=== 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: 72113d53c0cb11844f79f3c088677e10c62b6b2a ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, Cray-FFTW 3.3.8.10, # COSMA 2.6.6, ELPA 2022.11.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.2.2, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.2, # SIRIUS 7.4.3, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Last update: 21.06.2023 # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.2 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/72113d53c0cb11844f79f3c088677e10c62b6b2a_performance_tests/01 job id: 48094756 --- 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/72113d53c0cb11844f79f3c088677e10c62b6b2a_performance_tests/02 job id: 48094758 --- 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/72113d53c0cb11844f79f3c088677e10c62b6b2a_performance_tests/03 job id: 48094761 --- 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/72113d53c0cb11844f79f3c088677e10c62b6b2a_performance_tests/04 job id: 48094762 --- 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/72113d53c0cb11844f79f3c088677e10c62b6b2a_performance_tests/05 job id: 48094763 --- 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/72113d53c0cb11844f79f3c088677e10c62b6b2a_performance_tests/06 job id: 48094764 --- 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/72113d53c0cb11844f79f3c088677e10c62b6b2a_performance_tests/07 job id: 48094765 --- 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/72113d53c0cb11844f79f3c088677e10c62b6b2a_performance_tests/08 job id: 48094766 --- 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/72113d53c0cb11844f79f3c088677e10c62b6b2a_performance_tests/09 job id: 48094767 --- 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/72113d53c0cb11844f79f3c088677e10c62b6b2a_performance_tests/10 job id: 48094768 --- 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/72113d53c0cb11844f79f3c088677e10c62b6b2a_performance_tests/11 job id: 48094769 --- 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/72113d53c0cb11844f79f3c088677e10c62b6b2a_performance_tests/12 job id: 48094770 --- 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/72113d53c0cb11844f79f3c088677e10c62b6b2a_performance_tests/13 job id: 48094771 --- 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/72113d53c0cb11844f79f3c088677e10c62b6b2a_performance_tests/14 job id: 48094772 --- 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/72113d53c0cb11844f79f3c088677e10c62b6b2a_performance_tests/15 job id: 48094773 --- 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/72113d53c0cb11844f79f3c088677e10c62b6b2a_performance_tests/16 job id: 48094774 --- 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/72113d53c0cb11844f79f3c088677e10c62b6b2a_performance_tests/17 job id: 48094775 --- 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/72113d53c0cb11844f79f3c088677e10c62b6b2a_performance_tests/18 job id: 48094776 --- 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/72113d53c0cb11844f79f3c088677e10c62b6b2a_performance_tests/19 job id: 48094777 --- 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/72113d53c0cb11844f79f3c088677e10c62b6b2a_performance_tests/20 job id: 48094778 --- 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/72113d53c0cb11844f79f3c088677e10c62b6b2a_performance_tests/21 job id: 48094781 --- 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/72113d53c0cb11844f79f3c088677e10c62b6b2a_performance_tests/22 job id: 48094784 --- 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/72113d53c0cb11844f79f3c088677e10c62b6b2a_performance_tests/23 job id: 48094785 --- 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/72113d53c0cb11844f79f3c088677e10c62b6b2a_performance_tests/24 job id: 48094786 --- 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/72113d53c0cb11844f79f3c088677e10c62b6b2a_performance_tests/25 job id: 48094787 --- 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/72113d53c0cb11844f79f3c088677e10c62b6b2a_performance_tests/26 job id: 48094788 --- 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/72113d53c0cb11844f79f3c088677e10c62b6b2a_performance_tests/27 job id: 48094789 --- 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/72113d53c0cb11844f79f3c088677e10c62b6b2a_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 424 8. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.026 0.035 133.868 133.869 farming_run 1 2.0 133.443 133.446 133.836 133.841 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.489433E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 230 1134128. MP_Allreduce 571 1938539. MP_Sync 25 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 10 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.006 0.024 114.867 114.868 qs_energies 1 2.0 0.000 0.000 114.646 114.685 mp2_main 1 3.0 0.000 0.000 112.575 112.614 mp2_gpw_main 1 4.0 0.037 0.043 111.594 111.633 mp2_ri_gpw_compute_in 1 5.0 0.174 0.204 92.715 93.053 mp2_ri_gpw_compute_in_loop 1 6.0 0.003 0.004 54.846 55.185 mp2_eri_3c_integrate_gpw 272 7.0 0.152 0.167 41.237 46.313 get_2c_integrals 1 6.0 0.008 0.009 36.926 37.665 integrate_v_rspace 273 8.0 0.436 0.453 24.669 29.627 pw_transfer 6555 10.6 0.378 0.393 27.230 27.682 fft_wrap_pw1pw2 5465 11.4 0.047 0.049 25.874 26.391 grid_integrate_task_list 273 9.0 20.495 25.928 20.495 25.928 fft_wrap_pw1pw2_100 2178 12.4 1.145 1.198 23.394 23.927 compute_2c_integrals 1 7.0 0.002 0.003 19.552 19.552 compute_2c_integrals_loop_lm 1 8.0 0.004 0.004 19.087 19.278 mp2_eri_2c_integrate_gpw 1 9.0 2.378 2.416 19.084 19.274 rpa_ri_compute_en 1 5.0 0.020 0.021 18.753 18.903 cp_fm_cholesky_decompose 12 8.2 17.297 18.023 17.297 18.023 cholesky_decomp 1 7.0 0.000 0.000 16.214 16.936 fft3d_s 5443 13.4 16.042 16.515 16.063 16.535 ao_to_mo_and_store_B_mult_1 272 7.0 10.767 15.320 10.767 15.320 calculate_wavefunction 272 8.0 5.387 5.531 12.471 13.145 rpa_num_int 1 6.0 0.003 0.044 10.600 10.640 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.585 10.591 calc_mat_Q 8 8.0 0.000 0.000 9.467 9.535 contract_S_to_Q 8 9.0 0.000 0.000 8.887 8.956 calc_potential_gpw 544 9.5 0.005 0.005 8.249 8.599 parallel_gemm_fm 14 9.1 0.000 0.000 8.465 8.561 parallel_gemm_fm_cosma 14 10.1 8.465 8.561 8.465 8.561 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.139 8.337 potential_pw2rs 545 10.0 0.107 0.110 7.622 8.226 create_integ_mat 1 6.0 0.006 0.008 7.722 7.722 collocate_single_gaussian 272 10.0 0.040 0.042 7.403 7.588 array2fm 1 7.0 0.000 0.000 6.679 7.154 pw_scatter_s 2720 13.7 4.427 4.597 4.427 4.597 pw_gather_s 2722 13.2 3.854 4.238 3.854 4.238 array2fm_buffer_send 1 8.0 2.949 3.180 2.949 3.180 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.633022, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2796.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/72113d53c0cb11844f79f3c088677e10c62b6b2a_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 205321. MP_Allreduce 424 9. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.028 0.041 423.577 423.578 farming_run 1 2.0 422.588 422.600 423.533 423.536 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827141120 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788822 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.236681E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 757 478553. MP_Allreduce 2021 21391. MP_Sync 37 MP_Alltoall 77 27528027. MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.031 217.906 217.906 qs_energies 1 2.0 0.002 0.007 217.644 217.661 scf_env_do_scf 1 3.0 0.000 0.000 114.782 114.782 qs_ks_update_qs_env 5 5.0 0.000 0.000 113.868 113.876 rebuild_ks_matrix 4 6.0 0.000 0.000 113.866 113.875 qs_ks_build_kohn_sham_matrix 4 7.0 0.054 0.063 113.866 113.875 hfx_ks_matrix 4 8.0 0.001 0.001 113.486 113.491 integrate_four_center 4 9.0 0.152 0.460 113.485 113.490 integrate_four_center_main 4 10.0 0.128 0.598 101.779 104.824 integrate_four_center_bin 266 11.0 101.652 104.822 101.652 104.822 mp2_main 1 3.0 0.013 0.061 102.570 102.587 mp2_gpw_main 1 4.0 0.057 0.098 101.675 101.695 init_scf_loop 1 4.0 0.000 0.000 96.460 96.462 mp2_ri_gpw_compute_in 1 5.0 0.074 0.116 74.632 75.716 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.003 54.073 55.153 mp2_eri_3c_integrate_gpw 91 7.0 0.145 0.164 41.817 46.574 integrate_v_rspace 95 8.0 0.395 0.562 28.199 32.865 pw_transfer 2240 10.6 0.145 0.163 29.891 30.351 fft_wrap_pw1pw2 1868 11.4 0.018 0.021 28.907 29.370 mp2_ri_gpw_compute_en 1 5.0 0.057 0.077 26.740 28.914 grid_integrate_task_list 95 9.0 23.501 28.242 23.501 28.242 ao_to_mo_and_store_B_mult_1 91 7.0 10.574 28.220 10.574 28.220 fft_wrap_pw1pw2_100 730 12.4 1.274 1.447 26.618 27.013 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.842 1.886 24.999 25.011 get_2c_integrals 1 6.0 0.000 0.000 20.462 20.497 compute_2c_integrals 1 7.0 0.003 0.006 19.443 19.458 compute_2c_integrals_loop_lm 1 8.0 0.004 0.012 18.843 19.319 mp2_eri_2c_integrate_gpw 1 9.0 1.754 1.902 18.839 19.318 fft3d_s 1823 13.4 18.426 18.787 18.439 18.801 scf_env_do_scf_inner_loop 4 4.0 0.000 0.001 18.320 18.321 calculate_wavefunction 91 8.0 2.004 2.034 9.730 9.962 mp2_ri_gpw_compute_en_expansio 172 7.0 0.558 0.579 8.830 9.297 potential_pw2rs 186 10.0 0.033 0.035 8.603 9.103 local_gemm 172 8.0 8.272 8.747 8.272 8.747 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.220 8.499 mp2_ri_gpw_compute_en_comm 22 7.0 0.501 0.517 7.941 8.386 calc_potential_gpw 182 9.5 0.002 0.002 7.918 8.175 collocate_single_gaussian 91 10.0 0.021 0.048 7.877 8.128 mp_sync 37 10.5 3.666 7.143 3.666 7.143 integrate_four_center_load 4 10.0 0.000 0.000 6.777 6.782 hfx_load_balance 1 11.0 0.000 0.000 6.777 6.782 mp_sendrecv_dm3 2068 8.0 5.969 6.414 5.969 6.414 mp2_ri_gpw_compute_en_ener 172 7.0 6.342 6.386 6.342 6.386 pw_gather_s 912 13.2 4.883 5.330 4.883 5.330 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.677214, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1496.