7.9 0.001 0.001 1.431 1.431 fft3d_ps 1201 14.6 0.517 0.572 1.349 1.373 grid_collocate_task_list 119 9.7 1.230 1.367 1.230 1.367 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.355 1.363 mp_allgather_i34 2286 14.5 0.537 1.296 0.537 1.296 mp_alltoall_d11v 2130 13.8 1.148 1.267 1.148 1.267 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.176 1.227 mp_sum_d 4135 12.0 0.621 1.065 0.621 1.065 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 0.973 0.975 rs_pw_transfer_RS2PW_140 130 11.5 0.140 0.148 0.552 0.963 mp_waitany 5720 13.7 0.529 0.942 0.529 0.942 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.891 0.904 acc_transpose_blocks_kernels 27432 16.5 0.180 0.269 0.670 0.883 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=38.809000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=464.818182, yerr=1.641406 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/04dd581a73886059b5d5fb12035b1998b2b62f5a_performance_tests/05/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 59.051995E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3143552 0.0% 0.0% 100.0% average stack size 0.0 0.0 46.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 523.366400E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63497. MP_Allreduce 10075 307. MP_Sync 54 MP_Alltoall 1821 1519277. 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 0.014 0.029 31.753 31.755 qs_mol_dyn_low 1 2.0 0.003 0.005 31.507 31.514 qs_forces 11 3.9 0.003 0.004 31.446 31.447 qs_energies 11 4.9 0.002 0.007 29.904 29.906 scf_env_do_scf 11 5.9 0.001 0.001 25.174 25.174 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 22.681 22.682 dbcsr_multiply_generic 2286 12.5 0.094 0.096 16.144 16.222 velocity_verlet 10 3.0 0.001 0.002 14.971 14.973 qs_scf_new_mos 108 7.5 0.001 0.001 14.590 14.610 qs_scf_loop_do_ot 108 8.5 0.001 0.001 14.590 14.610 ot_scf_mini 108 9.5 0.002 0.004 13.876 13.889 multiply_cannon 2286 13.5 0.196 0.204 12.981 13.723 multiply_cannon_loop 2286 14.5 0.635 0.663 12.217 12.984 ot_mini 108 10.5 0.001 0.002 8.549 8.565 multiply_cannon_multrec 18288 15.5 1.913 2.880 6.797 7.079 qs_ot_get_derivative 108 11.5 0.001 0.002 7.060 7.074 rebuild_ks_matrix 119 8.3 0.000 0.000 6.320 6.341 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.016 6.319 6.341 dbcsr_mm_accdrv_process 38222 16.0 3.900 5.212 4.798 5.615 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.583 5.604 sum_up_and_integrate 119 10.3 0.032 0.034 3.905 3.918 integrate_v_rspace 119 11.3 0.003 0.003 3.873 3.886 mp_waitall_1 158411 16.6 2.496 3.671 2.496 3.671 init_scf_run 11 5.9 0.000 0.001 3.542 3.542 scf_env_initial_rho_setup 11 6.9 0.001 0.002 3.542 3.542 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.379 3.381 calculate_rho_elec 119 8.7 0.031 0.032 3.378 3.380 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.634 3.270 qs_ot_get_p 119 10.4 0.001 0.001 3.174 3.200 init_scf_loop 11 6.9 0.001 0.004 2.477 2.477 rs_pw_transfer 974 11.9 0.009 0.010 2.196 2.462 multiply_cannon_metrocomm3 18288 15.5 0.044 0.046 1.375 2.451 apply_preconditioner_dbcsr 119 12.6 0.000 0.001 2.044 2.364 apply_single 119 13.6 0.000 0.000 2.043 2.364 calculate_first_density_matrix 1 7.0 0.001 0.002 2.355 2.356 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.166 2.172 density_rs2pw 119 9.7 0.004 0.004 1.900 2.141 jit_kernel_multiply 10 15.8 0.846 1.945 0.846 1.945 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.905 1.905 pw_transfer 1439 11.6 0.065 0.070 1.863 1.872 calculate_dm_sparse 119 9.5 0.000 0.001 1.838 1.856 make_m2s 4572 13.5 0.044 0.045 1.703 1.854 grid_integrate_task_list 119 12.3 1.788 1.852 1.788 1.852 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.771 1.781 make_images 4572 14.5 0.190 0.202 1.620 1.770 prepare_preconditioner 11 7.9 0.000 0.000 1.697 1.701 make_preconditioner 11 8.9 0.000 0.002 1.697 1.701 potential_pw2rs 119 12.3 0.007 0.008 1.673 1.678 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.665 1.676 cp_fm_diag_elpa_base 50 14.0 1.643 1.655 1.663 1.674 multiply_cannon_sync_h2d 18288 15.5 1.397 1.672 1.397 1.672 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.644 1.646 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.551 1.635 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.517 1.523 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.515 1.522 ot_diis_step 108 11.5 0.012 0.022 1.470 1.477 fft_wrap_pw1pw2_140 487 13.2 0.258 0.263 1.436 1.447 mp_sum_l 7207 12.9 1.078 1.411 1.078 1.411 grid_collocate_task_list 119 9.7 1.218 1.373 1.218 1.373 fft3d_ps 1201 14.6 0.527 0.544 1.252 1.262 acc_transpose_blocks 18288 15.5 0.076 0.077 1.215 1.243 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.215 1.219 wfi_extrapolate 11 7.9 0.001 0.002 1.139 1.139 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 0.968 0.969 make_images_data 4572 15.5 0.045 0.049 0.771 0.923 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.863 0.887 hybrid_alltoall_any 4725 16.4 0.055 0.112 0.666 0.852 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.805 0.807 acc_transpose_blocks_kernels 18288 16.5 0.209 0.218 0.777 0.798 mp_waitany 9880 13.7 0.519 0.796 0.519 0.796 rs_pw_transfer_RS2PW_140 130 11.5 0.120 0.123 0.512 0.783 mp_alltoall_d11v 2130 13.8 0.643 0.753 0.643 0.753 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.643 0.699 cp_fm_cholesky_invert 11 10.9 0.688 0.692 0.688 0.692 mp_alltoall_z22v 1201 16.6 0.586 0.653 0.586 0.653 multiply_cannon_metrocomm1 18288 15.5 0.028 0.029 0.402 0.638 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=31.755000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=497.090909, yerr=2.502891 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/04dd581a73886059b5d5fb12035b1998b2b62f5a_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 558.981120E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63496. MP_Allreduce 10074 349. MP_Sync 54 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.032 34.599 34.601 qs_mol_dyn_low 1 2.0 0.003 0.003 34.437 34.445 qs_forces 11 3.9 0.002 0.003 34.372 34.373 qs_energies 11 4.9 0.001 0.001 32.664 32.672 scf_env_do_scf 11 5.9 0.000 0.001 27.810 27.811 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 24.473 24.473 dbcsr_multiply_generic 2286 12.5 0.098 0.104 17.943 18.067 velocity_verlet 10 3.0 0.001 0.001 17.586 17.588 qs_scf_new_mos 108 7.5 0.001 0.001 16.083 16.136 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.082 16.135 ot_scf_mini 108 9.5 0.002 0.003 15.150 15.201 multiply_cannon 2286 13.5 0.220 0.229 14.380 14.820 multiply_cannon_loop 2286 14.5 0.941 0.970 13.456 13.939 ot_mini 108 10.5 0.001 0.001 9.193 9.257 multiply_cannon_multrec 27432 15.5 2.345 3.036 8.628 8.970 qs_ot_get_derivative 108 11.5 0.001 0.001 7.376 7.427 dbcsr_mm_accdrv_process 47916 15.9 5.460 7.198 6.186 7.389 rebuild_ks_matrix 119 8.3 0.000 0.000 6.537 6.581 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 6.537 6.581 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.805 5.844 sum_up_and_integrate 119 10.3 0.035 0.038 3.772 3.781 integrate_v_rspace 119 11.3 0.003 0.003 3.736 3.746 init_scf_run 11 5.9 0.000 0.001 3.541 3.541 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.541 3.541 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.494 3.526 calculate_rho_elec 119 8.7 0.040 0.046 3.493 3.525 qs_ot_get_p 119 10.4 0.001 0.001 3.351 3.427 init_scf_loop 11 6.9 0.000 0.000 3.319 3.319 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.657 3.137 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.120 2.596 apply_single 119 13.6 0.000 0.000 2.120 2.596 prepare_preconditioner 11 7.9 0.000 0.000 2.485 2.493 make_preconditioner 11 8.9 0.000 0.000 2.485 2.493 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.094 2.419 mp_waitall_1 137007 16.6 1.695 2.271 1.695 2.271 make_m2s 4572 13.5 0.054 0.056 2.107 2.229 calculate_first_density_matrix 1 7.0 0.001 0.001 2.193 2.193 rs_pw_transfer 974 11.9 0.009 0.009 1.944 2.134 make_images 4572 14.5 0.274 0.330 2.000 2.120 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.110 2.119 density_rs2pw 119 9.7 0.004 0.004 1.907 2.095 calculate_dm_sparse 119 9.5 0.000 0.000 2.018 2.076 pw_transfer 1439 11.6 0.065 0.069 1.981 2.017 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.889 1.929 grid_integrate_task_list 119 12.3 1.820 1.921 1.820 1.921 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.886 1.911 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.800 1.801 ot_diis_step 108 11.5 0.012 0.012 1.773 1.773 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.767 1.768 jit_kernel_multiply 9 16.0 0.666 1.646 0.666 1.646 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.605 1.619 fft_wrap_pw1pw2_140 487 13.2 0.292 0.304 1.561 1.601 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.560 1.570 cp_fm_diag_elpa_base 50 14.0 1.526 1.544 1.558 1.567 mp_sum_l 7207 12.9 1.031 1.562 1.031 1.562 potential_pw2rs 119 12.3 0.009 0.010 1.529 1.533 acc_transpose_blocks 27432 15.5 0.113 0.115 1.446 1.472 grid_collocate_task_list 119 9.7 1.221 1.359 1.221 1.359 fft3d_ps 1201 14.6 0.556 0.607 1.322 1.352 wfi_extrapolate 11 7.9 0.001 0.001 1.300 1.300 multiply_cannon_metrocomm3 27432 15.5 0.038 0.040 0.761 1.277 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.203 1.213 cp_fm_upper_to_full 72 14.2 0.810 1.152 0.810 1.152 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.092 1.114 multiply_cannon_sync_h2d 27432 15.5 1.000 1.092 1.000 1.092 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.065 1.066 dbcsr_complete_redistribute 329 12.2 0.119 0.151 0.758 1.030 make_images_data 4572 15.5 0.045 0.048 0.820 0.946 hybrid_alltoall_any 4725 16.4 0.062 0.151 0.711 0.888 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.792 0.868 mp_alltoall_d11v 2130 13.8 0.708 0.854 0.708 0.854 acc_transpose_blocks_kernels 27432 16.5 0.268 0.275 0.823 0.833 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.822 0.827 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.551 0.816 cp_fm_cholesky_invert 11 10.9 0.714 0.717 0.714 0.717 mp_alltoall_i22 627 13.8 0.422 0.717 0.422 0.717 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=34.601000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=530.909091, yerr=3.449757 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/04dd581a73886059b5d5fb12035b1998b2b62f5a_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 601.473024E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63495. MP_Allreduce 10074 348. MP_Sync 54 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.032 27.522 27.523 qs_mol_dyn_low 1 2.0 0.003 0.003 27.338 27.340 qs_forces 11 3.9 0.002 0.003 27.222 27.223 qs_energies 11 4.9 0.001 0.001 25.519 25.522 scf_env_do_scf 11 5.9 0.000 0.001 20.758 20.758 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 18.329 18.330 velocity_verlet 10 3.0 0.001 0.001 14.109 14.112 dbcsr_multiply_generic 2286 12.5 0.092 0.097 11.869 11.940 qs_scf_new_mos 108 7.5 0.001 0.001 10.473 10.496 qs_scf_loop_do_ot 108 8.5 0.001 0.001 10.473 10.495 multiply_cannon 2286 13.5 0.229 0.236 9.471 9.907 ot_scf_mini 108 9.5 0.002 0.002 9.813 9.838 multiply_cannon_loop 2286 14.5 0.329 0.341 8.571 8.746 multiply_cannon_multrec 9144 15.5 1.565 1.801 5.842 6.068 rebuild_ks_matrix 119 8.3 0.000 0.000 5.917 5.941 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 5.916 5.940 ot_mini 108 10.5 0.001 0.001 5.442 5.471 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.271 5.293 dbcsr_mm_accdrv_process 12550 15.8 3.398 4.226 4.177 4.266 qs_ot_get_derivative 108 11.5 0.001 0.001 4.168 4.193 sum_up_and_integrate 119 10.3 0.038 0.042 3.617 3.623 integrate_v_rspace 119 11.3 0.003 0.003 3.579 3.585 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.439 3.444 calculate_rho_elec 119 8.7 0.060 0.061 3.438 3.444 init_scf_run 11 5.9 0.000 0.001 3.275 3.275 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.275 3.275 qs_ot_get_p 119 10.4 0.001 0.001 2.620 2.659 init_scf_loop 11 6.9 0.000 0.000 2.409 2.410 calculate_first_density_matrix 1 7.0 0.003 0.003 2.168 2.169 pw_transfer 1439 11.6 0.066 0.069 2.011 2.021 grid_integrate_task_list 119 12.3 1.844 1.948 1.844 1.948 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.917 1.928 density_rs2pw 119 9.7 0.004 0.004 1.793 1.928 jit_kernel_multiply 10 15.7 0.742 1.871 0.742 1.871 make_m2s 4572 13.5 0.034 0.035 1.665 1.826 rs_pw_transfer 974 11.9 0.008 0.008 1.641 1.803 calculate_dm_sparse 119 9.5 0.000 0.000 1.772 1.789 mp_waitall_1 115863 16.7 1.260 1.767 1.260 1.767 make_images 4572 14.5 0.269 0.302 1.578 1.736 qs_ot_p2m_diag 50 11.0 0.023 0.023 1.733 1.736 prepare_preconditioner 11 7.9 0.000 0.000 1.676 1.680 make_preconditioner 11 8.9 0.000 0.000 1.676 1.680 