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/72113d53c0cb11844f79f3c088677e10c62b6b2a_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 451.981312E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62379. MP_Allreduce 10329 270. MP_Sync 530 MP_Alltoall 2083 638459. 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 0.019 0.046 53.818 53.819 qs_mol_dyn_low 1 2.0 0.004 0.007 53.506 53.516 qs_forces 11 3.9 0.003 0.009 53.286 53.288 qs_energies 11 4.9 0.002 0.006 51.781 51.814 scf_env_do_scf 11 5.9 0.000 0.001 45.306 45.307 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 43.261 43.261 dbcsr_multiply_generic 2286 12.5 0.093 0.098 33.389 33.869 qs_scf_new_mos 108 7.5 0.000 0.001 33.213 33.489 qs_scf_loop_do_ot 108 8.5 0.000 0.001 33.212 33.488 ot_scf_mini 108 9.5 0.002 0.002 31.504 31.697 multiply_cannon 2286 13.5 0.188 0.199 26.617 27.829 multiply_cannon_loop 2286 14.5 1.831 1.936 25.975 27.219 velocity_verlet 10 3.0 0.001 0.002 25.977 25.980 ot_mini 108 10.5 0.001 0.001 18.937 19.276 qs_ot_get_derivative 108 11.5 0.001 0.001 15.986 16.162 mp_waitall_1 245248 16.5 8.155 14.637 8.155 14.637 multiply_cannon_metrocomm3 54864 15.5 0.072 0.080 5.966 13.344 multiply_cannon_multrec 54864 15.5 3.695 5.883 8.060 11.714 qs_ot_get_p 119 10.4 0.001 0.001 7.863 8.155 rebuild_ks_matrix 119 8.3 0.000 0.000 7.844 7.983 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.032 7.843 7.983 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.904 7.038 mp_sum_l 7287 12.8 4.892 6.479 4.892 6.479 multiply_cannon_sync_h2d 54864 15.5 5.242 6.377 5.242 6.377 dbcsr_mm_accdrv_process 76910 16.1 1.832 2.874 4.278 6.174 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.547 6.002 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.165 5.202 init_scf_run 11 5.9 0.000 0.001 5.157 5.157 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.157 5.157 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 4.893 5.014 sum_up_and_integrate 119 10.3 0.012 0.015 4.439 4.446 integrate_v_rspace 119 11.3 0.002 0.003 4.427 4.436 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.393 4.393 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.153 4.336 calculate_rho_elec 119 8.7 0.012 0.017 4.153 4.335 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.207 4.207 cp_fm_redistribute_end 50 14.0 2.148 4.183 2.154 4.186 cp_fm_diag_elpa_base 50 14.0 2.027 4.078 2.031 4.085 jit_kernel_multiply 13 15.8 2.382 3.272 2.382 3.272 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.891 3.093 apply_single 119 13.6 0.000 0.000 2.891 3.093 calculate_dm_sparse 119 9.5 0.000 0.001 2.980 3.089 calculate_first_density_matrix 1 7.0 0.000 0.003 2.997 3.004 acc_transpose_blocks 54864 15.5 0.231 0.260 2.345 2.933 ot_diis_step 108 11.5 0.021 0.132 2.714 2.823 rs_pw_transfer 974 11.9 0.011 0.013 2.643 2.733 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.006 2.500 2.502 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.388 2.440 multiply_cannon_metrocomm1 54864 15.5 0.057 0.062 1.597 2.403 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.237 2.290 density_rs2pw 119 9.7 0.004 0.004 2.116 2.238 grid_integrate_task_list 119 12.3 2.028 2.124 2.028 2.124 wfi_extrapolate 11 7.9 0.001 0.004 2.087 2.087 init_scf_loop 11 6.9 0.002 0.016 2.029 2.029 mp_sum_d 4135 12.0 1.320 1.947 1.320 1.947 potential_pw2rs 119 12.3 0.004 0.004 1.786 1.793 pw_transfer 1439 11.6 0.052 0.056 1.668 1.735 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.592 1.660 make_m2s 4572 13.5 0.054 0.055 1.573 1.613 make_images 4572 14.5 0.134 0.140 1.491 1.530 mp_alltoall_d11v 2130 13.8 1.231 1.494 1.231 1.494 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.001 1.439 1.479 mp_waitany 12084 13.8 1.247 1.438 1.247 1.438 acc_transpose_blocks_sync 164592 16.5 1.200 1.434 1.200 1.434 grid_collocate_task_list 119 9.7 1.355 1.420 1.355 1.420 fft3d_ps 1201 14.6 0.372 0.480 1.234 1.298 fft_wrap_pw1pw2_140 487 13.2 0.185 0.202 1.228 1.297 acc_transpose_blocks_kernels 54864 16.5 0.247 0.386 0.866 1.196 dbcsr_dot_sd 1205 11.9 0.049 0.061 0.724 1.156 parallel_gemm_fm 81 9.0 0.000 0.000 1.136 1.141 parallel_gemm_fm_cosma 81 10.0 1.136 1.141 1.136 1.141 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=53.819000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=431.272727, yerr=1.052349 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/72113d53c0cb11844f79f3c088677e10c62b6b2a_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 487.874560E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2194560 MPI messages size (bytes): total size 310.646604E+09 min size 0.000000E+00 max size 1.145520E+06 average size 141.553031E+03 MPI breakdown and total messages size (bytes): size <= 128 724648 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 281952 4619501568 32768 < size <= 131072 494448 39143342080 131072 < size <= 4194304 440000 264807943488 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62658. MP_Allreduce 10306 303. MP_Sync 54 MP_Alltoall 2060 1058300. MP_SendRecv 16779 37093. MP_ISendRecv 16779 37093. MP_Wait 23539 MP_comm_split 50 MP_ISend 5720 128509. MP_IRecv 5720 128509. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.032 40.424 40.426 qs_mol_dyn_low 1 2.0 0.003 0.004 40.200 40.209 qs_forces 11 3.9 0.002 0.002 40.131 40.133 qs_energies 11 4.9 0.005 0.007 38.423 38.428 scf_env_do_scf 11 5.9 0.001 0.001 32.586 32.586 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 30.109 30.110 dbcsr_multiply_generic 2286 12.5 0.102 0.106 22.856 23.222 qs_scf_new_mos 108 7.5 0.001 0.001 21.030 21.257 qs_scf_loop_do_ot 108 8.5 0.001 0.001 21.029 21.256 ot_scf_mini 108 9.5 0.003 0.003 20.119 20.288 multiply_cannon 2286 13.5 0.209 0.218 17.481 19.336 velocity_verlet 10 3.0 0.002 0.006 18.540 18.541 multiply_cannon_loop 2286 14.5 1.189 1.252 16.338 17.792 ot_mini 108 10.5 0.001 0.001 12.518 12.750 mp_waitall_1 200699 16.5 5.786 11.062 5.786 11.062 qs_ot_get_derivative 108 11.5 0.001 0.001 10.155 10.327 multiply_cannon_metrocomm3 27432 15.5 0.072 0.074 4.318 9.833 multiply_cannon_multrec 27432 15.5 1.822 4.076 6.353 9.149 rebuild_ks_matrix 119 8.3 0.000 0.000 7.196 7.334 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 7.195 7.333 dbcsr_mm_accdrv_process 47894 16.0 3.602 6.120 4.451 6.665 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.335 6.460 qs_ot_get_p 119 10.4 0.002 0.005 4.715 4.944 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.797 4.905 init_scf_run 11 5.9 0.000 0.001 4.588 4.588 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.587 4.588 mp_sum_l 7287 12.8 2.361 4.305 2.361 4.305 sum_up_and_integrate 119 10.3 0.024 0.028 4.098 4.104 integrate_v_rspace 119 11.3 0.002 0.003 4.073 4.082 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.340 4.045 apply_single 119 13.6 0.000 0.000 3.339 4.045 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.751 3.794 calculate_rho_elec 119 8.7 0.021 0.024 3.750 3.793 calculate_first_density_matrix 1 7.0 0.001 0.002 3.075 3.080 qs_ot_p2m_diag 50 11.0 0.009 0.013 2.973 2.992 rs_pw_transfer 974 11.9 0.010 0.011 2.454 2.943 make_m2s 4572 13.5 0.052 0.054 2.573 2.907 make_images 4572 14.5 0.205 0.245 2.483 2.820 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.577 2.577 density_rs2pw 119 9.7 0.004 0.004 2.010 2.536 init_scf_loop 11 6.9 0.001 0.003 2.457 2.459 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.299 2.388 ot_diis_step 108 11.5 0.011 0.011 2.314 2.315 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.310 2.312 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.271 2.271 cp_fm_redistribute_end 50 14.0 1.151 2.248 1.154 2.250 calculate_dm_sparse 119 9.5 0.000 0.001 2.133 2.216 multiply_cannon_sync_h2d 27432 15.5 1.719 2.202 1.719 2.202 cp_fm_diag_elpa_base 50 14.0 1.064 2.158 1.092 2.194 jit_kernel_multiply 10 16.4 0.790 2.173 0.790 2.173 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.114 2.153 acc_transpose_blocks 27432 15.5 0.113 0.119 1.707 2.110 grid_integrate_task_list 119 12.3 1.844 1.935 1.844 1.935 pw_transfer 1439 11.6 0.065 0.070 1.900 1.932 fft_wrap_pw1pw2 1201 12.6 0.008 0.009 1.808 1.842 potential_pw2rs 119 12.3 0.006 0.006 1.822 1.831 make_images_data 4572 15.5 0.048 0.054 1.226 1.566 prepare_preconditioner 11 7.9 0.000 0.000 1.516 1.541 make_preconditioner 11 8.9 0.000 0.001 1.516 1.541 fft_wrap_pw1pw2_140 487 13.2 0.203 0.216 1.449 1.483 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.417 1.474 wfi_extrapolate 11 7.9 0.001 0.001 1.447 1.447 hybrid_alltoall_any 4725 16.4 0.053 0.115 1.086 1.435 grid_collocate_task_list 119 9.7 1.280 1.432 1.280 1.432 fft3d_ps 1201 14.6 0.522 0.577 1.358 1.384 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.331 1.343 mp_alltoall_d11v 2130 13.8 1.157 1.301 1.157 1.301 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.190 1.236 mp_waitany 5720 13.7 0.536 1.069 0.536 1.069 rs_pw_transfer_RS2PW_140 130 11.5 0.138 0.147 0.564 1.050 acc_transpose_blocks_kernels 27432 16.5 0.188 0.276 0.752 1.035 mp_sum_d 4135 12.0 0.555 1.033 0.555 1.033 qs_energies_init_hamiltonians 11 5.9 0.002 0.008 0.980 0.986 mp_allgather_i34 2286 14.5 0.565 0.966 0.565 0.966 acc_transpose_blocks_sync 82296 16.5 0.815 0.940 0.815 0.940 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.918 0.931 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=40.426000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=464.545455, yerr=1.616035 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/72113d53c0cb11844f79f3c088677e10c62b6b2a_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 521.678848E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63490. MP_Allreduce 10155 305. MP_Sync 54 MP_Alltoall 1821 1607886. 