fft_wrap_pw1pw2_140 487 13.2 0.369 0.378 1.586 1.598 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.568 1.595 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.544 1.545 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.530 1.531 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.468 1.486 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.383 1.394 grid_collocate_task_list 119 9.7 1.271 1.388 1.271 1.388 potential_pw2rs 119 12.3 0.010 0.011 1.358 1.361 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.280 1.288 cp_fm_diag_elpa_base 50 14.0 1.253 1.269 1.278 1.287 ot_diis_step 108 11.5 0.012 0.013 1.262 1.263 fft3d_ps 1201 14.6 0.558 0.568 1.244 1.254 qs_energies_init_hamiltonians 11 5.9 0.002 0.004 1.252 1.253 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.208 1.214 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.143 1.169 apply_single 119 13.6 0.000 0.000 1.142 1.169 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.095 1.109 wfi_extrapolate 11 7.9 0.001 0.001 1.060 1.060 hybrid_alltoall_any 4725 16.4 0.063 0.177 0.739 0.986 make_images_data 4572 15.5 0.040 0.043 0.776 0.965 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.869 0.920 acc_transpose_blocks 9144 15.5 0.038 0.039 0.876 0.893 cp_fm_cholesky_invert 11 10.9 0.834 0.837 0.834 0.837 mp_alltoall_d11v 2130 13.8 0.701 0.832 0.701 0.832 multiply_cannon_sync_h2d 9144 15.5 0.709 0.798 0.709 0.798 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.768 0.770 qs_env_update_s_mstruct 11 6.9 0.000 0.001 0.705 0.754 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.702 0.712 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.315 0.658 acc_transpose_blocks_kernels 9144 16.5 0.116 0.120 0.637 0.650 mp_allgather_i34 2286 14.5 0.229 0.628 0.229 0.628 mp_waitany 5200 13.7 0.468 0.617 0.468 0.617 rs_pw_transfer_RS2PW_140 130 11.5 0.101 0.104 0.427 0.589 mp_alltoall_z22v 1201 16.6 0.561 0.588 0.561 0.588 qs_create_task_list 11 7.9 0.000 0.000 0.540 0.567 generate_qs_task_list 11 8.9 0.187 0.209 0.540 0.567 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=27.523000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=569.727273, yerr=4.574453 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/04dd581a73886059b5d5fb12035b1998b2b62f5a_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 763.772928E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63729. MP_Allreduce 10074 433. MP_Sync 54 MP_Alltoall 1582 7383731. MP_SendRecv 2499 189067. MP_ISendRecv 2499 189067. MP_Wait 6399 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.037 40.628 40.629 qs_mol_dyn_low 1 2.0 0.003 0.003 40.233 40.240 qs_forces 11 3.9 0.013 0.021 40.167 40.168 qs_energies 11 4.9 0.002 0.003 38.193 38.203 scf_env_do_scf 11 5.9 0.001 0.001 32.693 32.693 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 25.011 25.013 velocity_verlet 10 3.0 0.001 0.001 22.913 22.919 dbcsr_multiply_generic 2286 12.5 0.098 0.102 17.046 17.283 qs_scf_new_mos 108 7.5 0.001 0.001 15.394 15.491 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.393 15.491 ot_scf_mini 108 9.5 0.002 0.002 14.327 14.427 multiply_cannon 2286 13.5 0.300 0.309 13.389 14.374 multiply_cannon_loop 2286 14.5 0.343 0.348 12.168 13.186 ot_mini 108 10.5 0.001 0.001 8.576 8.692 multiply_cannon_multrec 9144 15.5 3.409 4.717 8.549 8.676 init_scf_loop 11 6.9 0.000 0.000 7.656 7.661 rebuild_ks_matrix 119 8.3 0.000 0.000 7.194 7.334 qs_ks_build_kohn_sham_matrix 119 9.3 0.014 0.014 7.193 7.333 prepare_preconditioner 11 7.9 0.000 0.000 6.697 6.711 make_preconditioner 11 8.9 0.000 0.000 6.697 6.711 qs_ot_get_derivative 108 11.5 0.001 0.001 6.590 6.689 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.501 6.629 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.256 6.592 dbcsr_mm_accdrv_process 12550 15.8 4.096 5.618 5.013 6.369 cp_fm_upper_to_full 72 14.2 3.164 4.552 3.164 4.552 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.160 4.169 calculate_rho_elec 119 8.7 0.119 0.122 4.159 4.169 sum_up_and_integrate 119 10.3 0.065 0.066 3.946 3.952 integrate_v_rspace 119 11.3 0.004 0.004 3.881 3.888 init_scf_run 11 5.9 0.000 0.001 3.472 3.472 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.472 3.472 qs_ot_get_p 119 10.4 0.001 0.001 3.064 3.198 mp_waitall_1 94719 16.7 2.071 3.122 2.071 3.122 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.369 2.823 dbcsr_complete_redistribute 329 12.2 0.289 0.293 1.952 2.760 pw_transfer 1439 11.6 0.069 0.069 2.604 2.607 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.505 2.509 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.656 2.456 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.186 2.427 apply_single 119 13.6 0.000 0.000 2.185 2.427 make_m2s 4572 13.5 0.038 0.038 2.118 2.262 mp_alltoall_i22 627 13.8 1.409 2.247 1.409 2.247 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.436 2.229 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 1.181 2.158 calculate_dm_sparse 119 9.5 0.000 0.000 2.132 2.150 make_images 4572 14.5 0.356 0.391 1.998 2.142 density_rs2pw 119 9.7 0.004 0.004 2.105 2.128 fft_wrap_pw1pw2_140 487 13.2 0.621 0.626 2.122 2.127 calculate_first_density_matrix 1 7.0 0.001 0.001 2.066 2.069 grid_integrate_task_list 119 12.3 2.043 2.067 2.043 2.067 ot_diis_step 108 11.5 0.014 0.014 1.962 1.962 qs_ot_p2m_diag 50 11.0 0.043 0.044 1.856 1.857 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.779 1.779 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.768 1.778 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.634 1.685 mp_sum_l 7207 12.9 1.018 1.680 1.018 1.680 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.587 1.587 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.538 1.578 fft3d_ps 1201 14.6 0.595 0.606 1.517 1.521 grid_collocate_task_list 119 9.7 1.478 1.490 1.478 1.490 rs_pw_transfer 974 11.9 0.009 0.009 1.415 1.447 jit_kernel_multiply 8 15.5 0.890 1.400 0.890 1.400 cp_fm_cholesky_invert 11 10.9 1.367 1.371 1.367 1.371 potential_pw2rs 119 12.3 0.014 0.015 1.352 1.355 wfi_extrapolate 11 7.9 0.001 0.001 1.342 1.342 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.311 1.312 cp_fm_diag_elpa_base 50 14.0 1.164 1.217 1.309 1.309 hybrid_alltoall_any 4725 16.4 0.087 0.147 1.029 1.225 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.211 1.218 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.115 1.133 make_images_data 4572 15.5 0.044 0.047 0.952 1.126 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.097 1.115 mp_alltoall_d11v 2130 13.8 1.033 1.053 1.033 1.053 multiply_cannon_sync_h2d 9144 15.5 1.043 1.047 1.043 1.047 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.983 1.020 qs_create_task_list 11 7.9 0.000 0.000 0.933 0.945 generate_qs_task_list 11 8.9 0.368 0.387 0.933 0.945 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.903 0.916 acc_transpose_blocks 9144 15.5 0.038 0.039 0.902 0.912 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=40.629000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=717.090909, yerr=16.692157 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/04dd581a73886059b5d5fb12035b1998b2b62f5a_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 501.682176E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66218. MP_Allreduce 9696 492. MP_Sync 52 MP_Alltoall 1938 656470. 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 0.019 0.041 80.685 80.687 qs_mol_dyn_low 1 2.0 0.003 0.003 80.401 80.418 qs_forces 11 3.9 0.003 0.003 80.331 80.331 qs_energies 11 4.9 0.001 0.002 77.490 77.502 scf_env_do_scf 11 5.9 0.000 0.001 68.679 68.681 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 63.308 63.308 dbcsr_multiply_generic 2055 12.4 0.104 0.106 50.058 50.332 qs_scf_new_mos 99 7.5 0.000 0.001 46.124 46.255 qs_scf_loop_do_ot 99 8.5 0.001 0.001 46.124 46.254 ot_scf_mini 99 9.5 0.002 0.002 43.828 43.907 multiply_cannon 2055 13.4 0.184 0.189 41.949 42.842 velocity_verlet 10 3.0 0.001 0.001 41.963 41.964 multiply_cannon_loop 2055 14.4 1.514 1.547 41.039 41.961 ot_mini 99 10.5 0.001 0.001 25.991 26.067 qs_ot_get_derivative 99 11.5 0.001 0.001 19.267 19.344 multiply_cannon_multrec 49320 15.4 12.471 13.130 17.449 18.142 rebuild_ks_matrix 110 8.3 0.000 0.001 14.482 14.593 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 14.481 14.593 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.699 12.795 mp_waitall_1 220248 16.4 10.265 11.183 10.265 11.183 multiply_cannon_sync_h2d 49320 15.4 10.400 10.834 10.400 10.834 qs_ot_get_p 110 10.4 0.001 0.001 9.583 9.692 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.172 7.760 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.166 7.684 apply_single 110 13.6 0.000 0.001 7.166 7.684 multiply_cannon_metrocomm3 49320 15.4 0.082 0.086 6.256 7.395 sum_up_and_integrate 110 10.3 0.037 0.043 7.055 7.070 integrate_v_rspace 110 11.3 0.003 0.004 7.018 7.043 init_scf_run 11 5.9 0.000 0.001 6.794 6.795 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.794 6.794 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.477 6.613 calculate_rho_elec 110 8.6 0.021 0.026 6.476 6.613 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.564 6.586 ot_diis_step 99 11.5 0.005 0.005 6.555 6.556 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.746 5.746 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.477 5.527 init_scf_loop 11 6.9 0.000 0.000 5.345 5.346 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.185 5.214 cp_fm_diag_elpa_base 48 14.0 5.173 5.201 5.183 5.212 dbcsr_mm_accdrv_process 87628 16.1 1.943 2.041 4.853 5.120 mp_sum_l 6514 12.8 3.931 4.590 3.931 4.590 rs_pw_transfer 902 11.9 0.012 0.014 3.661 4.251 wfi_extrapolate 11 7.9 0.001 0.001 3.982 3.983 density_rs2pw 110 9.6 0.004 0.005 3.316 3.920 make_m2s 4110 13.4 0.060 0.065 3.720 3.824 calculate_dm_sparse 110 9.5 0.001 0.001 3.672 3.765 make_images 4110 14.4 0.178 0.193 3.625 3.732 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.650 3.653 grid_integrate_task_list 110 12.3 3.243 3.428 3.243 3.428 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.322 3.396 multiply_cannon_metrocomm1 49320 15.4 0.061 0.063 2.258 3.385 prepare_preconditioner 11 7.9 0.000 0.000 3.202 3.226 make_preconditioner 11 8.9 0.000 0.000 3.202 3.226 pw_transfer 1331 11.6 0.056 0.065 3.046 3.117 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.009 3.070 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.000 3.047 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 2.957 3.032 calculate_first_density_matrix 1 7.0 0.001 0.001 2.723 2.728 jit_kernel_multiply 13 15.9 2.635 2.704 2.635 2.704 fft_wrap_pw1pw2_140 451 13.1 0.454 0.499 2.537 2.619 potential_pw2rs 110 12.3 0.006 0.007 2.594 2.616 mp_waitany 14300 13.8 1.847 2.515 1.847 2.515 mp_alltoall_d11v 2046 13.8 1.970 2.460 1.970 2.460 grid_collocate_task_list 110 9.6 2.086 2.299 2.086 2.299 fft3d_ps 1111 14.6 0.780 0.866 2.148 2.197 acc_transpose_blocks 49320 15.4 0.210 0.218 2.054 2.107 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.912 1.928 mp_sum_d 3889 11.9 1.276 1.804 1.276 1.804 make_images_data 4110 15.4 0.042 0.046 1.701 1.803 hybrid_alltoall_any 4261 16.3 0.083 0.484 1.489 1.757 cp_fm_cholesky_invert 11 10.9 1.737 1.741 1.737 1.741 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.646 1.676 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=80.687000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=477.181818, yerr=2.166614 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/04dd581a73886059b5d5fb12035b1998b2b62f5a_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 588.718080E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1972800 MPI messages size (bytes): total size 1.077520E+12 min size 0.000000E+00 max size 4.537280E+06 average size 546.188250E+03 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 222984 1826684928 8192 < size <= 32768 520356 13399818240 32768 < size <= 131072 372336 35386294272 131072 < size <= 4194304 787758 788321309808 4194304 < size <= 16777216 54450 238588003280 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66437. MP_Allreduce 9695 570. MP_Sync 52 MP_Alltoall 1717 2912639. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.026 0.075 69.181 69.181 qs_mol_dyn_low 1 2.0 0.003 0.003 68.787 68.796 qs_forces 11 3.9 0.004 0.009 68.360 68.361 qs_energies 11 4.9 0.002 0.006 65.032 65.036 scf_env_do_scf 11 5.9 0.000 0.001 56.420 56.424 scf_env_do_scf_inner_loop 99 6.5 0.004 0.019 48.963 48.964 dbcsr_multiply_generic 2055 12.4 0.178 0.205 37.366 37.511 velocity_verlet 10 3.0 0.001 0.001 36.172 36.181 qs_scf_new_mos 99 7.5 0.001 0.001 32.645 32.765 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.645 32.765 multiply_cannon 2055 13.4 0.220 0.243 30.814 31.970 ot_scf_mini 99 9.5 0.003 0.004 31.004 31.121 multiply_cannon_loop 2055 14.4 0.924 0.946 29.561 30.536 ot_mini 99 10.5 0.001 0.001 17.988 18.112 multiply_cannon_multrec 24660 15.4 7.614 9.096 13.819 15.420 rebuild_ks_matrix 110 8.3 0.000 0.001 13.778 13.883 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.015 13.778 13.883 qs_ot_get_derivative 99 11.5 0.001 0.001 12.224 12.343 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.126 12.218 mp_waitall_1 176588 16.5 7.498 9.715 7.498 9.715 multiply_cannon_sync_h2d 24660 15.4 6.960 7.889 6.960 7.889 multiply_cannon_metrocomm3 24660 15.4 0.068 0.070 5.085 7.725 init_scf_loop 11 6.9 0.001 0.003 7.423 7.423 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.498 7.179 apply_single 110 13.6 0.000 0.001 6.498 7.178 sum_up_and_integrate 110 10.3 0.052 0.059 6.656 6.669 integrate_v_rspace 110 11.3 0.003 0.003 6.603 6.618 qs_ot_get_p 110 10.4 0.001 0.001 6.430 6.561 dbcsr_mm_accdrv_process 52282 16.1 4.680 5.462 6.046 6.334 init_scf_run 11 5.9 0.000 0.001 6.247 6.248 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.247 6.248 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.184 6.195 calculate_rho_elec 110 8.6 0.040 0.048 6.183 6.194 ot_diis_step 99 11.5 0.010 0.010 5.716 5.716 prepare_preconditioner 11 7.9 0.000 0.000 5.394 5.416 make_preconditioner 11 8.9 0.000 0.002 5.394 5.416 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.609 5.382 make_full_inverse_cholesky 11 9.9 0.000 0.000 4.982 5.131 qs_ot_p2m_diag 48 11.0 0.030 0.053 4.592 4.612 make_m2s 4110 13.4 0.057 0.061 4.106 4.546 make_images 4110 14.4 0.401 0.447 4.000 4.437 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.163 4.164 density_rs2pw 110 9.6 0.004 0.005 3.288 3.831 pw_transfer 1331 11.6 0.067 0.077 3.636 3.769 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.528 3.663 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.589 3.603 cp_fm_diag_elpa_base 48 14.0 3.541 3.563 3.587 3.600 rs_pw_transfer 902 11.9 0.012 0.014 3.000 3.545 wfi_extrapolate 11 7.9 0.001 0.001 3.493 3.493 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.402 3.404 grid_integrate_task_list 110 12.3 3.143 3.320 3.143 3.320 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.120 3.181 fft_wrap_pw1pw2_140 451 13.1 0.527 0.541 3.022 3.158 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.010 3.066 calculate_dm_sparse 110 9.5 0.001 0.001 2.938 2.966 make_images_data 4110 15.4 0.047 0.051 2.303 2.747 hybrid_alltoall_any 4261 16.3 0.104 0.446 1.987 2.718 calculate_first_density_matrix 1 7.0 0.001 0.003 2.664 2.668 fft3d_ps 1111 14.6 1.103 1.335 2.493 2.643 cp_fm_cholesky_invert 11 10.9 2.524 2.530 2.524 2.530 grid_collocate_task_list 110 9.6 2.048 2.493 2.048 2.493 potential_pw2rs 110 12.3 0.008 0.009 2.447 2.460 mp_sum_l 6514 12.8 1.789 2.309 1.789 2.309 mp_alltoall_d11v 2046 13.8 1.729 2.035 1.729 2.035 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.932 1.949 jit_kernel_multiply 10 16.4 1.019 1.937 1.019 1.937 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.859 1.860 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.810 1.819 mp_waitany 10164 13.8 1.257 1.808 1.257 1.808 multiply_cannon_metrocomm4 22605 15.4 0.075 0.082 0.816 1.784 mp_irecv_dv 57340 16.2 0.692 1.665 0.692 1.665 mp_allgather_i34 2055 14.4 0.566 1.559 0.566 1.559 acc_transpose_blocks 24660 15.4 0.111 0.114 1.517 1.539 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.526 1.536 rs_pw_transfer_RS2PW_140 121 11.5 0.207 0.218 0.983 1.527 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.366 1.473 dbcsr_complete_redistribute 325 12.2 0.257 0.369 1.151 1.415 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=69.181000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=556.272727, yerr=8.057644 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/04dd581a73886059b5d5fb12035b1998b2b62f5a_performance_tests/11/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 404.681598E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 3346752 0.0% 0.0% 100.0% average stack size 0.0 0.0 294.1 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 660.840448E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66428. MP_Allreduce 9694 567. MP_Sync 52 MP_Alltoall 1496 4511006. MP_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.033 61.082 61.083 qs_mol_dyn_low 1 2.0 0.003 0.003 60.740 60.750 qs_forces 11 3.9 0.004 0.010 60.667 60.669 qs_energies 11 4.9 0.002 0.008 57.427 57.430 scf_env_do_scf 11 5.9 0.000 0.001 49.488 49.488 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 40.781 40.782 velocity_verlet 10 3.0 0.001 0.001 32.960 32.961 dbcsr_multiply_generic 2055 12.4 0.106 0.110 28.445 28.742 qs_scf_new_mos 99 7.5 0.001 0.001 25.551 25.654 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.551 25.653 ot_scf_mini 99 9.5 0.002 0.003 24.317 24.458 multiply_cannon 2055 13.4 0.211 0.227 22.043 23.123 multiply_cannon_loop 2055 14.4 0.619 0.633 20.878 21.958 ot_mini 99 10.5 0.001 0.001 13.680 13.818 rebuild_ks_matrix 110 8.3 0.000 0.000 12.474 12.625 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 12.473 12.624 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.980 11.113 multiply_cannon_multrec 16440 15.4 3.781 4.507 9.584 10.302 mp_waitall_1 139946 16.5 7.128 10.270 7.128 10.270 qs_ot_get_derivative 99 11.5 0.001 0.001 9.185 9.325 init_scf_loop 11 6.9 0.001 0.003 8.669 8.670 multiply_cannon_metrocomm3 16440 15.4 0.043 0.046 4.475 7.470 prepare_preconditioner 11 7.9 0.000 0.000 6.889 6.908 make_preconditioner 11 8.9 0.000 0.002 6.888 6.908 sum_up_and_integrate 110 10.3 0.060 0.061 6.650 6.666 integrate_v_rspace 110 11.3 0.003 0.004 6.589 6.605 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.202 6.546 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.086 6.093 calculate_rho_elec 110 8.6 0.059 0.059 6.085 6.093 qs_ot_get_p 110 10.4 0.001 0.001 5.857 6.017 dbcsr_mm_accdrv_process 34862 16.1 4.579 5.340 5.655 5.786 init_scf_run 11 5.9 0.000 0.001 5.495 5.496 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.495 5.495 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.996 5.392 apply_single 110 13.6 0.000 0.000 4.996 5.391 make_m2s 4110 13.4 0.051 0.052 4.216 4.550 ot_diis_step 99 11.5 0.011 0.011 4.466 4.466 make_images 4110 14.4 0.393 0.511 4.103 4.437 density_rs2pw 110 9.6 0.005 0.005 3.178 4.410 multiply_cannon_sync_h2d 16440 15.4 3.612 4.253 3.612 4.253 qs_ot_p2m_diag 48 11.0 0.042 0.044 4.174 4.181 rs_pw_transfer 902 11.9 0.011 0.012 2.775 3.994 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.822 3.822 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.148 3.769 pw_transfer 1331 11.6 0.066 0.074 3.747 3.757 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.639 3.651 grid_integrate_task_list 110 12.3 3.166 3.420 3.166 3.420 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.222 3.231 cp_fm_diag_elpa_base 48 14.0 3.153 3.183 3.220 3.229 fft_wrap_pw1pw2_140 451 13.1 0.646 0.654 3.103 3.118 wfi_extrapolate 11 7.9 0.001 0.001 3.001 3.001 make_images_data 4110 15.4 0.044 0.048 2.401 2.844 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.782 2.783 hybrid_alltoall_any 4261 16.3 0.106 0.376 2.132 2.773 cp_fm_cholesky_invert 11 10.9 2.736 2.744 2.736 2.744 calculate_dm_sparse 110 9.5 0.001 0.001 2.534 2.562 multiply_cannon_metrocomm4 14385 15.4 0.046 0.049 0.895 2.537 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.441 2.501 fft3d_ps 1111 14.6 1.090 1.097 2.451 2.463 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.370 2.446 grid_collocate_task_list 110 9.6 2.082 2.445 2.082 2.445 mp_waitany 17072 13.8 1.202 2.445 1.202 2.445 calculate_first_density_matrix 1 7.0 0.000 0.002 2.407 2.408 mp_irecv_dv 48980 15.7 0.825 2.408 0.825 2.408 potential_pw2rs 110 12.3 0.011 0.011 2.376 2.383 mp_alltoall_d11v 2046 13.8 1.842 2.247 1.842 2.247 mp_sum_l 6514 12.8 1.573 2.225 1.573 2.225 rs_pw_transfer_RS2PW_140 121 11.5 0.178 0.183 0.913 2.127 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.985 1.986 dbcsr_complete_redistribute 325 12.2 0.323 0.351 1.480 1.923 cp_fm_upper_to_full 70 14.2 1.391 1.767 1.391 1.767 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.692 1.708 cp_fm_cholesky_decompose 22 10.9 1.622 1.642 1.622 1.642 mp_allgather_i34 2055 14.4 0.488 1.544 0.488 1.544 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.483 1.498 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.356 1.480 jit_kernel_multiply 8 16.5 0.695 1.479 0.695 1.479 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.993 1.445 rs_gather_matrices 110 12.3 0.235 0.261 0.969 1.327 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.268 1.276 acc_transpose_blocks 16440 15.4 0.074 0.076 1.207 1.223 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=61.083000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=625.727273, yerr=8.147402 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/04dd581a73886059b5d5fb12035b1998b2b62f5a_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 740.089856E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66426. MP_Allreduce 9694 608. MP_Sync 52 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.022 0.034 64.517 64.519 qs_mol_dyn_low 1 2.0 0.003 0.004 64.208 64.217 qs_forces 11 3.9 0.004 0.010 64.141 64.142 qs_energies 11 4.9 0.002 0.006 60.714 60.719 scf_env_do_scf 11 5.9 0.001 0.001 52.503 52.506 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 40.938 40.939 velocity_verlet 10 3.0 0.001 0.001 36.617 36.619 dbcsr_multiply_generic 2055 12.4 0.112 0.118 29.152 29.365 qs_scf_new_mos 99 7.5 0.001 0.001 25.965 26.086 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.965 26.086 ot_scf_mini 99 9.5 0.003 0.004 24.350 24.462 multiply_cannon 2055 13.4 0.247 0.266 22.425 23.287 multiply_cannon_loop 2055 14.4 0.886 0.908 20.919 21.583 ot_mini 99 10.5 0.001 0.001 13.839 13.975 multiply_cannon_multrec 24660 15.4 4.187 6.795 12.628 13.893 rebuild_ks_matrix 110 8.3 0.000 0.000 12.183 12.304 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.015 12.183 12.303 init_scf_loop 11 6.9 0.005 0.024 11.525 11.530 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.765 10.877 qs_ot_get_derivative 99 11.5 0.001 0.001 9.701 9.821 prepare_preconditioner 11 7.9 0.000 0.000 9.755 9.774 make_preconditioner 11 8.9 0.000 0.002 9.755 9.774 make_full_inverse_cholesky 11 9.9 0.000 0.000 7.940 9.401 dbcsr_mm_accdrv_process 52304 16.0 6.865 8.146 8.292 9.253 sum_up_and_integrate 110 10.3 0.068 0.071 6.479 6.493 integrate_v_rspace 110 11.3 0.003 0.003 6.411 6.427 mp_waitall_1 121746 16.5 4.038 6.085 4.038 6.085 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.048 6.058 calculate_rho_elec 110 8.6 0.078 0.082 6.048 6.058 qs_ot_get_p 110 10.4 0.001 0.002 5.544 5.700 make_m2s 4110 13.4 0.060 0.062 5.150 5.612 make_images 4110 14.4 0.582 0.703 5.010 5.472 init_scf_run 11 5.9 0.000 0.001 5.431 5.431 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.431 5.431 cp_fm_upper_to_full 70 14.2 3.353 4.868 3.353 4.868 ot_diis_step 99 11.5 0.011 0.011 4.103 4.103 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.009 4.090 apply_single 110 13.6 0.000 0.000 4.008 4.090 dbcsr_complete_redistribute 325 12.2 0.433 0.484 2.685 3.851 qs_ot_p2m_diag 48 11.0 0.056 0.064 3.789 3.804 pw_transfer 1331 11.6 0.066 0.075 3.710 3.739 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.604 3.638 density_rs2pw 110 9.6 0.004 0.005 2.979 3.488 grid_integrate_task_list 110 12.3 3.262 3.433 3.262 3.433 multiply_cannon_sync_h2d 24660 15.4 3.184 3.390 3.184 3.390 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.172 3.325 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.318 3.318 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.216 3.271 fft_wrap_pw1pw2_140 451 13.1 0.676 0.696 3.111 3.147 multiply_cannon_metrocomm3 24660 15.4 0.036 0.037 1.413 3.141 make_images_data 4110 15.4 0.047 0.051 2.574 3.058 qs_ot_get_derivative_diag 47 12.0 0.002 0.006 2.994 3.053 hybrid_alltoall_any 4261 16.3 0.121 0.461 2.249 3.045 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.806 2.945 wfi_extrapolate 11 7.9 0.001 0.001 2.933 2.933 calculate_dm_sparse 110 9.5 0.001 0.001 2.895 2.919 rs_pw_transfer 902 11.9 0.010 0.011 2.353 2.857 mp_alltoall_i22 605 13.7 1.679 2.842 1.679 2.842 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.794 2.807 cp_fm_diag_elpa_base 48 14.0 2.647 2.709 2.792 2.806 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.800 2.802 cp_fm_cholesky_invert 11 10.9 2.580 2.588 2.580 2.588 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 2.433 2.467 grid_collocate_task_list 110 9.6 2.183 2.434 2.183 2.434 calculate_first_density_matrix 1 7.0 0.001 0.002 2.403 2.405 fft3d_ps 1111 14.6 1.081 1.112 2.355 2.370 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.286 2.287 jit_kernel_multiply 12 15.7 1.098 2.189 1.098 2.189 mp_alltoall_d11v 2046 13.8 1.707 2.164 1.707 2.164 potential_pw2rs 110 12.3 0.013 0.013 2.129 2.136 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.597 1.711 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.672 1.708 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.671 1.682 cp_fm_cholesky_decompose 22 10.9 1.607 1.662 1.607 1.662 mp_sum_l 6514 12.8 0.901 1.562 0.901 1.562 mp_waitany 13376 13.8 1.070 1.534 1.070 1.534 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.514 1.532 multiply_cannon_metrocomm4 20550 15.4 0.062 0.066 0.862 1.526 acc_transpose_blocks 24660 15.4 0.108 0.111 1.498 1.514 mp_irecv_dv 62702 16.1 0.759 1.448 0.759 1.448 rs_gather_matrices 110 12.3 0.268 0.296 0.940 1.365 rs_pw_transfer_RS2PW_140 121 11.5 0.170 0.177 0.836 1.344 qs_env_update_s_mstruct 11 6.9 0.001 0.002 1.228 1.340 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=64.519000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=700.181818, yerr=11.463921 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/04dd581a73886059b5d5fb12035b1998b2b62f5a_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 817.164288E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66424. MP_Allreduce 9694 649. MP_Sync 52 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.046 0.053 54.780 54.781 qs_mol_dyn_low 1 2.0 0.003 0.003 54.496 54.505 qs_forces 11 3.9 0.003 0.003 54.426 54.427 qs_energies 11 4.9 0.025 0.049 50.733 50.738 scf_env_do_scf 11 5.9 0.000 0.001 42.470 42.470 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 34.894 34.894 velocity_verlet 10 3.0 0.001 0.001 30.816 30.819 dbcsr_multiply_generic 2055 12.4 0.104 0.110 22.627 22.846 qs_scf_new_mos 99 7.5 0.001 0.001 19.893 19.942 qs_scf_loop_do_ot 99 8.5 0.001 0.001 19.893 19.941 multiply_cannon 2055 13.4 0.245 0.254 17.282 18.783 ot_scf_mini 99 9.5 0.002 0.002 18.662 18.684 multiply_cannon_loop 2055 14.4 0.321 0.333 15.929 16.355 rebuild_ks_matrix 110 8.3 0.000 0.000 11.913 11.951 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.015 11.913 11.951 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.574 10.609 ot_mini 99 10.5 0.001 0.001 10.240 10.257 multiply_cannon_multrec 8220 15.4 3.191 4.497 7.539 8.570 mp_waitall_1 103326 16.6 5.844 7.600 5.844 7.600 init_scf_loop 11 6.9 0.000 0.000 7.531 7.531 sum_up_and_integrate 110 10.3 0.079 0.080 6.502 6.514 qs_ot_get_derivative 99 11.5 0.001 0.001 6.456 6.477 integrate_v_rspace 110 11.3 0.003 0.003 6.423 6.434 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.237 6.254 calculate_rho_elec 110 8.6 0.115 0.116 6.236 6.253 prepare_preconditioner 11 7.9 0.000 0.000 5.872 5.876 make_preconditioner 11 8.9 0.000 0.000 5.872 5.876 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.462 5.531 dbcsr_mm_accdrv_process 17442 15.9 2.780 3.478 4.219 5.119 init_scf_run 11 5.9 0.000 0.001 5.101 5.101 scf_env_initial_rho_setup 11 6.9 0.015 0.015 5.100 5.101 qs_ot_get_p 110 10.4 0.001 0.001 4.517 4.554 make_m2s 4110 13.4 0.039 0.040 4.145 4.433 multiply_cannon_metrocomm3 8220 15.4 0.017 0.018 2.990 4.309 make_images 4110 14.4 0.643 0.697 4.017 4.303 pw_transfer 1331 11.6 0.066 0.074 3.902 3.932 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.794 3.827 ot_diis_step 99 11.5 0.012 0.012 3.764 3.764 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.681 3.721 apply_single 110 13.6 0.000 0.000 3.681 3.721 grid_integrate_task_list 110 12.3 3.360 3.512 3.360 3.512 density_rs2pw 110 9.6 0.004 0.004 2.980 3.349 fft_wrap_pw1pw2_140 451 13.1 0.840 0.855 3.292 3.336 qs_ot_p2m_diag 48 11.0 0.081 0.085 3.220 3.225 multiply_cannon_sync_h2d 8220 15.4 2.912 3.058 2.912 3.058 cp_dbcsr_syevd 48 12.0 0.003 0.003 2.927 2.927 cp_fm_cholesky_invert 11 10.9 2.826 2.829 2.826 2.829 make_images_data 4110 15.4 0.039 0.044 2.373 2.805 hybrid_alltoall_any 4261 16.3 0.202 0.874 2.295 2.702 wfi_extrapolate 11 7.9 0.001 0.001 2.694 2.695 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.675 2.675 calculate_dm_sparse 110 9.5 0.001 0.001 2.503 2.548 rs_pw_transfer 902 11.9 0.010 0.011 2.149 2.547 grid_collocate_task_list 110 9.6 2.277 2.539 2.277 2.539 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.506 2.507 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.426 2.433 cp_fm_diag_elpa_base 48 14.0 2.370 2.392 2.424 2.432 fft3d_ps 1111 14.6 1.132 1.174 2.329 2.349 calculate_first_density_matrix 1 7.0 0.001 0.001 2.294 2.298 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.173 2.190 potential_pw2rs 110 12.3 0.016 0.016 2.054 2.056 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.774 1.999 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.939 1.951 mp_alltoall_d11v 2046 13.8 1.570 1.811 1.570 1.811 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.742 1.751 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.667 1.670 cp_fm_cholesky_decompose 22 10.9 1.627 1.646 1.627 1.646 jit_kernel_multiply 9 15.8 1.137 1.645 1.137 1.645 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.509 1.637 mp_allgather_i34 2055 14.4 0.534 1.614 0.534 1.614 mp_waitany 9240 13.8 1.071 1.488 1.071 1.488 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.470 1.483 dbcsr_complete_redistribute 325 12.2 0.568 0.594 1.382 1.461 qs_create_task_list 11 7.9 0.000 0.000 1.222 1.330 generate_qs_task_list 11 8.9 0.374 0.443 1.222 1.330 rs_pw_transfer_RS2PW_140 121 11.5 0.163 0.166 0.797 1.197 multiply_cannon_metrocomm4 6165 15.4 0.017 0.019 0.488 1.186 mp_irecv_dv 24056 15.7 0.464 1.141 0.464 1.141 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.091 1.117 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=54.781000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=773.636364, yerr=10.645295 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/04dd581a73886059b5d5fb12035b1998b2b62f5a_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 1.370116E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67104. MP_Allreduce 9672 819. MP_Sync 52 MP_Alltoall 1474 16505187. MP_SendRecv 2310 360267. MP_ISendRecv 2310 360267. MP_Wait 5214 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.040 87.043 87.045 qs_mol_dyn_low 1 2.0 0.003 0.003 86.721 86.734 qs_forces 11 3.9 0.003 0.003 86.641 86.642 qs_energies 11 4.9 0.001 0.001 82.489 82.492 scf_env_do_scf 11 5.9 0.001 0.001 72.510 72.510 velocity_verlet 10 3.0 0.001 0.001 55.683 55.690 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 44.268 44.269 dbcsr_multiply_generic 2055 12.4 0.119 0.123 28.517 28.640 init_scf_loop 11 6.9 0.000 0.000 28.171 28.173 qs_scf_new_mos 99 7.5 0.001 0.001 26.090 26.177 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.089 26.176 prepare_preconditioner 11 7.9 0.000 0.000 26.124 26.136 make_preconditioner 11 8.9 0.000 0.000 26.124 26.136 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.443 25.586 ot_scf_mini 99 9.5 0.002 0.002 24.337 24.409 multiply_cannon 2055 13.4 0.341 0.367 21.498 22.183 multiply_cannon_loop 2055 14.4 0.341 0.347 19.684 20.036 cp_fm_upper_to_full 70 14.2 12.632 18.024 12.632 18.024 rebuild_ks_matrix 110 8.3 0.000 0.001 14.093 14.184 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.015 14.093 14.183 ot_mini 99 10.5 0.001 0.001 13.558 13.627 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.768 12.847 dbcsr_complete_redistribute 325 12.2 1.027 1.043 7.250 10.349 multiply_cannon_multrec 8220 15.4 4.353 4.564 9.616 9.718 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.222 9.322 qs_ot_get_derivative 99 11.5 0.001 0.001 8.896 8.966 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.666 8.745 mp_waitall_1 84994 16.7 7.457 8.458 7.457 8.458 mp_alltoall_i22 605 13.7 5.299 8.386 5.299 8.386 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.666 7.702 calculate_rho_elec 110 8.6 0.228 0.228 7.665 7.701 sum_up_and_integrate 110 10.3 0.150 0.152 7.289 7.303 integrate_v_rspace 110 11.3 0.004 0.004 7.138 7.152 make_m2s 4110 13.4 0.044 0.044 5.213 5.751 init_scf_run 11 5.9 0.000 0.001 5.716 5.716 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.715 5.715 qs_ot_get_p 110 10.4 0.001 0.001 5.614 5.700 make_images 4110 14.4 0.884 0.934 5.026 5.563 dbcsr_mm_accdrv_process 11614 15.7 3.301 3.672 5.118 5.346 cp_fm_cholesky_invert 11 10.9 5.282 5.286 5.282 5.286 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.728 5.257 apply_single 110 13.6 0.000 0.000 4.727 5.256 pw_transfer 1331 11.6 0.075 0.076 5.030 5.041 multiply_cannon_metrocomm3 8220 15.4 0.018 0.019 4.723 4.995 fft_wrap_pw1pw2 1111 12.6 0.009 0.010 4.913 4.925 ot_diis_step 99 11.5 0.015 0.016 4.643 4.643 fft_wrap_pw1pw2_140 451 13.1 1.341 1.346 4.287 4.302 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.030 4.037 multiply_cannon_sync_h2d 8220 15.4 3.946 3.953 3.946 3.953 density_rs2pw 110 9.6 0.005 0.005 3.708 3.745 grid_integrate_task_list 110 12.3 3.656 3.722 3.656 3.722 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.659 3.660 hybrid_alltoall_any 4261 16.3 0.257 0.550 2.845 3.597 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.587 3.588 make_images_data 4110 15.4 0.043 0.046 2.840 3.483 wfi_extrapolate 11 7.9 0.001 0.001 3.317 3.317 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.836 3.305 calculate_dm_sparse 110 9.5 0.001 0.001 3.111 3.131 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.993 2.993 cp_fm_diag_elpa_base 48 14.0 2.454 2.647 2.991 2.991 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.907 2.911 fft3d_ps 1111 14.6 1.302 1.317 2.837 2.846 grid_collocate_task_list 110 9.6 2.628 2.648 2.628 2.648 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.406 2.436 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.386 2.424 potential_pw2rs 110 12.3 0.022 0.022 2.415 2.422 rs_pw_transfer 902 11.9 0.011 0.011 2.274 2.294 calculate_first_density_matrix 1 7.0 0.001 0.001 2.279 2.279 qs_env_update_s_mstruct 11 6.9 0.001 0.001 2.190 2.255 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.092 2.154 cp_fm_cholesky_decompose 22 10.9 1.992 2.016 1.992 2.016 mp_alltoall_d11v 2046 13.8 1.925 2.013 1.925 2.013 qs_create_task_list 11 7.9 0.000 0.000 1.894 1.942 generate_qs_task_list 11 8.9 0.732 0.786 1.894 1.942 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.927 1.938 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.779 1.835 jit_kernel_multiply 10 15.3 1.619 1.781 1.619 1.781 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=87.045000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1214.727273, yerr=70.005313 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/04dd581a73886059b5d5fb12035b1998b2b62f5a_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430458527744 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 11613083000832 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 94.228655E+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 6806314816 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 631.947264E+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 433.881094E+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 3200148065288 4194304 < size <= 16777216 263934 1154745839544 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4040 57230. MP_Allreduce 11099 798. MP_Sync 87 MP_Alltoall 2242 3393782. 