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 0.013 0.030 33.321 33.321 qs_mol_dyn_low 1 2.0 0.003 0.004 33.080 33.087 qs_forces 11 3.9 0.002 0.003 33.024 33.025 qs_energies 11 4.9 0.001 0.002 31.437 31.439 scf_env_do_scf 11 5.9 0.000 0.001 26.051 26.051 scf_env_do_scf_inner_loop 108 6.5 0.007 0.015 23.613 23.613 dbcsr_multiply_generic 2286 12.5 0.096 0.097 17.740 17.840 qs_scf_new_mos 108 7.5 0.001 0.001 15.515 15.529 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.514 15.529 velocity_verlet 10 3.0 0.001 0.001 15.267 15.268 multiply_cannon 2286 13.5 0.194 0.198 14.351 15.074 ot_scf_mini 108 9.5 0.003 0.004 14.759 14.773 multiply_cannon_loop 2286 14.5 0.857 0.895 13.587 14.339 ot_mini 108 10.5 0.025 0.194 9.277 9.296 multiply_cannon_multrec 18288 15.5 1.960 2.974 7.633 7.849 qs_ot_get_derivative 108 11.5 0.001 0.001 7.607 7.622 dbcsr_mm_accdrv_process 38222 16.0 5.492 6.458 5.580 6.522 rebuild_ks_matrix 119 8.3 0.000 0.001 6.353 6.368 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.021 6.352 6.368 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.603 5.616 init_scf_run 11 5.9 0.000 0.001 4.230 4.230 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.230 4.230 sum_up_and_integrate 119 10.3 0.031 0.032 3.894 3.899 integrate_v_rspace 119 11.3 0.002 0.003 3.863 3.872 mp_waitall_1 158411 16.6 2.405 3.486 2.405 3.486 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.815 3.424 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.407 3.411 calculate_rho_elec 119 8.7 0.031 0.032 3.406 3.411 qs_ot_get_p 119 10.4 0.004 0.008 3.265 3.289 calculate_first_density_matrix 1 7.0 0.000 0.000 3.010 3.011 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.186 2.523 apply_single 119 13.6 0.000 0.000 2.186 2.523 init_scf_loop 11 6.9 0.000 0.001 2.423 2.423 multiply_cannon_metrocomm3 18288 15.5 0.047 0.049 1.470 2.417 rs_pw_transfer 974 11.9 0.009 0.010 2.158 2.394 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.173 2.180 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.178 2.180 density_rs2pw 119 9.7 0.004 0.004 1.899 2.149 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.028 2.034 calculate_dm_sparse 119 9.5 0.000 0.001 2.008 2.013 make_m2s 4572 13.5 0.044 0.045 1.769 1.919 grid_integrate_task_list 119 12.3 1.800 1.898 1.800 1.898 acc_transpose_blocks 18288 15.5 0.078 0.080 1.816 1.896 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.879 1.879 pw_transfer 1439 11.6 0.065 0.069 1.862 1.868 make_images 4572 14.5 0.192 0.205 1.683 1.832 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.770 1.780 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.693 1.702 potential_pw2rs 119 12.3 0.007 0.008 1.654 1.660 ot_diis_step 108 11.5 0.011 0.014 1.629 1.653 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.642 1.651 cp_fm_diag_elpa_base 50 14.0 1.619 1.632 1.641 1.649 mp_sum_l 7287 12.8 1.231 1.622 1.231 1.622 prepare_preconditioner 11 7.9 0.000 0.000 1.603 1.605 make_preconditioner 11 8.9 0.000 0.000 1.603 1.605 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.462 1.547 fft_wrap_pw1pw2_140 487 13.2 0.253 0.259 1.439 1.450 grid_collocate_task_list 119 9.7 1.246 1.408 1.246 1.408 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.348 1.351 fft3d_ps 1201 14.6 0.529 0.545 1.259 1.270 multiply_cannon_sync_h2d 18288 15.5 1.043 1.172 1.043 1.172 wfi_extrapolate 11 7.9 0.001 0.001 1.168 1.168 acc_transpose_blocks_kernels 18288 16.5 0.217 0.225 0.979 1.020 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.941 0.958 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.937 0.938 jit_kernel_multiply 4 15.0 0.029 0.937 0.029 0.937 make_images_data 4572 15.5 0.047 0.052 0.749 0.916 hybrid_alltoall_any 4725 16.4 0.058 0.116 0.646 0.831 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.820 0.822 acc_transpose_blocks_sync 54864 16.5 0.739 0.814 0.739 0.814 jit_kernel_transpose 5 15.6 0.761 0.801 0.761 0.801 mp_alltoall_d11v 2130 13.8 0.629 0.752 0.629 0.752 mp_waitany 9880 13.7 0.497 0.749 0.497 0.749 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.676 0.745 rs_pw_transfer_RS2PW_140 130 11.5 0.118 0.122 0.495 0.736 cp_fm_cholesky_invert 11 10.9 0.677 0.680 0.677 0.680 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=33.321000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.727273, yerr=1.762793 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/72113d53c0cb11844f79f3c088677e10c62b6b2a_performance_tests/06/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 114.044384E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3805952 0.0% 0.0% 100.0% average stack size 0.0 0.0 38.6 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 555.638784E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63489. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.028 37.620 37.622 qs_mol_dyn_low 1 2.0 0.003 0.003 37.453 37.468 qs_forces 11 3.9 0.002 0.002 37.393 37.394 qs_energies 11 4.9 0.001 0.001 35.664 35.671 scf_env_do_scf 11 5.9 0.000 0.001 30.306 30.307 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 26.935 26.935 dbcsr_multiply_generic 2286 12.5 0.100 0.103 20.841 20.983 qs_scf_new_mos 108 7.5 0.001 0.001 18.538 18.592 qs_scf_loop_do_ot 108 8.5 0.001 0.001 18.538 18.591 velocity_verlet 10 3.0 0.001 0.001 18.464 18.466 multiply_cannon 2286 13.5 0.218 0.224 17.141 17.581 ot_scf_mini 108 9.5 0.002 0.003 17.493 17.566 multiply_cannon_loop 2286 14.5 1.503 1.577 16.211 16.590 ot_mini 108 10.5 0.001 0.001 11.131 11.197 multiply_cannon_multrec 27432 15.5 2.871 3.667 9.720 10.080 qs_ot_get_derivative 108 11.5 0.001 0.001 9.277 9.335 dbcsr_mm_accdrv_process 47916 15.9 6.307 8.236 6.748 8.583 rebuild_ks_matrix 119 8.3 0.000 0.000 6.555 6.606 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.555 6.605 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.805 5.849 init_scf_run 11 5.9 0.000 0.001 4.031 4.031 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.030 4.031 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.435 3.923 sum_up_and_integrate 119 10.3 0.035 0.038 3.747 3.756 integrate_v_rspace 119 11.3 0.003 0.003 3.712 3.721 qs_ot_get_p 119 10.4 0.009 0.019 3.566 3.642 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.514 3.554 calculate_rho_elec 119 8.7 0.040 0.046 3.513 3.554 init_scf_loop 11 6.9 0.004 0.010 3.353 3.353 acc_transpose_blocks 27432 15.5 0.117 0.119 2.640 2.861 mp_waitall_1 137007 16.6 1.809 2.659 1.809 2.659 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.181 2.646 apply_single 119 13.6 0.000 0.000 2.181 2.646 calculate_first_density_matrix 1 7.0 0.000 0.000 2.639 2.641 prepare_preconditioner 11 7.9 0.000 0.000 2.479 2.491 make_preconditioner 11 8.9 0.000 0.000 2.479 2.491 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.095 2.423 make_m2s 4572 13.5 0.054 0.055 2.203 2.317 calculate_dm_sparse 119 9.5 0.000 0.000 2.255 2.310 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.278 2.305 make_images 4572 14.5 0.270 0.331 2.095 2.208 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.150 2.161 rs_pw_transfer 974 11.9 0.009 0.009 1.902 2.083 density_rs2pw 119 9.7 0.004 0.004 1.871 2.063 pw_transfer 1439 11.6 0.065 0.070 1.953 1.986 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.959 1.960 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.861 1.898 grid_integrate_task_list 119 12.3 1.810 1.883 1.810 1.883 ot_diis_step 108 11.5 0.012 0.012 1.808 1.808 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.794 1.807 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.751 1.751 multiply_cannon_metrocomm3 27432 15.5 0.041 0.042 0.898 1.660 mp_sum_l 7287 12.8 1.070 1.635 1.070 1.635 fft_wrap_pw1pw2_140 487 13.2 0.287 0.300 1.543 1.582 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.522 1.530 cp_fm_diag_elpa_base 50 14.0 1.488 1.504 1.520 1.529 potential_pw2rs 119 12.3 0.009 0.009 1.500 1.502 acc_transpose_blocks_sync 82296 16.5 1.379 1.471 1.379 1.471 grid_collocate_task_list 119 9.7 1.251 1.371 1.251 1.371 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.335 1.346 wfi_extrapolate 11 7.9 0.001 0.001 1.338 1.338 fft3d_ps 1201 14.6 0.556 0.606 1.300 1.325 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.272 1.286 acc_transpose_blocks_kernels 27432 16.5 0.270 0.277 1.113 1.278 jit_kernel_multiply 6 16.3 0.374 1.227 0.374 1.227 cp_fm_upper_to_full 72 14.2 0.807 1.148 0.807 1.148 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.075 1.076 jit_kernel_transpose 5 15.6 0.843 1.009 0.843 1.009 dbcsr_complete_redistribute 329 12.2 0.121 0.147 0.731 0.998 make_images_data 4572 15.5 0.047 0.051 0.813 0.923 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.800 0.874 mp_alltoall_d11v 2130 13.8 0.734 0.864 0.734 0.864 hybrid_alltoall_any 4725 16.4 0.066 0.154 0.696 0.845 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.839 0.844 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.543 0.804 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=37.622000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=528.363636, yerr=2.267266 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/72113d53c0cb11844f79f3c088677e10c62b6b2a_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 621.182976E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63488. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.033 0.051 30.175 30.179 qs_mol_dyn_low 1 2.0 0.003 0.003 29.691 29.699 qs_forces 11 3.9 0.046 0.054 29.596 29.596 qs_energies 11 4.9 0.013 0.016 27.835 27.861 scf_env_do_scf 11 5.9 0.000 0.001 22.045 22.046 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 19.571 19.572 velocity_verlet 10 3.0 0.002 0.002 14.847 14.852 dbcsr_multiply_generic 2286 12.5 0.108 0.124 13.527 13.610 qs_scf_new_mos 108 7.5 0.001 0.001 11.630 11.651 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.629 11.650 multiply_cannon 2286 13.5 0.222 0.230 10.946 11.453 ot_scf_mini 108 9.5 0.002 0.002 10.929 10.951 multiply_cannon_loop 2286 14.5 0.648 0.676 10.043 10.214 multiply_cannon_multrec 9144 15.5 1.682 1.869 6.110 6.312 ot_mini 108 10.5 0.001 0.001 6.252 6.280 rebuild_ks_matrix 119 8.3 0.000 0.000 5.953 5.976 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.952 5.976 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.294 5.315 qs_ot_get_derivative 108 11.5 0.001 0.001 4.979 5.002 dbcsr_mm_accdrv_process 12550 15.8 3.625 4.309 4.318 4.381 init_scf_run 11 5.9 0.000 0.001 3.987 3.987 scf_env_initial_rho_setup 11 6.9 0.002 0.002 3.987 3.987 sum_up_and_integrate 119 10.3 0.037 0.041 3.632 3.638 integrate_v_rspace 119 11.3 0.003 0.003 3.595 3.601 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.510 3.516 calculate_rho_elec 119 8.7 0.060 0.061 3.509 3.515 qs_ot_get_p 119 10.4 0.014 0.015 2.827 2.859 calculate_first_density_matrix 1 7.0 0.000 0.000 2.836 2.836 init_scf_loop 11 6.9 0.007 0.008 2.455 2.457 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.128 2.129 pw_transfer 1439 11.6 0.066 0.069 2.030 2.041 density_rs2pw 119 9.7 0.004 0.004 1.815 1.953 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.937 1.948 acc_transpose_blocks 9144 15.5 0.042 0.045 1.880 1.945 make_m2s 4572 13.5 0.034 0.036 1.793 1.941 grid_integrate_task_list 119 12.3 1.854 1.921 1.854 1.921 calculate_dm_sparse 119 9.5 0.000 0.000 1.890 1.908 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.841 1.850 make_images 4572 14.5 0.268 0.300 1.705 1.850 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.839 1.841 mp_waitall_1 115863 16.7 1.261 1.824 1.261 1.824 rs_pw_transfer 974 11.9 0.008 0.008 1.639 1.788 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.767 1.786 prepare_preconditioner 11 7.9 0.000 0.000 1.695 1.698 make_preconditioner 11 8.9 0.000 0.000 1.695 1.698 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.588 1.612 fft_wrap_pw1pw2_140 487 13.2 0.364 0.374 1.597 1.607 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.597 1.598 qs_energies_init_hamiltonians 11 5.9 0.010 0.015 1.513 1.538 grid_collocate_task_list 119 9.7 1.296 1.409 1.296 1.409 jit_kernel_multiply 8 15.8 0.655 1.385 0.655 1.385 potential_pw2rs 119 12.3 0.010 0.010 1.359 1.362 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.331 1.341 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.312 1.320 cp_fm_diag_elpa_base 50 14.0 1.285 1.302 1.310 1.319 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.285 1.290 fft3d_ps 1201 14.6 0.558 0.571 1.271 1.283 ot_diis_step 108 11.5 0.012 0.013 1.261 1.261 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.202 1.237 apply_single 119 13.6 0.000 0.000 1.202 1.236 acc_transpose_blocks_kernels 9144 16.5 0.118 0.120 1.092 1.107 wfi_extrapolate 11 7.9 0.001 0.001 1.097 1.097 jit_kernel_transpose 5 15.6 0.974 0.989 0.974 0.989 qs_env_update_s_mstruct 11 6.9 0.000 0.001 0.946 0.989 hybrid_alltoall_any 4725 16.4 0.065 0.176 0.769 0.980 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.873 0.922 make_images_data 4572 15.5 0.042 0.046 0.757 0.914 mp_alltoall_d11v 2130 13.8 0.772 0.854 0.772 0.854 cp_fm_cholesky_invert 11 10.9 0.838 0.841 0.838 0.841 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.815 0.821 qs_create_task_list 11 7.9 0.239 0.249 0.788 0.812 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.780 0.782 acc_transpose_blocks_sync 27432 16.5 0.729 0.777 0.729 0.777 dbcsr_complete_redistribute 329 12.2 0.276 0.304 0.615 0.640 mp_allgather_i34 2286 14.5 0.221 0.637 0.221 0.637 multiply_cannon_metrocomm3 9144 15.5 0.021 0.021 0.312 0.620 mp_alltoall_z22v 1201 16.6 0.587 0.613 0.587 0.613 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=30.179000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=588.818182, yerr=5.041973 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/72113d53c0cb11844f79f3c088677e10c62b6b2a_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 770.973696E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63723. MP_Allreduce 10154 429. MP_Sync 54 MP_Alltoall 1582 7383731. MP_SendRecv 2499 189067. MP_ISendRecv 2499 189067. MP_Wait 6399 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.048 44.529 44.530 qs_mol_dyn_low 1 2.0 0.005 0.021 43.943 43.952 qs_forces 11 3.9 0.016 0.019 43.882 43.884 qs_energies 11 4.9 0.001 0.001 41.830 41.834 scf_env_do_scf 11 5.9 0.001 0.001 35.201 35.201 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 27.313 27.314 velocity_verlet 10 3.0 0.002 0.002 24.324 24.330 dbcsr_multiply_generic 2286 12.5 0.101 0.104 20.037 20.162 qs_scf_new_mos 108 7.5 0.001 0.001 17.518 17.622 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.517 17.622 ot_scf_mini 108 9.5 0.002 0.002 16.357 16.464 multiply_cannon 2286 13.5 0.297 0.301 15.666 16.444 multiply_cannon_loop 2286 14.5 0.873 0.899 14.382 15.215 ot_mini 108 10.5 0.001 0.001 10.050 10.169 multiply_cannon_multrec 9144 15.5 3.439 5.012 9.137 9.254 qs_ot_get_derivative 108 11.5 0.001 0.001 7.980 8.087 init_scf_loop 11 6.9 0.001 0.002 7.861 7.864 rebuild_ks_matrix 119 8.3 0.000 0.000 7.347 7.486 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.346 7.486 dbcsr_mm_accdrv_process 12550 15.8 4.578 5.924 5.562 6.936 prepare_preconditioner 11 7.9 0.000 0.000 6.878 6.891 make_preconditioner 11 8.9 0.000 0.000 6.878 6.891 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.396 6.772 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.612 6.739 cp_fm_upper_to_full 72 14.2 3.195 4.624 3.195 4.624 init_scf_run 11 5.9 0.000 0.001 4.561 4.561 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.561 4.561 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.260 4.263 calculate_rho_elec 119 8.7 0.118 0.121 4.259 4.263 sum_up_and_integrate 119 10.3 0.064 0.067 4.028 4.033 integrate_v_rspace 119 11.3 0.003 0.003 3.964 3.970 qs_ot_get_p 119 10.4 0.001 0.001 3.374 3.512 mp_waitall_1 94719 16.7 2.430 3.299 2.430 3.299 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.786 3.204 calculate_first_density_matrix 1 7.0 0.000 0.000 3.080 3.082 make_m2s 4572 13.5 0.038 0.038 2.773 2.941 dbcsr_complete_redistribute 329 12.2 0.321 0.331 2.061 2.907 make_images 4572 14.5 0.353 0.385 2.652 2.821 pw_transfer 1439 11.6 0.069 0.069 2.641 2.644 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.603 2.604 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.747 2.574 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.542 2.545 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.237 2.524 apply_single 119 13.6 0.000 0.000 2.237 2.524 calculate_dm_sparse 119 9.5 0.000 0.000 2.450 2.466 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.314 2.359 mp_alltoall_i22 627 13.8 1.454 2.311 1.454 2.311 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.476 2.297 multiply_cannon_metrocomm3 9144 15.5 0.021 0.022 1.416 2.186 density_rs2pw 119 9.7 0.004 0.004 2.147 2.173 fft_wrap_pw1pw2_140 487 13.2 0.615 0.619 2.151 2.155 grid_integrate_task_list 119 12.3 2.087 2.141 2.087 2.141 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.048 2.103 ot_diis_step 108 11.5 0.014 0.015 2.043 2.043 qs_ot_p2m_diag 50 11.0 0.043 0.044 1.997 1.999 acc_transpose_blocks 9144 15.5 0.044 0.046 1.934 1.964 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.782 1.783 mp_sum_l 7287 12.8 1.061 1.718 1.061 1.718 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.639 1.640 grid_collocate_task_list 119 9.7 1.528 1.570 1.528 1.570 fft3d_ps 1201 14.6 0.595 0.607 1.559 1.564 rs_pw_transfer 974 11.9 0.009 0.009 1.449 1.492 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.454 1.460 cp_fm_cholesky_invert 11 10.9 1.442 1.446 1.442 1.446 wfi_extrapolate 11 7.9 0.001 0.001 1.412 1.412 potential_pw2rs 119 12.3 0.014 0.014 1.361 1.364 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.357 1.357 cp_fm_diag_elpa_base 50 14.0 1.211 1.265 1.355 1.356 hybrid_alltoall_any 4725 16.4 0.090 0.150 1.076 1.339 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.303 1.323 jit_kernel_multiply 6 15.5 0.956 1.304 0.956 1.304 make_images_data 4572 15.5 0.045 0.048 1.019 1.236 acc_transpose_blocks_sync 27432 16.5 1.115 1.157 1.115 1.157 mp_alltoall_d11v 2130 13.8 1.100 1.140 1.100 1.140 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.104 1.118 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 0.995 1.047 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.945 0.957 qs_create_task_list 11 7.9 0.000 0.000 0.934 0.948 generate_qs_task_list 11 8.9 0.368 0.387 0.934 0.948 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=44.530000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=718.818182, yerr=17.887620 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/72113d53c0cb11844f79f3c088677e10c62b6b2a_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 501.108736E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66212. MP_Allreduce 9776 488. MP_Sync 52 MP_Alltoall 1938 1614740. 