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 0.011 0.027 206.781 206.782 qs_mol_dyn_low 1 2.0 0.003 0.003 206.412 206.422 qs_forces 11 3.9 0.005 0.005 206.318 206.320 qs_energies 11 4.9 0.001 0.002 200.717 200.729 scf_env_do_scf 11 5.9 0.001 0.001 184.201 184.205 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 163.971 163.972 dbcsr_multiply_generic 2527 12.6 0.174 0.178 124.125 124.784 velocity_verlet 10 3.0 0.001 0.001 124.042 124.043 qs_scf_new_mos 118 7.6 0.001 0.001 123.516 123.710 qs_scf_loop_do_ot 118 8.6 0.001 0.001 123.515 123.710 ot_scf_mini 118 9.6 0.003 0.003 116.895 117.103 multiply_cannon 2527 13.6 0.243 0.251 101.139 103.105 multiply_cannon_loop 2527 14.6 2.086 2.138 98.932 100.803 ot_mini 118 10.6 0.001 0.001 64.944 65.215 multiply_cannon_multrec 60648 15.6 33.720 35.914 41.938 43.710 qs_ot_get_derivative 118 11.6 0.001 0.001 40.066 40.246 rebuild_ks_matrix 129 8.3 0.001 0.001 33.984 34.239 qs_ks_build_kohn_sham_matrix 129 9.3 0.016 0.018 33.983 34.238 mp_waitall_1 269254 16.5 27.622 30.966 27.622 30.966 qs_ks_update_qs_env 129 7.6 0.001 0.001 30.552 30.745 qs_ot_get_p 129 10.4 0.001 0.001 29.832 30.112 multiply_cannon_sync_h2d 60648 15.6 27.950 29.439 27.950 29.439 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 24.511 25.014 apply_single 129 13.6 0.001 0.001 24.511 25.014 ot_diis_step 118 11.6 0.007 0.008 24.673 24.674 qs_ot_p2m_diag 83 11.4 0.079 0.091 23.246 23.293 cp_dbcsr_syevd 83 12.4 0.005 0.005 20.626 20.628 init_scf_loop 11 6.9 0.000 0.000 20.156 20.157 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 18.832 18.990 cp_fm_diag_elpa 83 13.4 0.000 0.000 17.664 17.708 cp_fm_diag_elpa_base 83 14.4 17.596 17.640 17.660 17.702 multiply_cannon_metrocomm3 60648 15.6 0.118 0.122 15.503 17.498 prepare_preconditioner 11 7.9 0.000 0.000 15.566 15.613 make_preconditioner 11 8.9 0.000 0.000 15.566 15.613 make_full_inverse_cholesky 11 9.9 0.000 0.000 14.794 14.976 sum_up_and_integrate 129 10.3 0.090 0.111 14.338 14.355 integrate_v_rspace 129 11.3 0.004 0.005 14.248 14.267 make_m2s 5054 13.6 0.105 0.113 13.818 14.169 make_images 5054 14.6 0.410 0.429 13.641 14.008 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.846 14.002 calculate_rho_elec 129 8.7 0.047 0.065 13.845 14.001 init_scf_run 11 5.9 0.000 0.001 12.329 12.329 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.328 12.329 density_rs2pw 129 9.7 0.006 0.007 7.154 10.451 mp_sum_l 7930 13.1 7.887 9.386 7.887 9.386 rs_pw_transfer 1054 12.0 0.017 0.020 5.871 9.179 wfi_extrapolate 11 7.9 0.001 0.001 9.156 9.156 cp_fm_cholesky_invert 11 10.9 8.801 8.809 8.801 8.809 calculate_dm_sparse 129 9.5 0.001 0.001 8.426 8.531 dbcsr_mm_accdrv_process 125468 16.2 3.205 3.318 7.775 8.272 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 8.001 8.118 qs_ot_get_orbitals 118 10.6 0.001 0.001 7.924 7.993 pw_transfer 1559 11.6 0.076 0.088 7.809 7.978 fft_wrap_pw1pw2 1301 12.7 0.011 0.013 7.603 7.773 make_images_data 5054 15.6 0.067 0.073 6.786 7.719 grid_integrate_task_list 129 12.3 7.107 7.632 7.107 7.632 multiply_cannon_metrocomm1 60648 15.6 0.088 0.092 5.580 7.480 hybrid_alltoall_any 5240 16.5 0.298 2.293 5.937 7.162 fft_wrap_pw1pw2_140 527 13.2 1.289 1.334 6.668 6.840 cp_dbcsr_sm_fm_multiply 37 9.5 0.003 0.003 6.634 6.642 mp_alltoall_d11v 2423 14.1 4.226 6.058 4.226 6.058 mp_waitany 16140 13.9 2.741 5.994 2.741 5.994 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.773 5.857 grid_collocate_task_list 129 9.7 4.591 5.822 4.591 5.822 fft3d_ps 1301 14.7 2.150 2.796 5.342 5.664 rs_pw_transfer_RS2PW_140 140 11.5 0.284 0.303 2.148 5.461 potential_pw2rs 129 12.3 0.010 0.011 4.673 4.695 cp_fm_cholesky_decompose 22 10.9 4.581 4.595 4.581 4.595 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=206.782000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=596.363636, yerr=6.568142 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/04dd581a73886059b5d5fb12035b1998b2b62f5a_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 833.478656E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100942E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834332048 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58363. MP_Allreduce 10978 967. MP_Sync 87 MP_Alltoall 1969 5871115. 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 0.017 0.040 188.562 188.562 qs_mol_dyn_low 1 2.0 0.003 0.003 188.190 188.204 qs_forces 11 3.9 0.005 0.012 188.043 188.045 qs_energies 11 4.9 0.002 0.003 181.254 181.264 scf_env_do_scf 11 5.9 0.001 0.001 164.953 164.963 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 132.493 132.495 velocity_verlet 10 3.0 0.001 0.001 118.741 118.744 dbcsr_multiply_generic 2507 12.6 0.194 0.204 97.113 98.265 qs_scf_new_mos 117 7.6 0.001 0.001 92.547 93.072 qs_scf_loop_do_ot 117 8.6 0.001 0.001 92.547 93.071 ot_scf_mini 117 9.6 0.004 0.004 87.759 88.321 multiply_cannon 2507 13.6 0.474 0.527 77.377 81.410 multiply_cannon_loop 2507 14.6 1.255 1.292 74.104 76.546 ot_mini 117 10.6 0.001 0.001 49.376 49.899 mp_waitall_1 214728 16.6 24.178 39.214 24.178 39.214 multiply_cannon_multrec 30084 15.6 22.034 26.966 31.745 36.891 rebuild_ks_matrix 128 8.3 0.001 0.001 33.312 33.882 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.019 33.311 33.881 init_scf_loop 11 6.9 0.000 0.000 32.374 32.375 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.976 30.492 multiply_cannon_metrocomm3 30084 15.6 0.093 0.100 15.656 29.149 qs_ot_get_derivative 117 11.6 0.001 0.002 27.516 28.073 prepare_preconditioner 11 7.9 0.000 0.000 27.964 28.035 make_preconditioner 11 8.9 0.000 0.000 27.964 28.035 make_full_inverse_cholesky 11 9.9 0.000 0.000 26.655 27.173 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 21.915 23.036 apply_single 128 13.6 0.001 0.001 21.915 23.036 multiply_cannon_sync_h2d 30084 15.6 19.278 21.878 19.278 21.878 ot_diis_step 117 11.6 0.014 0.015 21.680 21.682 qs_ot_get_p 128 10.4 0.001 0.001 20.163 20.798 cp_fm_cholesky_invert 11 10.9 16.336 16.348 16.336 16.348 make_m2s 5014 13.6 0.092 0.096 14.043 15.683 make_images 5014 14.6 1.162 1.355 13.841 15.482 qs_ot_p2m_diag 83 11.4 0.188 0.216 15.415 15.447 sum_up_and_integrate 128 10.3 0.119 0.147 14.635 14.667 integrate_v_rspace 128 11.3 0.004 0.005 14.516 14.553 cp_dbcsr_syevd 83 12.4 0.005 0.006 14.288 14.289 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.937 13.969 calculate_rho_elec 128 8.7 0.089 0.106 13.936 13.969 init_scf_run 11 5.9 0.000 0.001 11.556 11.558 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.556 11.557 cp_fm_diag_elpa 83 13.4 0.000 0.001 11.201 11.228 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 10.813 11.226 cp_fm_diag_elpa_base 83 14.4 10.948 11.022 11.196 11.225 multiply_cannon_metrocomm4 27577 15.6 0.099 0.115 3.738 10.600 make_images_data 5014 15.6 0.065 0.073 8.559 10.468 hybrid_alltoall_any 5200 16.5 0.343 1.495 7.297 10.213 mp_irecv_dv 69486 16.3 3.542 10.203 3.542 10.203 density_rs2pw 128 9.7 0.006 0.007 7.325 9.835 dbcsr_mm_accdrv_process 62242 16.2 4.515 5.288 9.166 9.733 pw_transfer 1547 11.6 0.086 0.098 8.710 8.777 fft_wrap_pw1pw2 1291 12.7 0.011 0.011 8.484 8.552 wfi_extrapolate 11 7.9 0.001 0.001 8.340 8.340 rs_pw_transfer 1046 11.9 0.015 0.017 5.563 8.079 fft_wrap_pw1pw2_140 523 13.2 1.347 1.364 7.534 7.624 grid_integrate_task_list 128 12.3 7.189 7.594 7.189 7.594 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.138 6.915 cp_fm_cholesky_decompose 22 10.9 6.768 6.864 6.768 6.864 calculate_dm_sparse 128 9.5 0.001 0.001 6.464 6.622 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.215 6.223 grid_collocate_task_list 128 9.7 4.703 5.875 4.703 5.875 mp_sum_l 7870 13.0 3.900 5.848 3.900 5.848 fft3d_ps 1291 14.7 2.809 2.977 5.774 5.817 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.376 5.521 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.270 5.340 mp_waitany 11748 13.9 2.467 4.975 2.467 4.975 mp_alltoall_d11v 2415 14.1 4.055 4.960 4.055 4.960 potential_pw2rs 128 12.3 0.015 0.018 4.858 4.873 mp_allgather_i34 2507 14.6 1.839 4.806 1.839 4.806 rs_pw_transfer_RS2PW_140 139 11.5 0.354 0.375 2.070 4.555 mp_sum_d 4465 12.1 2.615 3.957 2.615 3.957 dbcsr_complete_redistribute 395 12.7 0.774 0.850 3.069 3.896 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.726 3.776 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=188.562000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=793.000000, yerr=3.275252 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/04dd581a73886059b5d5fb12035b1998b2b62f5a_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 2.928533E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 3984192 0.0% 0.0% 100.0% average stack size 0.0 0.0 1695.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 933.990400E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1042912 MPI messages size (bytes): total size 2.716210E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.604448E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 281856 36943429632 131072 < size <= 4194304 660064 996105256960 4194304 < size <= 16777216 65632 931530938576 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58357. MP_Allreduce 10977 1007. 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 0.015 0.035 173.974 173.975 qs_mol_dyn_low 1 2.0 0.003 0.003 173.405 173.416 qs_forces 11 3.9 0.004 0.005 173.286 173.292 qs_energies 11 4.9 0.001 0.002 166.688 166.699 scf_env_do_scf 11 5.9 0.001 0.001 151.391 151.392 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 116.966 116.967 velocity_verlet 10 3.0 0.001 0.001 111.768 111.770 dbcsr_multiply_generic 2507 12.6 0.178 0.185 81.251 82.414 qs_scf_new_mos 117 7.6 0.001 0.001 78.780 79.086 qs_scf_loop_do_ot 117 8.6 0.001 0.001 78.779 79.085 ot_scf_mini 117 9.6 0.003 0.004 74.670 75.034 multiply_cannon 2507 13.6 0.505 0.531 61.708 66.440 multiply_cannon_loop 2507 14.6 0.859 0.880 58.616 61.345 ot_mini 117 10.6 0.001 0.001 42.324 42.702 mp_waitall_1 170520 16.6 24.820 35.292 24.820 35.292 init_scf_loop 11 6.9 0.000 0.000 34.329 34.330 rebuild_ks_matrix 128 8.3 0.001 0.001 30.720 31.146 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.018 30.719 31.145 prepare_preconditioner 11 7.9 0.000 0.000 30.301 30.345 make_preconditioner 11 8.9 0.000 0.000 30.301 30.345 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.927 29.350 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.665 28.061 multiply_cannon_metrocomm3 20056 15.6 0.059 0.064 15.167 25.440 multiply_cannon_multrec 20056 15.6 13.491 16.549 22.139 25.229 qs_ot_get_derivative 117 11.6 0.002 0.002 22.674 23.046 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.789 20.844 apply_single 128 13.6 0.001 0.001 19.789 20.844 ot_diis_step 117 11.6 0.017 0.018 19.549 19.549 qs_ot_get_p 128 10.4 0.001 0.001 17.736 18.235 multiply_cannon_sync_h2d 20056 15.6 14.309 15.798 14.309 15.798 make_m2s 5014 13.6 0.084 0.091 14.544 15.592 make_images 5014 14.6 1.178 1.269 14.314 15.363 sum_up_and_integrate 128 10.3 0.134 0.146 14.564 14.592 integrate_v_rspace 128 11.3 0.004 0.005 14.430 14.462 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.274 14.298 calculate_rho_elec 128 8.7 0.132 0.148 14.273 14.297 cp_fm_cholesky_invert 11 10.9 13.999 14.008 13.999 14.008 qs_ot_p2m_diag 83 11.4 0.266 0.273 13.506 13.514 cp_dbcsr_syevd 83 12.4 0.005 0.006 12.521 12.522 make_images_data 5014 15.6 0.062 0.070 8.985 10.394 init_scf_run 11 5.9 0.000 0.001 10.332 10.332 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.332 10.332 hybrid_alltoall_any 5200 16.5 0.436 1.985 7.694 9.803 cp_fm_diag_elpa 83 13.4 0.000 0.001 9.412 9.427 cp_fm_diag_elpa_base 83 14.4 9.008 9.161 9.409 9.424 multiply_cannon_metrocomm4 17549 15.6 0.063 0.075 3.422 9.275 density_rs2pw 128 9.7 0.007 0.007 7.102 9.177 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 8.884 9.147 mp_irecv_dv 50230 16.2 3.299 9.034 3.299 9.034 pw_transfer 1547 11.6 0.086 0.104 8.822 8.932 fft_wrap_pw1pw2 1291 12.7 0.011 0.012 8.597 8.712 dbcsr_mm_accdrv_process 41502 16.2 4.407 5.216 8.097 8.199 grid_integrate_task_list 128 12.3 7.306 7.833 7.306 7.833 fft_wrap_pw1pw2_140 523 13.2 1.429 1.461 7.659 7.791 wfi_extrapolate 11 7.9 0.001 0.001 7.393 7.393 cp_fm_upper_to_full 105 14.8 5.730 7.292 5.730 7.292 cp_fm_cholesky_decompose 22 10.9 7.162 7.190 7.162 7.190 rs_pw_transfer 1046 11.9 0.014 0.015 5.027 7.110 dbcsr_complete_redistribute 395 12.7 1.171 1.197 4.540 6.328 calculate_dm_sparse 128 9.5 0.001 0.001 5.703 5.837 fft3d_ps 1291 14.7 2.736 2.967 5.693 5.781 grid_collocate_task_list 128 9.7 4.901 5.720 4.901 5.720 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.384 5.389 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.529 5.253 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.397 5.180 mp_alltoall_d11v 2415 14.1 4.304 5.075 4.304 5.075 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.518 4.644 potential_pw2rs 128 12.3 0.021 0.023 4.627 4.636 mp_sum_l 7870 13.0 3.179 4.556 3.179 4.556 mp_waitany 11748 13.9 2.315 4.428 2.315 4.428 mp_allgather_i34 2507 14.6 1.599 4.342 1.599 4.342 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.354 4.082 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.970 3.996 mp_alltoall_i22 716 14.1 1.997 3.977 1.997 3.977 rs_pw_transfer_RS2PW_140 139 11.5 0.331 0.352 1.899 3.964 qs_energies_init_hamiltonians 11 5.9 0.009 0.024 3.751 3.751 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.553 3.589 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=173.975000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=888.545455, yerr=5.416534 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/04dd581a73886059b5d5fb12035b1998b2b62f5a_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 4.320339E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 5927808 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.2 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.128997E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1133160 MPI messages size (bytes): total size 2.008142E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.772161E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 315952 35695099904 131072 < size <= 4194304 709496 778939400192 4194304 < size <= 16777216 69840 660837692480 16777216 < size 30480 532676608000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4007 58138. MP_Allreduce 11015 1090. MP_Sync 86 MP_Alltoall 1700 12496371. MP_SendRecv 5842 75008. MP_ISendRecv 5842 75008. MP_Wait 22272 MP_ISend 14840 244848. MP_IRecv 14840 244848. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.037 0.101 185.988 185.990 qs_mol_dyn_low 1 2.0 0.003 0.004 185.535 185.548 qs_forces 11 3.9 0.005 0.009 185.433 185.439 qs_energies 11 4.9 0.003 0.008 178.285 178.292 scf_env_do_scf 11 5.9 0.001 0.004 161.573 161.580 velocity_verlet 10 3.0 0.001 0.001 122.879 122.883 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 115.627 115.629 dbcsr_multiply_generic 2485 12.5 0.185 0.192 78.611 79.673 qs_scf_new_mos 116 7.6 0.001 0.001 78.457 78.762 qs_scf_loop_do_ot 116 8.6 0.001 0.001 78.456 78.762 ot_scf_mini 116 9.6 0.003 0.004 74.033 74.337 multiply_cannon 2485 13.5 0.551 0.593 54.457 57.527 multiply_cannon_loop 2485 14.5 1.171 1.200 50.658 53.064 init_scf_loop 11 6.9 0.001 0.003 45.823 45.824 ot_mini 116 10.6 0.001 0.001 42.007 42.298 prepare_preconditioner 11 7.9 0.000 0.000 41.730 41.752 make_preconditioner 11 8.9 0.001 0.005 41.730 41.752 make_full_inverse_cholesky 11 9.9 0.000 0.001 35.421 40.450 multiply_cannon_multrec 29820 15.5 13.939 19.036 25.794 30.468 rebuild_ks_matrix 127 8.3 0.001 0.001 29.664 30.001 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.020 29.664 30.000 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.699 27.020 mp_waitall_1 146592 16.7 17.079 26.014 17.079 26.014 qs_ot_get_derivative 116 11.6 0.001 0.002 22.491 22.789 make_m2s 4970 13.5 0.098 0.102 19.795 21.062 make_images 4970 14.5 1.925 2.288 19.492 20.760 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 18.889 19.624 apply_single 127 13.6 0.001 0.001 18.889 19.624 ot_diis_step 116 11.6 0.017 0.018 19.396 19.398 qs_ot_get_p 127 10.4 0.001 0.002 18.467 18.754 cp_fm_upper_to_full 104 14.8 11.451 16.871 11.451 16.871 cp_fm_cholesky_invert 11 10.9 15.814 15.822 15.814 15.822 multiply_cannon_metrocomm3 29820 15.5 0.047 0.049 6.422 15.497 sum_up_and_integrate 127 10.3 0.139 0.150 14.589 14.614 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.472 14.497 calculate_rho_elec 127 8.7 0.175 0.191 14.472 14.496 integrate_v_rspace 127 11.3 0.004 0.005 14.449 14.479 qs_ot_p2m_diag 82 11.4 0.339 0.385 14.342 14.395 cp_dbcsr_syevd 82 12.4 0.007 0.021 13.093 13.095 multiply_cannon_sync_h2d 29820 15.5 11.687 12.891 11.687 12.891 dbcsr_complete_redistribute 393 12.7 1.506 1.631 8.860 12.623 make_images_data 4970 15.5 0.064 0.070 10.647 12.529 dbcsr_mm_accdrv_process 61748 16.2 7.395 8.479 11.437 11.960 hybrid_alltoall_any 5155 16.4 0.523 2.199 9.472 11.620 copy_fm_to_dbcsr 208 11.6 0.002 0.002 7.506 11.298 init_scf_run 11 5.9 0.000 0.001 10.841 10.842 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.841 10.842 cp_fm_diag_elpa 82 13.4 0.001 0.002 10.056 10.069 cp_fm_diag_elpa_base 82 14.4 9.100 9.415 10.047 10.062 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.286 9.968 qs_ot_get_derivative_diag 76 12.4 0.002 0.003 9.265 9.465 mp_alltoall_i22 712 14.1 5.522 9.421 5.522 9.421 pw_transfer 1535 11.6 0.086 0.102 9.067 9.146 fft_wrap_pw1pw2 1281 12.7 0.011 0.011 8.842 8.925 density_rs2pw 127 9.7 0.006 0.007 6.946 8.518 fft_wrap_pw1pw2_140 519 13.2 1.561 1.591 7.876 7.975 grid_integrate_task_list 127 12.3 7.455 7.844 7.455 7.844 wfi_extrapolate 11 7.9 0.001 0.001 7.677 7.677 cp_fm_cholesky_decompose 22 10.9 7.424 7.499 7.424 7.499 multiply_cannon_metrocomm4 24850 15.5 0.082 0.095 2.755 6.638 mp_irecv_dv 75445 16.2 2.605 6.367 2.605 6.367 rs_pw_transfer 1038 11.9 0.013 0.015 4.620 6.277 calculate_dm_sparse 127 9.5 0.001 0.001 6.102 6.213 fft3d_ps 1281 14.7 2.809 2.883 5.774 5.843 grid_collocate_task_list 127 9.7 5.026 5.722 5.026 5.722 mp_alltoall_d11v 2401 14.1 4.857 5.574 4.857 5.574 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.475 5.540 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 4.394 4.479 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.363 4.473 potential_pw2rs 127 12.3 0.023 0.024 4.448 4.464 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 4.452 4.453 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.086 4.140 mp_waitany 14840 13.9 2.227 3.868 2.227 3.868 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=185.990000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1070.272727, yerr=16.498435 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/04dd581a73886059b5d5fb12035b1998b2b62f5a_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 5.865088E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 1960712 0.0% 0.0% 100.0% average stack size 0.0 0.0 3445.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.512587E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 240672 MPI messages size (bytes): total size 1.331455E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.532237E+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 113904 59718500352 4194304 < size <= 16777216 104976 550376570880 16777216 < size 20208 721350092304 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8931 51. MP_Alltoall 9654 799394. MP_ISend 40068 2102572. MP_IRecv 40068 2101675. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58352. MP_Allreduce 10977 1175. MP_Sync 87 MP_Alltoall 1712 18838210. MP_SendRecv 3840 122880. MP_ISendRecv 3840 122880. MP_Wait 16122 MP_ISend 10680 423556. MP_IRecv 10680 423556. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.043 0.071 173.373 173.380 qs_mol_dyn_low 1 2.0 0.003 0.003 172.951 172.964 qs_forces 11 3.9 0.004 0.005 172.807 172.814 qs_energies 11 4.9 0.002 0.002 165.238 165.247 scf_env_do_scf 11 5.9 0.001 0.001 148.056 148.066 velocity_verlet 10 3.0 0.001 0.001 113.517 113.520 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 112.624 112.625 dbcsr_multiply_generic 2507 12.6 0.179 0.191 74.091 74.508 qs_scf_new_mos 117 7.6 0.001 0.001 74.041 74.125 qs_scf_loop_do_ot 117 8.6 0.001 0.001 74.040 74.124 ot_scf_mini 117 9.6 0.003 0.004 69.607 69.674 multiply_cannon 2507 13.6 0.587 0.621 54.345 59.406 multiply_cannon_loop 2507 14.6 0.445 0.453 49.738 50.444 ot_mini 117 10.6 0.001 0.001 39.553 39.614 init_scf_loop 11 6.9 0.000 0.000 35.285 35.286 mp_waitall_1 125778 16.7 25.963 33.222 25.963 33.222 prepare_preconditioner 11 7.9 0.000 0.000 31.346 31.371 make_preconditioner 11 8.9 0.000 0.000 31.346 31.371 rebuild_ks_matrix 128 8.3 0.001 0.001 29.908 29.955 qs_ks_build_kohn_sham_matrix 128 9.3 0.018 0.019 29.907 29.955 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.219 29.495 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.157 27.206 multiply_cannon_multrec 10028 15.6 10.392 14.358 17.650 20.393 ot_diis_step 117 11.6 0.020 0.022 19.936 19.936 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.484 19.755 apply_single 128 13.6 0.001 0.001 19.484 19.755 qs_ot_get_derivative 117 11.6 0.002 0.002 19.551 19.618 multiply_cannon_metrocomm3 10028 15.6 0.023 0.025 12.324 19.344 make_m2s 5014 13.6 0.070 0.073 16.215 18.483 make_images 5014 14.6 2.302 2.590 15.910 18.174 cp_fm_cholesky_invert 11 10.9 17.848 17.853 17.848 17.853 qs_ot_get_p 128 10.4 0.001 0.001 17.162 17.236 qs_rho_update_rho_low 128 7.7 0.001 0.001 15.294 15.344 calculate_rho_elec 128 8.7 0.259 0.270 15.293 15.344 sum_up_and_integrate 128 10.3 0.180 0.190 15.012 15.061 integrate_v_rspace 128 11.3 0.004 0.005 14.831 14.888 qs_ot_p2m_diag 83 11.4 0.496 0.501 13.459 13.478 make_images_data 5014 15.6 0.054 0.062 9.894 12.532 cp_dbcsr_syevd 83 12.4 0.005 0.006 12.342 12.345 hybrid_alltoall_any 5200 16.5 0.794 3.642 9.731 12.269 multiply_cannon_sync_h2d 10028 15.6 11.655 12.045 11.655 12.045 init_scf_run 11 5.9 0.000 0.001 10.343 10.343 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.343 10.343 pw_transfer 1547 11.6 0.087 0.097 9.486 9.509 cp_fm_diag_elpa 83 13.4 0.000 0.000 9.354 9.365 cp_fm_diag_elpa_base 83 14.4 9.101 9.191 9.347 9.358 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 9.262 9.291 grid_integrate_task_list 128 12.3 7.729 8.226 7.729 8.226 fft_wrap_pw1pw2_140 523 13.2 1.931 1.957 8.191 8.224 cp_fm_cholesky_decompose 22 10.9 7.829 7.948 7.829 7.948 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 7.872 7.923 mp_allgather_i34 2507 14.6 2.797 7.847 2.797 7.847 density_rs2pw 128 9.7 0.006 0.006 6.866 7.757 dbcsr_mm_accdrv_process 20762 16.1 2.664 3.540 6.888 7.499 multiply_cannon_metrocomm1 10028 15.6 0.028 0.029 4.470 7.375 wfi_extrapolate 11 7.9 0.001 0.001 7.374 7.374 calculate_dm_sparse 128 9.5 0.001 0.001 6.138 6.228 grid_collocate_task_list 128 9.7 5.394 6.078 5.394 6.078 mp_alltoall_d11v 2415 14.1 5.029 5.797 5.029 5.797 fft3d_ps 1291 14.7 2.770 2.830 5.714 5.746 dbcsr_complete_redistribute 395 12.7 2.125 2.186 5.122 5.533 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 5.319 5.321 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.140 5.148 rs_pw_transfer 1046 11.9 0.013 0.013 4.159 5.082 potential_pw2rs 128 12.3 0.027 0.028 4.427 4.440 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.069 4.095 multiply_cannon_metrocomm4 7521 15.6 0.024 0.027 1.852 3.965 mp_irecv_dv 28860 15.9 1.816 3.889 1.816 3.889 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.554 3.871 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.668 3.700 copy_fm_to_dbcsr 209 11.7 0.001 0.002 3.377 3.699 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.662 3.668 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.499 3.580 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=173.380000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1415.272727, yerr=50.634323 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/04dd581a73886059b5d5fb12035b1998b2b62f5a_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430454546432 0.0% 0.0% 100.0% flops 32 x 32 x 32 1975684956160 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992006770688 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 11613065416704 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239182565376 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239182565376 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 94.245913E+12 0.0% 0.0% 100.0% flops max/rank 11.787674E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806580192 0.0% 0.0% 100.0% number of processed stacks 1982496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3433.3 marketing flops 145.663816E+12 ------------------------------------------------------------------------------- # multiplications 2535 max memory usage/rank 3.137581E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101400 MPI messages size (bytes): total size 1.145171E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.293599E+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 45888 35634806784 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592540656 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4041 58592. MP_Allreduce 11099 1511. MP_Sync 88 MP_Alltoall 1724 36993632. MP_SendRecv 1806 218624. MP_ISendRecv 1806 218624. MP_Wait 9876 MP_ISend 6456 1080169. MP_IRecv 6456 1080169. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.025 0.040 298.361 298.363 qs_mol_dyn_low 1 2.0 0.003 0.003 297.838 297.851 qs_forces 11 3.9 0.005 0.005 297.741 297.745 qs_energies 11 4.9 0.002 0.002 288.388 288.401 scf_env_do_scf 11 5.9 0.001 0.001 266.150 266.160 velocity_verlet 10 3.0 0.001 0.001 215.170 215.178 scf_env_do_scf_inner_loop 118 6.6 0.005 0.009 141.521 141.523 init_scf_loop 11 6.9 0.000 0.000 124.368 124.370 prepare_preconditioner 11 7.9 0.000 0.000 119.334 119.360 make_preconditioner 11 8.9 0.000 0.000 119.334 119.360 make_full_inverse_cholesky 11 9.9 0.000 0.000 95.365 116.490 qs_scf_new_mos 118 7.6 0.001 0.001 90.441 90.565 qs_scf_loop_do_ot 118 8.6 0.001 0.001 90.440 90.564 ot_scf_mini 118 9.6 0.004 0.004 85.634 85.667 dbcsr_multiply_generic 2535 12.6 0.214 0.226 83.443 83.990 cp_fm_upper_to_full 106 14.8 53.396 76.336 53.396 76.336 multiply_cannon 2535 13.6 0.703 0.739 58.832 59.293 multiply_cannon_loop 2535 14.6 0.476 0.483 55.227 56.585 ot_mini 118 10.6 0.001 0.001 44.445 44.476 dbcsr_complete_redistribute 397 12.7 3.996 4.045 29.631 42.553 copy_fm_to_dbcsr 210 11.7 0.002 0.002 26.229 39.127 rebuild_ks_matrix 129 8.3 0.001 0.001 38.519 38.558 qs_ks_build_kohn_sham_matrix 129 9.3 0.020 0.020 38.518 38.557 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 23.924 36.638 qs_ks_update_qs_env 129 7.6 0.001 0.001 35.534 35.575 mp_alltoall_i22 720 14.1 21.699 34.746 21.699 34.746 cp_fm_cholesky_invert 11 10.9 33.003 33.009 33.003 33.009 mp_waitall_1 104820 16.8 28.208 32.198 28.208 32.198 qs_ot_get_p 129 10.4 0.001 0.001 25.633 25.639 qs_ot_get_derivative 118 11.6 0.002 0.002 24.233 24.266 qs_ot_p2m_diag 84 11.4 0.889 0.895 21.593 21.622 make_m2s 5070 13.6 0.080 0.082 20.100 20.973 qs_rho_update_rho_low 129 7.7 0.001 0.001 20.598 20.618 calculate_rho_elec 129 8.7 0.487 0.488 20.597 20.618 make_images 5070 14.6 3.806 3.986 19.619 20.493 ot_diis_step 118 11.6 0.022 0.022 20.188 20.188 multiply_cannon_metrocomm3 10140 15.6 0.024 0.024 18.746 19.900 cp_dbcsr_syevd 84 12.4 0.006 0.006 19.867 19.869 sum_up_and_integrate 129 10.3 0.324 0.326 19.743 19.821 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.373 19.558 apply_single 129 13.6 0.001 0.001 19.372 19.557 integrate_v_rspace 129 11.3 0.004 0.004 19.419 19.499 multiply_cannon_multrec 10140 15.6 10.625 12.471 18.131 18.208 cp_fm_diag_elpa 84 13.4 0.000 0.000 16.666 16.666 cp_fm_diag_elpa_base 84 14.4 12.204 13.815 16.662 16.662 multiply_cannon_sync_h2d 10140 15.6 15.777 15.790 15.777 15.790 hybrid_alltoall_any 5257 16.5 1.313 3.078 10.927 12.814 make_images_data 5070 15.6 0.063 0.067 10.738 12.446 init_scf_run 11 5.9 0.000 0.001 12.141 12.142 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.141 12.142 pw_transfer 1559 11.6 0.094 0.095 11.450 11.455 fft_wrap_pw1pw2 1301 12.7 0.012 0.012 11.214 11.220 fft_wrap_pw1pw2_140 527 13.2 3.107 3.135 9.967 9.973 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 9.509 9.543 dbcsr_mm_accdrv_process 20958 16.1 3.960 5.788 7.267 9.170 mp_alltoall_d11v 2429 14.1 8.218 9.058 8.218 9.058 wfi_extrapolate 11 7.9 0.001 0.001 9.040 9.040 cp_fm_cholesky_decompose 22 10.9 8.809 8.842 8.809 8.842 grid_integrate_task_list 129 12.3 8.655 8.827 8.655 8.827 density_rs2pw 129 9.7 0.006 0.006 7.998 8.124 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 8.075 8.076 calculate_dm_sparse 129 9.5 0.001 0.001 6.656 6.764 grid_collocate_task_list 129 9.7 6.415 6.458 6.415 6.458 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.297 6.389 rs_scatter_matrices 140 9.7 3.819 4.806 6.158 6.387 fft3d_ps 1301 14.7 2.861 2.871 6.234 6.264 copy_dbcsr_to_fm 187 11.8 0.004 0.004 6.072 6.144 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=298.363000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2821.090909, yerr=179.786972 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/04dd581a73886059b5d5fb12035b1998b2b62f5a_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 1.262129E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255669. MP_Allreduce 3059 6274. MP_Sync 4 MP_Alltoall 54 26147635. 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 0.016 0.031 84.980 84.981 qs_energies 1 2.0 0.000 0.000 84.570 84.576 ls_scf 1 3.0 0.000 0.000 83.668 83.674 dbcsr_multiply_generic 111 6.7 0.015 0.016 72.801 72.985 multiply_cannon 111 7.7 0.017 0.020 55.976 57.382 multiply_cannon_loop 111 8.7 0.208 0.222 52.522 53.926 ls_scf_main 1 4.0 0.000 0.000 52.404 52.415 density_matrix_trs4 2 5.0 0.002 0.003 46.890 47.009 ls_scf_init_scf 1 4.0 0.000 0.000 28.204 28.205 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.119 27.173 mp_waitall_1 11031 10.9 22.646 25.833 22.646 25.833 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.023 25.045 multiply_cannon_multrec 2664 9.7 8.207 8.984 15.516 17.229 multiply_cannon_sync_h2d 2664 9.7 13.514 14.961 13.514 14.961 make_m2s 222 7.7 0.009 0.012 13.056 13.611 make_images 222 8.7 0.102 0.112 13.034 13.590 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 9.862 12.007 make_images_data 222 9.7 0.005 0.006 7.616 8.113 dbcsr_mm_accdrv_process 4760 10.4 0.510 0.641 6.932 7.880 multiply_cannon_metrocomm3 2664 9.7 0.010 0.010 5.505 7.795 hybrid_alltoall_any 227 10.6 0.218 1.833 6.542 7.790 dbcsr_mm_accdrv_process_sort 4760 11.4 6.223 7.113 6.223 7.113 calculate_norms 4752 9.8 5.492 6.096 5.492 6.096 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.061 5.168 mp_sum_l 807 5.4 3.229 4.652 3.229 4.652 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.442 3.471 multiply_cannon_metrocomm4 2442 9.7 0.011 0.014 2.045 3.359 mp_irecv_dv 6231 10.9 2.028 3.332 2.028 3.332 make_images_sizes 222 9.7 0.000 0.000 0.680 3.320 mp_alltoall_i44 222 10.7 0.679 3.320 0.679 3.320 arnoldi_extremal 4 6.8 0.000 0.000 3.209 3.242 arnoldi_normal_ev 4 7.8 0.001 0.002 3.209 3.242 build_subspace 16 8.4 0.009 0.012 3.117 3.120 ls_scf_post 1 4.0 0.000 0.000 3.061 3.067 ls_scf_store_result 1 5.0 0.000 0.000 2.878 2.923 dbcsr_special_finalize 555 9.7 0.006 0.007 2.396 2.810 dbcsr_merge_single_wm 555 10.7 0.460 0.591 2.387 2.801 make_images_pack 222 9.7 2.214 2.647 2.216 2.648 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.327 2.559 dbcsr_sort_data 658 11.4 2.181 2.527 2.181 2.527 dbcsr_matrix_vector_mult_local 304 10.0 2.066 2.458 2.068 2.460 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.261 2.349 buffer_matrices_ensure_size 222 8.7 1.758 2.106 1.758 2.106 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.773 1.774 rebuild_ks_matrix 3 7.3 0.000 0.000 1.764 1.765 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.764 1.765 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=84.981000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1134.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/04dd581a73886059b5d5fb12035b1998b2b62f5a_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 2.119303E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 10339. MP_Sync 4 MP_Alltoall 47 15335933. MP_SendRecv 141 57600. MP_ISendRecv 141 57600. MP_Wait 687 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.043 0.079 91.327 91.327 qs_energies 1 2.0 0.000 0.000 90.838 90.849 ls_scf 1 3.0 0.001 0.007 89.511 89.521 dbcsr_multiply_generic 111 6.7 0.016 0.017 75.382 75.784 multiply_cannon 111 7.7 0.029 0.043 53.474 56.912 ls_scf_main 1 4.0 0.000 0.003 55.139 55.144 multiply_cannon_loop 111 8.7 0.117 0.124 50.133 53.256 density_matrix_trs4 2 5.0 0.002 0.004 49.404 49.595 ls_scf_init_scf 1 4.0 0.001 0.006 30.809 30.811 mp_waitall_1 9105 10.9 21.154 30.206 21.154 30.206 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.584 29.717 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 27.146 27.162 multiply_cannon_multrec 1332 9.7 13.312 16.185 22.593 26.424 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.747 20.969 make_m2s 222 7.7 0.008 0.008 15.387 16.079 make_images 222 8.7 1.581 1.977 15.357 16.049 dbcsr_mm_accdrv_process 4041 10.4 0.329 0.454 8.872 10.492 dbcsr_mm_accdrv_process_sort 4041 11.4 8.460 10.038 8.460 10.038 make_images_data 222 9.7 0.004 0.005 8.854 9.886 hybrid_alltoall_any 227 10.6 0.521 2.482 8.316 9.293 mp_sum_l 807 5.4 5.190 7.910 5.190 7.910 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.236 7.773 mp_irecv_dv 3311 11.0 3.216 7.706 3.216 7.706 calculate_norms 2376 9.8 6.004 6.793 6.004 6.793 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.985 6.228 multiply_cannon_sync_h2d 1332 9.7 4.790 6.124 4.790 6.124 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.064 5.249 arnoldi_extremal 4 6.8 0.000 0.002 4.635 4.660 arnoldi_normal_ev 4 7.8 0.001 0.006 4.635 4.659 build_subspace 16 8.4 0.014 0.021 4.378 4.383 ls_scf_post 1 4.0 0.002 0.012 3.561 3.572 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.142 3.370 ls_scf_store_result 1 5.0 0.000 0.000 3.183 3.295 dbcsr_matrix_vector_mult_local 304 10.0 2.763 3.238 2.766 3.239 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.572 2.669 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.188 2.573 mp_allgather_i34 111 8.7 1.029 2.512 1.029 2.512 make_images_pack 222 9.7 2.036 2.453 2.039 2.456 dbcsr_sort_data 436 11.2 1.816 2.065 1.816 2.065 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.879 1.881 rebuild_ks_matrix 3 7.3 0.000 0.000 1.866 1.868 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.866 1.868 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.327000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1729.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/04dd581a73886059b5d5fb12035b1998b2b62f5a_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 2.763284E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265470. MP_Allreduce 3058 11181. MP_Sync 4 MP_Alltoall 47 23526250. MP_SendRecv 93 57600. MP_ISendRecv 93 57600. MP_Wait 639 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.030 0.044 94.468 94.469 qs_energies 1 2.0 0.000 0.000 93.916 93.920 ls_scf 1 3.0 0.000 0.000 92.518 92.522 dbcsr_multiply_generic 111 6.7 0.016 0.017 77.183 77.476 ls_scf_main 1 4.0 0.000 0.000 57.967 57.973 multiply_cannon 111 7.7 0.037 0.064 53.326 57.356 multiply_cannon_loop 111 8.7 0.100 0.105 49.775 53.790 density_matrix_trs4 2 5.0 0.002 0.003 52.049 52.226 mp_waitall_1 7281 11.0 24.215 34.743 24.215 34.743 ls_scf_init_scf 1 4.0 0.000 0.000 30.987 30.990 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.805 29.882 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.379 27.394 multiply_cannon_multrec 888 9.7 12.628 15.386 21.197 24.522 multiply_cannon_metrocomm3 888 9.7 0.004 0.005 11.385 24.172 make_m2s 222 7.7 0.007 0.008 17.109 18.329 make_images 222 8.7 1.976 2.301 17.071 18.288 hybrid_alltoall_any 227 10.6 0.623 2.863 9.399 10.879 make_images_data 222 9.7 0.004 0.005 9.684 10.698 dbcsr_mm_accdrv_process 3754 10.4 0.251 0.430 8.091 9.316 mp_sum_l 807 5.4 5.174 9.098 5.174 9.098 dbcsr_mm_accdrv_process_sort 3754 11.4 7.715 8.885 7.715 8.885 multiply_cannon_sync_h2d 888 9.7 6.021 7.418 6.021 7.418 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.020 7.292 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.459 6.724 mp_irecv_dv 2335 11.1 2.444 6.687 2.444 6.687 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.772 6.563 arnoldi_extremal 4 6.8 0.000 0.000 5.078 5.098 arnoldi_normal_ev 4 7.8 0.001 0.004 5.078 5.098 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.872 5.091 build_subspace 16 8.4 0.014 0.020 4.777 4.784 calculate_norms 1584 9.8 4.308 4.689 4.308 4.689 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.451 3.785 mp_allgather_i34 111 8.7 1.385 3.728 1.385 3.728 dbcsr_matrix_vector_mult_local 304 10.0 3.038 3.614 3.040 3.616 ls_scf_post 1 4.0 0.000 0.000 3.563 3.568 ls_scf_store_result 1 5.0 0.000 0.000 3.302 3.391 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.822 2.920 make_images_sizes 222 9.7 0.000 0.000 1.115 2.289 mp_alltoall_i44 222 10.7 1.114 2.289 1.114 2.289 dbcsr_sort_data 325 11.1 1.862 2.157 1.862 2.157 make_images_pack 222 9.7 1.814 2.118 1.817 2.121 dbcsr_data_release 9322 10.9 1.326 1.968 1.326 1.968 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.904 1.906 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=94.469000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2198.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/04dd581a73886059b5d5fb12035b1998b2b62f5a_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 3.343741E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 13371. MP_Sync 4 MP_Alltoall 47 30278988. MP_SendRecv 69 86400. MP_ISendRecv 69 86400. MP_Wait 531 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.042 0.094 97.117 97.118 qs_energies 1 2.0 0.000 0.000 96.415 96.420 ls_scf 1 3.0 0.000 0.001 94.588 94.595 dbcsr_multiply_generic 111 6.7 0.018 0.023 78.367 78.635 ls_scf_main 1 4.0 0.000 0.002 58.687 58.688 multiply_cannon 111 7.7 0.057 0.152 51.821 55.918 density_matrix_trs4 2 5.0 0.002 0.004 52.589 52.707 multiply_cannon_loop 111 8.7 0.115 0.126 46.831 49.254 ls_scf_init_scf 1 4.0 0.000 0.001 32.644 32.645 ls_scf_init_matrix_S 1 5.0 0.000 0.001 31.385 31.454 mp_waitall_1 6369 11.0 22.844 30.778 22.844 30.778 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 28.896 28.911 multiply_cannon_multrec 1332 9.7 14.276 17.223 22.194 24.937 make_m2s 222 7.7 0.008 0.009 21.236 22.618 make_images 222 8.7 3.150 3.624 21.185 22.569 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.385 17.161 make_images_data 222 9.7 0.004 0.005 11.829 13.429 hybrid_alltoall_any 227 10.6 0.803 3.772 11.259 12.930 dbcsr_mm_accdrv_process 3641 10.4 0.222 0.404 7.554 9.075 dbcsr_mm_accdrv_process_sort 3641 11.4 7.179 8.660 7.179 8.660 mp_sum_l 807 5.4 3.750 6.681 3.750 6.681 multiply_cannon_sync_h2d 1332 9.7 5.550 6.101 5.550 6.101 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.089 6.091 mp_irecv_dv 3229 10.9 2.066 6.027 2.066 6.027 arnoldi_extremal 4 6.8 0.000 0.000 5.162 5.176 arnoldi_normal_ev 4 7.8 0.001 0.004 5.162 5.176 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.830 5.127 build_subspace 16 8.4 0.014 0.021 4.820 4.827 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.471 4.808 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.535 4.738 calculate_norms 2376 9.8 4.202 4.579 4.202 4.579 mp_allgather_i34 111 8.7 2.101 4.286 2.101 4.286 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.570 3.862 dbcsr_matrix_vector_mult_local 304 10.0 3.184 3.684 3.186 3.686 dbcsr_sort_data 658 11.4 3.101 3.461 3.101 3.461 ls_scf_post 1 4.0 0.000 0.001 3.257 3.263 dbcsr_special_finalize 555 9.7 0.006 0.007 2.843 3.239 dbcsr_merge_single_wm 555 10.7 0.537 0.663 2.835 3.231 ls_scf_store_result 1 5.0 0.000 0.000 2.991 3.059 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.983 3.030 dbcsr_data_release 10477 10.7 1.602 2.416 1.602 2.416 dbcsr_finalize 304 7.8 0.050 0.061 1.821 1.984 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.955 1.957 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.118000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2712.