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 0.033 0.100 84.090 84.103 qs_mol_dyn_low 1 2.0 0.050 0.385 82.410 82.429 qs_forces 11 3.9 0.004 0.006 81.951 81.952 qs_energies 11 4.9 0.001 0.002 79.081 79.096 scf_env_do_scf 11 5.9 0.000 0.001 69.460 69.462 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 64.058 64.059 dbcsr_multiply_generic 2055 12.4 0.105 0.107 51.378 51.713 qs_scf_new_mos 99 7.5 0.000 0.001 46.809 46.940 qs_scf_loop_do_ot 99 8.5 0.001 0.001 46.809 46.940 ot_scf_mini 99 9.5 0.002 0.002 44.458 44.545 multiply_cannon 2055 13.4 0.184 0.189 43.273 44.104 multiply_cannon_loop 2055 14.4 1.797 1.842 42.360 43.188 velocity_verlet 10 3.0 0.001 0.002 42.942 42.943 ot_mini 99 10.5 0.001 0.001 26.381 26.462 qs_ot_get_derivative 99 11.5 0.001 0.001 19.571 19.668 multiply_cannon_multrec 49320 15.4 11.397 12.228 17.865 18.858 rebuild_ks_matrix 110 8.3 0.000 0.001 14.547 14.687 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.547 14.687 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.741 12.865 mp_waitall_1 220248 16.4 10.554 11.424 10.554 11.424 multiply_cannon_sync_h2d 49320 15.4 9.686 10.359 9.686 10.359 qs_ot_get_p 110 10.4 0.001 0.001 9.633 9.780 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.556 8.115 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.241 7.788 apply_single 110 13.6 0.000 0.000 7.241 7.788 multiply_cannon_metrocomm3 49320 15.4 0.083 0.086 6.514 7.582 init_scf_run 11 5.9 0.000 0.001 7.448 7.449 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.448 7.448 sum_up_and_integrate 110 10.3 0.036 0.042 7.015 7.030 integrate_v_rspace 110 11.3 0.003 0.003 6.979 7.002 dbcsr_mm_accdrv_process 87628 16.1 3.024 3.106 6.336 6.949 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.562 6.715 calculate_rho_elec 110 8.6 0.021 0.026 6.561 6.714 ot_diis_step 99 11.5 0.005 0.006 6.608 6.608 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.542 6.556 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.699 5.699 init_scf_loop 11 6.9 0.005 0.014 5.375 5.376 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.206 5.265 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.155 5.176 cp_fm_diag_elpa_base 48 14.0 5.142 5.165 5.153 5.175 mp_sum_l 6594 12.7 3.865 4.567 3.865 4.567 rs_pw_transfer 902 11.9 0.011 0.013 3.544 4.284 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.004 4.103 4.108 density_rs2pw 110 9.6 0.004 0.005 3.285 4.028 wfi_extrapolate 11 7.9 0.001 0.001 4.001 4.001 make_m2s 4110 13.4 0.060 0.065 3.816 3.927 calculate_dm_sparse 110 9.5 0.001 0.001 3.785 3.892 make_images 4110 14.4 0.177 0.190 3.721 3.835 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.780 3.830 multiply_cannon_metrocomm1 49320 15.4 0.066 0.068 2.310 3.449 grid_integrate_task_list 110 12.3 3.259 3.434 3.259 3.434 jit_kernel_multiply 13 15.9 3.033 3.374 3.033 3.374 calculate_first_density_matrix 1 7.0 0.000 0.005 3.346 3.351 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.139 3.204 prepare_preconditioner 11 7.9 0.000 0.000 3.169 3.185 make_preconditioner 11 8.9 0.000 0.000 3.169 3.185 pw_transfer 1331 11.6 0.055 0.072 3.060 3.133 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 2.972 3.048 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.968 3.017 acc_transpose_blocks 49320 15.4 0.213 0.219 2.869 2.982 fft_wrap_pw1pw2_140 451 13.1 0.456 0.506 2.559 2.647 mp_waitany 14300 13.8 1.814 2.596 1.814 2.596 potential_pw2rs 110 12.3 0.006 0.006 2.524 2.539 mp_alltoall_d11v 2046 13.8 2.029 2.406 2.029 2.406 grid_collocate_task_list 110 9.6 2.156 2.365 2.156 2.365 fft3d_ps 1111 14.6 0.795 0.879 2.163 2.213 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.934 1.956 mp_sum_d 3889 11.9 1.338 1.913 1.338 1.913 make_images_data 4110 15.4 0.044 0.048 1.740 1.886 hybrid_alltoall_any 4261 16.3 0.084 0.483 1.499 1.766 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.724 1.751 cp_fm_cholesky_invert 11 10.9 1.714 1.718 1.714 1.718 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=84.103000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=475.181818, yerr=2.405228 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/72113d53c0cb11844f79f3c088677e10c62b6b2a_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 591.835136E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1972800 MPI messages size (bytes): total size 1.077520E+12 min size 0.000000E+00 max size 4.537280E+06 average size 546.188250E+03 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 222984 1826684928 8192 < size <= 32768 520356 13399818240 32768 < size <= 131072 372336 35386294272 131072 < size <= 4194304 787758 788321309808 4194304 < size <= 16777216 54450 238588003280 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66430. MP_Allreduce 9775 566. MP_Sync 52 MP_Alltoall 1717 2586898. 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 0.092 0.116 70.650 70.651 qs_mol_dyn_low 1 2.0 0.003 0.003 70.270 70.280 qs_forces 11 3.9 0.003 0.004 70.198 70.205 qs_energies 11 4.9 0.002 0.006 66.834 66.844 scf_env_do_scf 11 5.9 0.001 0.001 56.896 56.900 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 49.409 49.409 dbcsr_multiply_generic 2055 12.4 0.116 0.120 38.595 38.777 velocity_verlet 10 3.0 0.001 0.002 36.329 36.330 qs_scf_new_mos 99 7.5 0.001 0.001 33.109 33.258 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.109 33.258 multiply_cannon 2055 13.4 0.223 0.244 32.100 33.167 multiply_cannon_loop 2055 14.4 1.159 1.186 30.851 31.669 ot_scf_mini 99 9.5 0.003 0.003 31.445 31.583 ot_mini 99 10.5 0.001 0.001 18.418 18.566 multiply_cannon_multrec 24660 15.4 6.977 8.475 14.640 16.023 rebuild_ks_matrix 110 8.3 0.000 0.000 13.773 13.891 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 13.773 13.891 qs_ot_get_derivative 99 11.5 0.001 0.001 12.608 12.748 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.101 12.199 mp_waitall_1 176588 16.5 7.555 9.978 7.555 9.978 multiply_cannon_metrocomm3 24660 15.4 0.072 0.077 5.211 7.895 dbcsr_mm_accdrv_process 52282 16.1 5.987 6.817 7.492 7.785 init_scf_loop 11 6.9 0.001 0.002 7.453 7.454 init_scf_run 11 5.9 0.000 0.001 7.375 7.376 scf_env_initial_rho_setup 11 6.9 0.001 0.004 7.375 7.376 multiply_cannon_sync_h2d 24660 15.4 6.356 7.269 6.356 7.269 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.501 7.231 apply_single 110 13.6 0.000 0.001 6.501 7.231 sum_up_and_integrate 110 10.3 0.052 0.059 6.635 6.647 integrate_v_rspace 110 11.3 0.002 0.003 6.583 6.598 qs_ot_get_p 110 10.4 0.001 0.001 6.403 6.579 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.189 6.197 calculate_rho_elec 110 8.6 0.040 0.048 6.188 6.196 ot_diis_step 99 11.5 0.010 0.010 5.763 5.763 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.792 5.461 prepare_preconditioner 11 7.9 0.000 0.000 5.433 5.452 make_preconditioner 11 8.9 0.000 0.001 5.433 5.452 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.023 5.179 make_m2s 4110 13.4 0.057 0.061 4.162 4.600 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.538 4.558 make_images 4110 14.4 0.406 0.465 4.052 4.485 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.076 4.077 density_rs2pw 110 9.6 0.004 0.005 3.273 3.769 calculate_first_density_matrix 1 7.0 0.050 0.400 3.760 3.761 pw_transfer 1331 11.6 0.066 0.071 3.618 3.749 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 3.511 3.643 calculate_dm_sparse 110 9.5 0.001 0.001 3.559 3.590 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.512 3.524 cp_fm_diag_elpa_base 48 14.0 3.466 3.479 3.510 3.521 wfi_extrapolate 11 7.9 0.001 0.001 3.516 3.516 rs_pw_transfer 902 11.9 0.012 0.014 2.959 3.497 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.408 3.409 grid_integrate_task_list 110 12.3 3.166 3.356 3.166 3.356 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.264 3.338 fft_wrap_pw1pw2_140 451 13.1 0.518 0.535 3.012 3.145 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.070 3.110 hybrid_alltoall_any 4261 16.3 0.106 0.449 2.001 2.777 make_images_data 4110 15.4 0.049 0.053 2.260 2.734 fft3d_ps 1111 14.6 1.109 1.334 2.487 2.629 jit_kernel_multiply 10 16.3 1.141 2.600 1.141 2.600 grid_collocate_task_list 110 9.6 2.097 2.537 2.097 2.537 cp_fm_cholesky_invert 11 10.9 2.519 2.526 2.519 2.526 potential_pw2rs 110 12.3 0.008 0.009 2.407 2.426 mp_sum_l 6594 12.7 1.721 2.407 1.721 2.407 acc_transpose_blocks 24660 15.4 0.116 0.119 2.110 2.234 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 2.130 2.140 qs_energies_init_hamiltonians 11 5.9 0.000 0.004 2.048 2.055 mp_alltoall_d11v 2046 13.8 1.746 2.042 1.746 2.042 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.965 1.992 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.829 1.845 mp_waitany 10164 13.8 1.237 1.799 1.237 1.799 dbcsr_complete_redistribute 325 12.2 0.238 0.295 1.445 1.757 multiply_cannon_metrocomm4 22605 15.4 0.081 0.085 0.793 1.569 mp_allgather_i34 2055 14.4 0.559 1.567 0.559 1.567 rs_pw_transfer_RS2PW_140 121 11.5 0.205 0.217 0.982 1.509 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.385 1.488 mp_sum_d 3889 11.9 1.071 1.437 1.071 1.437 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=70.651000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=559.909091, yerr=6.721127 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/72113d53c0cb11844f79f3c088677e10c62b6b2a_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 668.041216E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66421. MP_Allreduce 9774 562. MP_Sync 52 MP_Alltoall 1496 4511006. MP_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.036 60.695 60.697 qs_mol_dyn_low 1 2.0 0.003 0.003 60.390 60.401 qs_forces 11 3.9 0.003 0.008 60.327 60.328 qs_energies 11 4.9 0.002 0.005 57.092 57.096 scf_env_do_scf 11 5.9 0.000 0.001 48.736 48.736 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 40.346 40.346 velocity_verlet 10 3.0 0.001 0.002 32.634 32.636 dbcsr_multiply_generic 2055 12.4 0.107 0.111 29.061 29.345 qs_scf_new_mos 99 7.5 0.001 0.001 25.253 25.368 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.253 25.367 ot_scf_mini 99 9.5 0.002 0.002 24.004 24.143 multiply_cannon 2055 13.4 0.212 0.223 22.887 23.942 multiply_cannon_loop 2055 14.4 0.814 0.856 21.685 22.728 ot_mini 99 10.5 0.001 0.001 13.817 13.959 rebuild_ks_matrix 110 8.3 0.000 0.000 12.422 12.571 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.422 12.570 multiply_cannon_multrec 16440 15.4 3.614 4.999 9.884 11.167 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.904 11.041 mp_waitall_1 139946 16.5 7.010 10.172 7.010 10.172 qs_ot_get_derivative 99 11.5 0.001 0.001 9.311 9.453 init_scf_loop 11 6.9 0.003 0.017 8.356 8.356 multiply_cannon_metrocomm3 16440 15.4 0.046 0.047 4.493 7.458 prepare_preconditioner 11 7.9 0.000 0.000 6.584 6.603 make_preconditioner 11 8.9 0.000 0.000 6.583 6.603 sum_up_and_integrate 110 10.3 0.060 0.061 6.556 6.571 integrate_v_rspace 110 11.3 0.003 0.003 6.497 6.511 dbcsr_mm_accdrv_process 34862 16.1 5.447 5.816 6.114 6.313 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.935 6.267 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.025 6.032 calculate_rho_elec 110 8.6 0.059 0.060 6.024 6.031 init_scf_run 11 5.9 0.000 0.001 5.915 5.915 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.914 5.915 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.000 5.574 apply_single 110 13.6 0.000 0.000 4.999 5.574 qs_ot_get_p 110 10.4 0.001 0.001 5.350 5.502 make_m2s 4110 13.4 0.049 0.051 4.116 4.535 ot_diis_step 99 11.5 0.010 0.011 4.476 4.476 make_images 4110 14.4 0.396 0.518 4.001 4.420 density_rs2pw 110 9.6 0.004 0.005 3.091 4.372 rs_pw_transfer 902 11.9 0.010 0.011 2.681 3.958 multiply_cannon_sync_h2d 16440 15.4 3.226 3.862 3.226 3.862 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.245 3.822 qs_ot_p2m_diag 48 11.0 0.042 0.045 3.748 3.752 pw_transfer 1331 11.6 0.066 0.074 3.644 3.650 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.536 3.546 grid_integrate_task_list 110 12.3 3.181 3.444 3.181 3.444 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.379 3.379 fft_wrap_pw1pw2_140 451 13.1 0.636 0.642 3.064 3.075 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.991 2.993 wfi_extrapolate 11 7.9 0.001 0.001 2.980 2.980 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.846 2.856 cp_fm_diag_elpa_base 48 14.0 2.781 2.808 2.845 2.854 calculate_first_density_matrix 1 7.0 0.000 0.002 2.843 2.845 make_images_data 4110 15.4 0.045 0.050 2.312 2.799 hybrid_alltoall_any 4261 16.3 0.109 0.384 2.061 2.726 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.653 2.705 calculate_dm_sparse 110 9.5 0.001 0.001 2.585 2.615 cp_fm_cholesky_invert 11 10.9 2.568 2.574 2.568 2.574 mp_waitany 17072 13.8 1.219 2.560 1.219 2.560 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.438 2.517 grid_collocate_task_list 110 9.6 2.124 2.514 2.124 2.514 multiply_cannon_metrocomm4 14385 15.4 0.048 0.052 0.895 2.472 fft3d_ps 1111 14.6 1.091 1.098 2.361 2.371 mp_irecv_dv 48980 15.7 0.820 2.339 0.820 2.339 potential_pw2rs 110 12.3 0.011 0.011 2.261 2.266 rs_pw_transfer_RS2PW_140 121 11.5 0.177 0.181 0.939 2.211 mp_sum_l 6594 12.7 1.437 2.161 1.437 2.161 mp_alltoall_d11v 2046 13.8 1.758 2.099 1.758 2.099 qs_energies_init_hamiltonians 11 5.9 0.001 0.006 1.998 1.998 dbcsr_complete_redistribute 325 12.2 0.334 0.374 1.394 1.841 acc_transpose_blocks 16440 15.4 0.078 0.081 1.758 1.826 cp_fm_upper_to_full 70 14.2 1.371 1.740 1.371 1.740 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.710 1.727 mp_allgather_i34 2055 14.4 0.516 1.557 0.516 1.557 cp_fm_cholesky_decompose 22 10.9 1.532 1.552 1.532 1.552 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.522 1.534 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.377 1.466 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.947 1.385 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.284 1.292 rs_gather_matrices 110 12.3 0.231 0.259 0.973 1.291 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=60.697000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=632.909091, yerr=8.805520 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/72113d53c0cb11844f79f3c088677e10c62b6b2a_performance_tests/12/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 601.317074E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 4916280 0.0% 0.0% 100.0% average stack size 0.0 0.0 200.2 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 736.710656E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66419. MP_Allreduce 9774 603. MP_Sync 52 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.031 66.903 66.908 qs_mol_dyn_low 1 2.0 0.003 0.012 66.358 66.369 qs_forces 11 3.9 0.021 0.036 66.292 66.295 qs_energies 11 4.9 0.001 0.002 62.818 62.826 scf_env_do_scf 11 5.9 0.000 0.001 54.074 54.077 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 42.372 42.373 velocity_verlet 10 3.0 0.002 0.002 37.396 37.398 dbcsr_multiply_generic 2055 12.4 0.115 0.121 31.132 31.360 qs_scf_new_mos 99 7.5 0.001 0.001 27.231 27.334 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.230 27.334 ot_scf_mini 99 9.5 0.003 0.003 25.553 25.669 multiply_cannon 2055 13.4 0.238 0.260 23.989 25.118 multiply_cannon_loop 2055 14.4 1.399 1.491 22.573 23.240 ot_mini 99 10.5 0.001 0.001 14.710 14.845 multiply_cannon_multrec 24660 15.4 4.122 6.819 13.227 14.768 rebuild_ks_matrix 110 8.3 0.000 0.000 12.362 12.463 qs_ks_build_kohn_sham_matrix 110 9.3 0.020 0.075 12.361 12.462 init_scf_loop 11 6.9 0.001 0.003 11.662 11.663 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.017 11.124 qs_ot_get_derivative 99 11.5 0.001 0.001 10.540 10.653 dbcsr_mm_accdrv_process 52304 16.0 7.762 9.433 8.948 10.147 prepare_preconditioner 11 7.9 0.000 0.000 9.695 9.709 make_preconditioner 11 8.9 0.000 0.000 9.695 9.709 make_full_inverse_cholesky 11 9.9 0.000 0.000 7.913 9.385 mp_waitall_1 121746 16.5 4.487 6.581 4.487 6.581 sum_up_and_integrate 110 10.3 0.067 0.070 6.503 6.515 integrate_v_rspace 110 11.3 0.003 0.003 6.436 6.451 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.163 6.175 calculate_rho_elec 110 8.6 0.078 0.081 6.163 6.174 init_scf_run 11 5.9 0.000 0.001 6.007 6.008 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.007 6.008 make_m2s 4110 13.4 0.059 0.061 5.497 5.881 qs_ot_get_p 110 10.4 0.001 0.001 5.682 5.848 make_images 4110 14.4 0.573 0.696 5.356 5.738 cp_fm_upper_to_full 70 14.2 3.319 4.839 3.319 4.839 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.048 4.148 apply_single 110 13.6 0.000 0.000 4.048 4.148 ot_diis_step 99 11.5 0.011 0.011 4.130 4.130 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.840 3.857 dbcsr_complete_redistribute 325 12.2 0.424 0.472 2.644 3.794 pw_transfer 1331 11.6 0.065 0.074 3.727 3.758 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.621 3.656 density_rs2pw 110 9.6 0.004 0.004 3.017 3.542 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.483 3.541 grid_integrate_task_list 110 12.3 3.275 3.436 3.275 3.436 multiply_cannon_metrocomm3 24660 15.4 0.038 0.040 1.398 3.376 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.310 3.372 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.283 3.285 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.280 3.281 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.120 3.259 hybrid_alltoall_any 4261 16.3 0.122 0.458 2.288 3.156 fft_wrap_pw1pw2_140 451 13.1 0.666 0.689 3.119 3.155 make_images_data 4110 15.4 0.048 0.052 2.760 3.153 calculate_dm_sparse 110 9.5 0.001 0.001 3.021 3.065 wfi_extrapolate 11 7.9 0.001 0.001 3.003 3.003 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.926 2.960 calculate_first_density_matrix 1 7.0 0.000 0.000 2.904 2.908 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.773 2.895 rs_pw_transfer 