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/04dd581a73886059b5d5fb12035b1998b2b62f5a_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 4.666024E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265558. MP_Allreduce 3049 15663. MP_Sync 4 MP_Alltoall 47 46208988. MP_SendRecv 45 115200. MP_ISendRecv 45 115200. MP_Wait 528 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.042 0.056 99.234 99.234 qs_energies 1 2.0 0.000 0.000 98.432 98.437 ls_scf 1 3.0 0.000 0.000 96.504 96.513 dbcsr_multiply_generic 111 6.7 0.018 0.019 78.124 78.327 ls_scf_main 1 4.0 0.000 0.000 62.224 62.225 multiply_cannon 111 7.7 0.098 0.233 55.439 60.753 density_matrix_trs4 2 5.0 0.002 0.003 55.080 55.181 multiply_cannon_loop 111 8.7 0.070 0.081 51.029 52.375 mp_waitall_1 5436 11.0 26.680 32.385 26.680 32.385 ls_scf_init_scf 1 4.0 0.000 0.000 30.630 30.635 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.362 29.410 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.217 27.225 multiply_cannon_multrec 444 9.7 14.065 16.663 21.116 22.792 make_m2s 222 7.7 0.006 0.007 17.863 20.246 make_images 222 8.7 3.731 4.435 17.799 20.184 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.044 16.330 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.408 14.508 make_images_data 222 9.7 0.003 0.004 10.032 12.367 hybrid_alltoall_any 227 10.6 0.790 3.759 9.862 12.264 dbcsr_mm_accdrv_process 3003 10.4 0.177 0.341 6.759 7.904 multiply_cannon_sync_h2d 444 9.7 6.567 7.570 6.567 7.570 dbcsr_mm_accdrv_process_sort 3003 11.4 6.444 7.558 6.444 7.558 mp_allgather_i34 111 8.7 2.622 7.022 2.622 7.022 arnoldi_extremal 4 6.8 0.000 0.000 5.708 5.717 arnoldi_normal_ev 4 7.8 0.001 0.004 5.708 5.717 build_subspace 16 8.4 0.015 0.020 5.317 5.329 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.614 4.775 mp_sum_l 807 5.4 2.863 4.596 2.863 4.596 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.706 4.491 mp_irecv_dv 1241 11.2 1.688 4.453 1.688 4.453 dbcsr_matrix_vector_mult 304 9.0 0.011 0.022 4.103 4.295 dbcsr_matrix_vector_mult_local 304 10.0 3.629 4.104 3.631 4.106 ls_scf_post 1 4.0 0.000 0.000 3.650 3.655 calculate_norms 792 9.8 3.547 3.634 3.547 3.634 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.537 3.614 make_images_sizes 222 9.7 0.000 0.000 1.161 3.497 mp_alltoall_i44 222 10.7 1.161 3.497 1.161 3.497 ls_scf_store_result 1 5.0 0.000 0.000 3.421 3.469 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.935 3.347 dbcsr_finalize 304 7.8 0.062 0.078 2.205 2.299 dbcsr_merge_all 275 8.9 0.475 0.530 2.051 2.136 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.065 2.066 rebuild_ks_matrix 3 7.3 0.000 0.000 2.033 2.034 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.033 2.034 dbcsr_data_release 10123 10.8 1.338 2.000 1.338 2.000 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=99.234000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3681.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/04dd581a73886059b5d5fb12035b1998b2b62f5a_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 8.770118E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284111. MP_Allreduce 3043 21950. MP_Sync 4 MP_Alltoall 47 88727262. MP_SendRecv 42 732600. MP_ISendRecv 42 732600. MP_Wait 267 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.110 0.115 108.088 108.088 qs_energies 1 2.0 0.000 0.000 106.649 106.663 ls_scf 1 3.0 0.000 0.000 103.738 103.751 dbcsr_multiply_generic 111 6.7 0.023 0.028 77.658 77.793 ls_scf_main 1 4.0 0.000 0.000 65.841 65.842 density_matrix_trs4 2 5.0 0.002 0.003 56.668 56.717 multiply_cannon 111 7.7 0.184 0.341 50.087 52.138 multiply_cannon_loop 111 8.7 0.068 0.071 46.547 47.520 ls_scf_init_scf 1 4.0 0.000 0.000 34.169 34.170 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.627 32.646 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.808 29.818 mp_waitall_1 4527 11.1 22.279 26.137 22.279 26.137 make_m2s 222 7.7 0.007 0.007 23.904 24.924 make_images 222 8.7 4.602 5.021 23.794 24.813 multiply_cannon_multrec 444 9.7 17.902 18.630 22.519 23.101 hybrid_alltoall_any 227 10.6 1.664 3.638 12.905 15.603 make_images_data 222 9.7 0.004 0.004 13.128 15.577 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.579 10.957 multiply_cannon_sync_h2d 444 9.7 8.846 8.886 8.846 8.886 arnoldi_extremal 4 6.8 0.000 0.000 7.107 7.123 arnoldi_normal_ev 4 7.8 0.002 0.009 7.107 7.123 build_subspace 16 8.4 0.026 0.036 6.553 6.566 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.250 5.382 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.030 5.283 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.117 5.204 dbcsr_matrix_vector_mult_local 304 10.0 4.853 5.120 4.856 5.122 dbcsr_mm_accdrv_process 1814 10.4 0.209 0.323 4.447 4.573 dbcsr_mm_accdrv_process_sort 1814 11.4 4.147 4.278 4.147 4.278 ls_scf_post 1 4.0 0.000 0.000 3.728 3.740 mp_allgather_i34 111 8.7 1.155 3.619 1.155 3.619 make_images_sizes 222 9.7 0.000 0.000 1.416 3.495 mp_alltoall_i44 222 10.7 1.416 3.494 1.416 3.494 ls_scf_store_result 1 5.0 0.000 0.000 3.427 3.442 calculate_norms 792 9.8 3.242 3.276 3.242 3.276 dbcsr_finalize 304 7.8 0.082 0.090 3.101 3.191 dbcsr_merge_all 275 8.9 0.895 0.917 2.886 2.970 dbcsr_complete_redistribute 5 7.6 1.441 1.472 2.771 2.886 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.881 2.881 dbcsr_data_release 12724 10.6 2.331 2.858 2.331 2.858 matrix_ls_to_qs 2 6.0 0.000 0.000 2.404 2.518 dbcsr_sort_data 325 11.1 2.444 2.516 2.444 2.516 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.462 2.463 rebuild_ks_matrix 3 7.3 0.000 0.000 2.395 2.397 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.395 2.397 dbcsr_new_transposed 4 7.5 0.243 0.251 2.284 2.310 dbcsr_frobenius_norm 74 6.6 2.056 2.126 2.190 2.234 dbcsr_add_d 103 6.2 0.000 0.000 2.137 2.208 dbcsr_add_anytype 103 7.2 0.860 0.892 2.136 2.208 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=108.088000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6822.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/04dd581a73886059b5d5fb12035b1998b2b62f5a_performance_tests/27/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 7009386627072 0.0% 0.0% 100.0% flops 9 x 9 x 32 7335108845568 0.0% 0.0% 100.0% flops 9 x 22 x 32 9866241589248 0.0% 0.0% 100.0% flops 22 x 9 x 32 9884108906496 0.0% 0.0% 100.0% flops 22 x 22 x 32 13354440523776 0.0% 0.0% 100.0% flops 32 x 32 x 9 20607185977344 0.0% 0.0% 100.0% flops 32 x 32 x 22 25186560638976 0.0% 0.0% 100.0% flops 9 x 32 x 32 28458319085568 0.0% 0.0% 100.0% flops 22 x 32 x 32 34782389993472 0.0% 0.0% 100.0% flops 9 x 32 x 9 42881542373376 0.0% 0.0% 100.0% flops 22 x 32 x 9 55680402235392 0.0% 0.0% 100.0% flops 9 x 32 x 22 55680402235392 0.0% 0.0% 100.0% flops 22 x 32 x 22 72328573419520 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 383.054662E+12 0.0% 0.0% 100.0% flops max/rank 733.641090E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 26899403712 0.0% 0.0% 100.0% number of processed stacks 118860288 0.0% 0.0% 100.0% average stack size 0.0 0.0 226.3 marketing flops 780.439111E+12 ------------------------------------------------------------------------------- # multiplications 1445 max memory usage/rank 590.848000E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 102097920 MPI messages size (bytes): total size 37.227590E+12 min size 0.000000E+00 max size 4.551360E+06 average size 364.626312E+03 MPI breakdown and total messages size (bytes): size <= 128 731472 0 128 < size <= 8192 11922720 97670922240 8192 < size <= 32768 24718992 614677610496 32768 < size <= 131072 20000256 1970081366016 131072 < size <= 4194304 42515668 24886801223040 4194304 < size <= 16777216 2208812 9656099886720 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4640 76559. MP_Allreduce 13152 2094. MP_Sync 1064 MP_Alltoall 2588 3686719. 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 0.028 0.110 223.061 223.063 qs_mol_dyn_low 1 2.0 0.004 0.017 222.085 222.105 qs_forces 5 3.8 0.005 0.041 221.921 221.927 qs_energies 5 4.8 0.003 0.096 218.937 218.966 scf_env_do_scf 5 5.8 0.000 0.009 204.177 204.179 scf_env_do_scf_inner_loop 105 6.6 0.003 0.020 177.808 177.822 qs_scf_new_mos 105 7.6 0.000 0.001 137.893 138.086 qs_scf_loop_do_ot 105 8.6 0.001 0.001 137.892 138.085 dbcsr_multiply_generic 1445 12.2 0.122 0.130 130.670 131.130 ot_scf_mini 105 9.6 0.003 0.005 128.207 128.358 multiply_cannon 1445 13.2 0.275 0.289 112.579 114.874 multiply_cannon_loop 1445 14.2 2.381 2.510 110.940 112.287 velocity_verlet 4 3.0 0.001 0.025 105.095 105.095 ot_mini 105 10.6 0.001 0.003 58.963 59.087 multiply_cannon_multrec 69360 15.2 31.990 36.874 39.895 45.330 mp_waitall_1 488190 16.1 32.742 41.268 32.742 41.268 qs_ot_get_p 112 10.4 0.001 0.001 39.902 40.270 qs_ot_get_derivative 55 11.6 0.001 0.002 37.370 37.537 multiply_cannon_sync_h2d 69360 15.2 30.814 35.214 30.814 35.214 multiply_cannon_metrocomm3 69360 15.2 0.194 0.208 24.274 33.071 qs_ot_p2m_diag 40 11.0 0.020 0.031 29.228 29.305 rebuild_ks_matrix 110 8.4 0.000 0.000 28.747 28.948 qs_ks_build_kohn_sham_matrix 110 9.4 0.015 0.031 28.747 28.947 qs_ks_update_qs_env 112 7.6 0.001 0.001 26.356 26.531 init_scf_loop 7 6.6 0.000 0.012 26.338 26.340 cp_dbcsr_syevd 40 12.0 0.002 0.002 26.264 26.265 apply_preconditioner_dbcsr 62 12.6 0.000 0.005 22.926 23.303 apply_single 62 13.6 0.000 0.000 22.926 23.302 prepare_preconditioner 7 7.6 0.000 0.001 21.472 21.518 make_preconditioner 7 8.6 0.000 0.004 21.472 21.518 cp_fm_syevd 40 13.0 0.000 0.001 21.081 21.222 ot_new_cg_direction 55 11.6 0.001 0.004 20.946 20.946 qs_rho_update_rho_low 110 7.6 0.001 0.002 16.962 17.284 calculate_rho_elec 110 8.6 0.030 0.033 16.962 17.283 cp_fm_redistribute_end 40 14.0 8.237 16.436 8.242 16.437 cp_fm_syevd_base 40 14.0 8.188 16.389 8.188 16.389 make_full_inverse_cholesky 7 9.6 0.000 0.000 14.371 14.459 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.210 14.426 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 13.629 13.725 rs_pw_transfer 690 11.5 0.012 0.014 11.232 12.483 density_rs2pw 110 9.6 0.006 0.007 11.085 12.364 mp_sum_l 4684 12.4 11.560 12.258 11.560 12.258 pw_transfer 1645 12.4 0.084 0.102 11.750 11.933 init_scf_run 5 5.8 0.000 0.001 11.830 11.831 scf_env_initial_rho_setup 5 6.8 0.001 0.003 11.830 11.831 fft_wrap_pw1pw2 1425 13.5 0.013 0.016 11.606 11.793 calculate_dm_sparse 110 9.5 0.000 0.001 11.346 11.541 qs_vxc_create 110 10.4 0.003 0.005 10.220 10.255 cp_fm_cholesky_invert 7 10.6 10.171 10.179 10.171 10.179 fft_wrap_pw1pw2_240 915 15.0 1.208 1.286 9.853 10.059 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 9.945 10.005 check_diag 80 13.5 8.588 8.859 9.489 9.631 sum_up_and_integrate 60 10.3 0.028 0.039 8.575 8.586 integrate_v_rspace 60 11.3 0.002 0.003 8.546 8.558 dbcsr_mm_accdrv_process 154766 15.8 4.077 4.276 7.772 8.553 fft3d_pb 915 16.0 2.390 2.573 7.875 8.047 xc_rho_set_and_dset_create 110 12.4 0.077 0.097 7.205 7.446 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.002 7.321 7.338 multiply_cannon_metrocomm1 69360 15.2 0.090 0.098 4.255 7.228 acc_transpose_blocks 69360 15.2 0.360 0.386 6.545 7.047 calculate_first_density_matrix 1 7.0 0.001 0.005 6.943 6.956 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 6.876 6.943 make_full_single_inverse 7 9.6 0.001 0.003 6.817 6.863 xc_vxc_pw_create 60 11.3 0.039 0.051 6.826 6.859 make_m2s 2890 13.2 0.080 0.089 6.067 6.628 make_images 2890 14.2 0.245 0.263 5.963 6.523 xc_pw_derive 510 13.4 0.005 0.007 6.018 6.091 mp_waitany 7680 13.5 4.409 5.550 4.409 5.550 mp_alltoall_z22v 2340 17.7 4.909 5.209 4.909 5.209 acc_transpose_blocks_kernels 69360 16.2 0.852 0.902 4.668 5.092 potential_pw2rs 60 12.3 0.003 0.003 4.777 4.798 wfi_extrapolate 5 7.8 0.000 0.002 4.583 4.583 grid_collocate_task_list 110 9.6 4.085 4.571 4.085 4.571 multiply_cannon_metrocomm4 67915 15.2 0.182 0.201 2.044 4.551 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=223.063000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=559.600000, yerr=4.029888 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 04dd581a73886059b5d5fb12035b1998b2b62f5a Summary: empty Status: OK