902 11.9 0.010 0.011 2.353 2.880 mp_alltoall_i22 605 13.7 1.666 2.845 1.666 2.845 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.759 2.773 cp_fm_diag_elpa_base 48 14.0 2.612 2.673 2.757 2.772 cp_fm_cholesky_invert 11 10.9 2.585 2.592 2.585 2.592 acc_transpose_blocks 24660 15.4 0.110 0.114 2.407 2.590 multiply_cannon_sync_h2d 24660 15.4 2.379 2.510 2.379 2.510 grid_collocate_task_list 110 9.6 2.223 2.469 2.223 2.469 fft3d_ps 1111 14.6 1.083 1.115 2.385 2.405 qs_energies_init_hamiltonians 11 5.9 0.000 0.002 2.247 2.252 potential_pw2rs 110 12.3 0.012 0.013 2.136 2.141 mp_alltoall_d11v 2046 13.8 1.728 1.991 1.728 1.991 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.778 1.820 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.703 1.716 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.002 1.611 1.705 jit_kernel_multiply 9 15.8 0.842 1.704 0.842 1.704 cp_fm_cholesky_decompose 22 10.9 1.612 1.659 1.612 1.659 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.585 1.600 multiply_cannon_metrocomm4 20550 15.4 0.063 0.066 0.854 1.553 mp_waitany 13376 13.8 1.062 1.547 1.062 1.547 acc_transpose_blocks_sync 73980 16.4 1.374 1.541 1.374 1.541 mp_sum_l 6594 12.7 0.984 1.536 0.984 1.536 mp_irecv_dv 62702 16.1 0.748 1.464 0.748 1.464 mp_allgather_i34 2055 14.4 0.500 1.428 0.500 1.428 rs_pw_transfer_RS2PW_140 121 11.5 0.170 0.176 0.825 1.364 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=66.908000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=697.090909, yerr=11.881648 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/72113d53c0cb11844f79f3c088677e10c62b6b2a_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 854.581248E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66417. MP_Allreduce 9774 644. MP_Sync 52 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.030 56.236 56.238 qs_mol_dyn_low 1 2.0 0.003 0.003 55.999 56.010 qs_forces 11 3.9 0.003 0.003 55.935 55.936 qs_energies 11 4.9 0.001 0.001 52.210 52.214 scf_env_do_scf 11 5.9 0.000 0.001 43.447 43.447 scf_env_do_scf_inner_loop 99 6.5 0.009 0.013 35.815 35.816 velocity_verlet 10 3.0 0.002 0.002 31.257 31.260 dbcsr_multiply_generic 2055 12.4 0.107 0.112 23.889 24.013 qs_scf_new_mos 99 7.5 0.001 0.001 20.784 20.827 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.783 20.827 multiply_cannon 2055 13.4 0.236 0.250 18.393 19.788 ot_scf_mini 99 9.5 0.002 0.002 19.541 19.563 multiply_cannon_loop 2055 14.4 0.601 0.630 17.135 17.440 rebuild_ks_matrix 110 8.3 0.000 0.000 11.902 11.931 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 11.902 11.930 ot_mini 99 10.5 0.001 0.001 11.040 11.057 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.556 10.581 multiply_cannon_multrec 8220 15.4 3.563 4.893 8.300 9.144 init_scf_loop 11 6.9 0.000 0.001 7.585 7.588 mp_waitall_1 103326 16.6 5.708 7.398 5.708 7.398 qs_ot_get_derivative 99 11.5 0.001 0.001 7.367 7.390 sum_up_and_integrate 110 10.3 0.079 0.080 6.497 6.512 integrate_v_rspace 110 11.3 0.003 0.003 6.419 6.433 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.287 6.302 calculate_rho_elec 110 8.6 0.114 0.115 6.286 6.301 dbcsr_mm_accdrv_process 17442 15.9 3.148 4.605 4.599 5.967 prepare_preconditioner 11 7.9 0.000 0.000 5.917 5.921 make_preconditioner 11 8.9 0.000 0.000 5.917 5.921 init_scf_run 11 5.9 0.000 0.001 5.633 5.633 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.632 5.632 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.509 5.586 qs_ot_get_p 110 10.4 0.001 0.001 4.566 4.576 make_m2s 4110 13.4 0.038 0.039 4.189 4.488 make_images 4110 14.4 0.648 0.704 4.060 4.359 multiply_cannon_metrocomm3 8220 15.4 0.019 0.020 2.960 4.284 pw_transfer 1331 11.6 0.066 0.074 3.906 3.939 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.799 3.835 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.715 3.773 apply_single 110 13.6 0.000 0.000 3.715 3.773 ot_diis_step 99 11.5 0.012 0.012 3.648 3.649 grid_integrate_task_list 110 12.3 3.361 3.539 3.361 3.539 density_rs2pw 110 9.6 0.004 0.004 2.997 3.367 fft_wrap_pw1pw2_140 451 13.1 0.831 0.843 3.297 3.351 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.228 3.233 cp_dbcsr_syevd 48 12.0 0.003 0.003 2.892 2.893 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.873 2.874 cp_fm_cholesky_invert 11 10.9 2.824 2.828 2.824 2.828 calculate_first_density_matrix 1 7.0 0.000 0.000 2.818 2.819 hybrid_alltoall_any 4261 16.3 0.201 0.853 2.222 2.738 wfi_extrapolate 11 7.9 0.001 0.001 2.714 2.714 qs_energies_init_hamiltonians 11 5.9 0.003 0.010 2.675 2.676 make_images_data 4110 15.4 0.041 0.047 2.250 2.640 grid_collocate_task_list 110 9.6 2.315 2.571 2.315 2.571 rs_pw_transfer 902 11.9 0.010 0.010 2.143 2.541 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.512 2.528 calculate_dm_sparse 110 9.5 0.001 0.001 2.465 2.501 multiply_cannon_sync_h2d 8220 15.4 2.387 2.488 2.387 2.488 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.390 2.397 cp_fm_diag_elpa_base 48 14.0 2.335 2.358 2.389 2.396 fft3d_ps 1111 14.6 1.134 1.196 2.347 2.366 potential_pw2rs 110 12.3 0.015 0.015 2.030 2.034 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.784 2.021 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.984 2.012 mp_alltoall_d11v 2046 13.8 1.619 1.905 1.619 1.905 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.881 1.896 jit_kernel_multiply 8 15.6 1.140 1.895 1.140 1.895 cp_fm_cholesky_decompose 22 10.9 1.675 1.691 1.675 1.691 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.674 1.677 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.497 1.615 acc_transpose_blocks 8220 15.4 0.039 0.040 1.514 1.595 mp_allgather_i34 2055 14.4 0.433 1.529 0.433 1.529 dbcsr_complete_redistribute 325 12.2 0.567 0.603 1.408 1.506 mp_waitany 9240 13.8 1.064 1.486 1.064 1.486 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.427 1.440 qs_create_task_list 11 7.9 0.000 0.001 1.214 1.319 generate_qs_task_list 11 8.9 0.375 0.445 1.213 1.318 rs_pw_transfer_RS2PW_140 121 11.5 0.162 0.166 0.808 1.207 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.132 1.152 rs_gather_matrices 110 12.3 0.321 0.365 0.939 1.150 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=56.238000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=805.636364, yerr=13.178997 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/72113d53c0cb11844f79f3c088677e10c62b6b2a_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 1.385460E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67098. MP_Allreduce 9752 812. MP_Sync 52 MP_Alltoall 1474 16505187. MP_SendRecv 2310 360267. MP_ISendRecv 2310 360267. MP_Wait 5214 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.028 0.055 89.997 89.999 qs_mol_dyn_low 1 2.0 0.003 0.004 89.595 89.604 qs_forces 11 3.9 0.004 0.005 89.525 89.527 qs_energies 11 4.9 0.002 0.002 85.283 85.286 scf_env_do_scf 11 5.9 0.001 0.001 74.314 74.314 velocity_verlet 10 3.0 0.002 0.002 57.143 57.150 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 45.579 45.581 dbcsr_multiply_generic 2055 12.4 0.120 0.126 30.314 30.366 init_scf_loop 11 6.9 0.001 0.001 28.663 28.665 qs_scf_new_mos 99 7.5 0.001 0.001 27.108 27.143 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.107 27.143 prepare_preconditioner 11 7.9 0.000 0.000 26.559 26.567 make_preconditioner 11 8.9 0.000 0.000 26.559 26.567 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.716 26.039 ot_scf_mini 99 9.5 0.002 0.002 25.292 25.325 multiply_cannon 2055 13.4 0.332 0.364 23.025 23.727 multiply_cannon_loop 2055 14.4 0.832 0.849 21.222 21.661 cp_fm_upper_to_full 70 14.2 12.859 18.423 12.859 18.423 rebuild_ks_matrix 110 8.3 0.000 0.000 14.316 14.364 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 14.315 14.364 ot_mini 99 10.5 0.001 0.001 14.264 14.292 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.943 12.987 dbcsr_complete_redistribute 325 12.2 1.054 1.089 7.453 10.754 multiply_cannon_multrec 8220 15.4 4.048 4.231 9.946 10.066 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.425 9.738 qs_ot_get_derivative 99 11.5 0.001 0.001 9.564 9.595 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.827 9.108 mp_alltoall_i22 605 13.7 5.448 8.783 5.448 8.783 mp_waitall_1 84994 16.7 7.788 8.613 7.788 8.613 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.802 7.843 calculate_rho_elec 110 8.6 0.224 0.224 7.802 7.843 sum_up_and_integrate 110 10.3 0.150 0.151 7.423 7.438 integrate_v_rspace 110 11.3 0.004 0.004 7.273 7.288 init_scf_run 11 5.9 0.000 0.001 6.405 6.405 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.404 6.404 make_m2s 4110 13.4 0.043 0.044 5.534 6.042 dbcsr_mm_accdrv_process 11614 15.7 3.852 4.115 5.751 5.966 make_images 4110 14.4 0.880 0.921 5.345 5.852 qs_ot_get_p 110 10.4 0.024 0.027 5.719 5.747 cp_fm_cholesky_invert 11 10.9 5.358 5.362 5.358 5.362 multiply_cannon_metrocomm3 8220 15.4 0.019 0.020 4.899 5.338 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.839 5.282 apply_single 110 13.6 0.000 0.000 4.839 5.282 pw_transfer 1331 11.6 0.075 0.076 5.021 5.030 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 4.904 4.913 ot_diis_step 99 11.5 0.015 0.015 4.675 4.675 fft_wrap_pw1pw2_140 451 13.1 1.333 1.336 4.275 4.284 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.086 4.092 qs_energies_init_hamiltonians 11 5.9 0.008 0.012 3.955 3.957 grid_integrate_task_list 110 12.3 3.737 3.851 3.737 3.851 density_rs2pw 110 9.6 0.004 0.004 3.763 3.825 calculate_dm_sparse 110 9.5 0.001 0.001 3.734 3.758 hybrid_alltoall_any 4261 16.3 0.264 0.565 2.984 3.751 make_images_data 4110 15.4 0.044 0.048 3.000 3.644 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.569 3.569 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.032 3.492 wfi_extrapolate 11 7.9 0.001 0.001 3.412 3.412 multiply_cannon_sync_h2d 8220 15.4 3.114 3.154 3.114 3.154 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.979 2.979 cp_fm_diag_elpa_base 48 14.0 2.439 2.636 2.977 2.977 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.960 2.961 calculate_first_density_matrix 1 7.0 0.000 0.000 2.872 2.872 fft3d_ps 1111 14.6 1.299 1.308 2.837 2.848 grid_collocate_task_list 110 9.6 2.719 2.804 2.719 2.804 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.671 2.688 qs_env_update_s_mstruct 11 6.9 0.244 0.252 2.458 2.506 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.416 2.435 rs_pw_transfer 902 11.9 0.010 0.011 2.332 2.420 potential_pw2rs 110 12.3 0.021 0.021 2.406 2.413 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 2.218 2.234 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.105 2.179 mp_alltoall_d11v 2046 13.8 1.979 2.112 1.979 2.112 acc_transpose_blocks 8220 15.4 0.041 0.041 2.061 2.094 cp_fm_cholesky_decompose 22 10.9 2.003 2.022 2.003 2.022 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.972 1.977 jit_kernel_multiply 10 15.2 1.704 1.970 1.704 1.970 qs_create_task_list 11 7.9 0.004 0.004 1.881 1.929 generate_qs_task_list 11 8.9 0.730 0.781 1.877 1.925 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=89.999000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1231.181818, yerr=66.534773 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/72113d53c0cb11844f79f3c088677e10c62b6b2a_performance_tests/15/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 1.094965E+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 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 629.403648E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175954870160 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 57905. MP_Allreduce 11059 797. MP_Sync 87 MP_Alltoall 2226 2185448. MP_SendRecv 24320 18752. MP_ISendRecv 24320 18752. MP_Wait 42476 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.050 209.186 209.187 qs_mol_dyn_low 1 2.0 0.003 0.004 207.911 208.173 qs_forces 11 3.9 0.005 0.005 207.826 207.827 qs_energies 11 4.9 0.001 0.002 202.166 202.191 scf_env_do_scf 11 5.9 0.001 0.001 185.082 185.086 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 164.665 164.667 dbcsr_multiply_generic 2507 12.6 0.180 0.184 126.083 127.349 qs_scf_new_mos 117 7.6 0.001 0.001 124.359 124.687 qs_scf_loop_do_ot 117 8.6 0.001 0.001 124.358 124.686 velocity_verlet 10 3.0 0.001 0.002 124.115 124.116 ot_scf_mini 117 9.6 0.003 0.003 117.672 117.993 multiply_cannon 2507 13.6 0.237 0.248 102.584 105.341 multiply_cannon_loop 2507 14.6 2.381 2.490 100.404 103.203 ot_mini 117 10.6 0.001 0.001 66.277 66.562 multiply_cannon_multrec 60168 15.6 32.229 34.352 42.519 44.107 qs_ot_get_derivative 117 11.6 0.001 0.001 41.260 41.552 rebuild_ks_matrix 128 8.3 0.001 0.001 33.837 34.382 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.018 33.837 34.381 mp_waitall_1 267128 16.5 28.667 31.311 28.667 31.311 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.352 30.874 qs_ot_get_p 128 10.4 0.001 0.002 29.164 29.549 multiply_cannon_sync_h2d 60168 15.6 26.618 28.703 26.618 28.703 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.514 25.491 apply_single 128 13.6 0.001 0.001 24.514 25.490 ot_diis_step 117 11.6 0.007 0.008 24.751 24.752 qs_ot_p2m_diag 83 11.4 0.079 0.092 22.535 22.622 init_scf_loop 11 6.9 0.000 0.001 20.343 20.344 cp_dbcsr_syevd 83 12.4 0.005 0.005 19.834 19.834 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.163 19.363 multiply_cannon_metrocomm3 60168 15.6 0.117 0.125 15.899 18.793 cp_fm_diag_elpa 83 13.4 0.000 0.000 16.865 16.893 cp_fm_diag_elpa_base 83 14.4 16.796 16.832 16.861 16.887 prepare_preconditioner 11 7.9 0.000 0.000 15.720 15.754 make_preconditioner 11 8.9 0.000 0.000 15.720 15.754 make_full_inverse_cholesky 11 9.9 0.000 0.000 14.926 15.110 sum_up_and_integrate 128 10.3 0.088 0.107 14.244 14.260 integrate_v_rspace 128 11.3 0.004 0.004 14.155 14.173 make_m2s 5014 13.6 0.105 0.113 13.741 14.109 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.873 14.009 calculate_rho_elec 128 8.7 0.046 0.065 13.872 14.008 make_images 5014 14.6 0.399 0.419 13.558 13.935 init_scf_run 11 5.9 0.000 0.001 12.934 12.935 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.934 12.935 dbcsr_mm_accdrv_process 124484 16.2 4.563 4.898 9.853 10.540 density_rs2pw 128 9.7 0.006 0.007 7.115 10.327 mp_sum_l 7950 12.9 8.503 9.538 8.503 9.538 wfi_extrapolate 11 7.9 0.001 0.001 9.112 9.112 rs_pw_transfer 1046 11.9 0.016 0.019 5.779 8.976 cp_fm_cholesky_invert 11 10.9 8.943 8.950 8.943 8.950 calculate_dm_sparse 128 9.5 0.001 0.001 8.861 8.936 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.372 8.511 multiply_cannon_metrocomm1 60168 15.6 0.096 0.101 6.316 8.360 qs_ot_get_orbitals 117 10.6 0.001 0.001 8.041 8.143 pw_transfer 1547 11.6 0.074 0.096 7.765 7.944 fft_wrap_pw1pw2 1291 12.7 0.011 0.013 7.562 7.745 make_images_data 5014 15.6 0.068 0.073 6.727 7.657 grid_integrate_task_list 128 12.3 7.031 7.653 7.031 7.653 hybrid_alltoall_any 5200 16.5 0.295 2.266 5.890 7.271 fft_wrap_pw1pw2_140 523 13.2 1.274 1.321 6.625 6.823 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.744 6.758 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.873 6.005 mp_waitany 16020 13.9 2.708 5.995 2.708 5.995 grid_collocate_task_list 128 9.7 4.684 5.932 4.684 5.932 fft3d_ps 1291 14.7 2.144 2.787 5.328 5.627 mp_alltoall_d11v 2415 14.1 4.310 5.505 4.310 5.505 rs_pw_transfer_RS2PW_140 139 11.5 0.279 0.304 2.143 5.351 potential_pw2rs 128 12.3 0.009 0.011 4.596 4.622 cp_fm_cholesky_decompose 22 10.9 4.553 4.569 4.553 4.569 mp_sum_d 4465 12.1 3.476 4.361 3.476 4.361 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=209.187000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=596.363636, yerr=5.927942 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/72113d53c0cb11844f79f3c088677e10c62b6b2a_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 844.103680E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100943E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834845264 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 58214. MP_Allreduce 11083 960. MP_Sync 87 MP_Alltoall 1969 6645697. 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 0.014 0.030 190.139 190.140 qs_mol_dyn_low 1 2.0 0.003 0.003 189.792 189.807 qs_forces 11 3.9 0.004 0.005 189.710 189.711 qs_energies 11 4.9 0.009 0.011 182.848 182.858 scf_env_do_scf 11 5.9 0.001 0.001 165.603 165.613 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 133.292 133.294 velocity_verlet 10 3.0 0.001 0.002 119.099 119.101 dbcsr_multiply_generic 2507 12.6 0.191 0.196 97.739 98.975 qs_scf_new_mos 117 7.6 0.001 0.001 93.548 94.004 qs_scf_loop_do_ot 117 8.6 0.001 0.001 93.548 94.004 ot_scf_mini 117 9.6 0.004 0.004 88.736 89.292 multiply_cannon 2507 13.6 0.503 0.562 77.834 82.321 multiply_cannon_loop 2507 14.6 1.561 1.633 74.454 76.932 ot_mini 117 10.6 0.001 0.001 50.006 50.498 mp_waitall_1 214728 16.6 23.576 39.019 23.576 39.019 multiply_cannon_multrec 30084 15.6 21.214 25.179 32.219 36.507 rebuild_ks_matrix 128 8.3 0.001 0.001 32.980 33.592 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.019 32.979 33.591 init_scf_loop 11 6.9 0.001 0.001 32.223 32.224 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.606 30.164 multiply_cannon_metrocomm3 30084 15.6 0.097 0.102 15.381 28.701 qs_ot_get_derivative 117 11.6 0.001 0.002 28.114 28.656 prepare_preconditioner 11 7.9 0.000 0.000 27.875 27.947 make_preconditioner 11 8.9 0.000 0.000 27.875 27.947 make_full_inverse_cholesky 11 9.9 0.000 0.000 26.562 27.094 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.050 23.208 apply_single 128 13.6 0.001 0.001 22.049 23.207 ot_diis_step 117 11.6 0.014 0.015 21.719 21.721 qs_ot_get_p 128 10.4 0.001 0.001 20.760 21.393 multiply_cannon_sync_h2d 30084 15.6 18.161 20.698 18.161 20.698 cp_fm_cholesky_invert 11 10.9 16.261 16.273 16.261 16.273 qs_ot_p2m_diag 83 11.4 0.189 0.217 15.918 15.952 make_m2s 5014 13.6 0.090 0.096 14.069 15.769 make_images 5014 14.6 1.161 1.357 13.855 15.553 cp_dbcsr_syevd 83 12.4 0.005 0.006 14.723 14.724 sum_up_and_integrate 128 10.3 0.116 0.132 14.638 14.668 integrate_v_rspace 128 11.3 0.004 0.004 14.522 14.553 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.072 14.097 calculate_rho_elec 128 8.7 0.088 0.104 14.072 14.096 init_scf_run 11 5.9 0.000 0.001 12.412 12.413 scf_env_initial_rho_setup 11 6.9 0.001 0.002 12.412 12.413 cp_fm_diag_elpa 83 13.4 0.000 0.000 11.598 11.637 cp_fm_diag_elpa_base 83 14.4 11.346 11.438 11.593 11.627 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 11.127 11.541 multiply_cannon_metrocomm4 27577 15.6 0.111 0.125 3.818 11.064 dbcsr_mm_accdrv_process 62242 16.2 5.642 6.365 10.451 10.995 mp_irecv_dv 69486 16.3 3.606 10.648 3.606 10.648 make_images_data 5014 15.6 0.067 0.079 8.249 10.394 density_rs2pw 128 9.7 0.007 0.007 7.378 9.865 hybrid_alltoall_any 5200 16.5 0.349 1.512 6.985 9.712 pw_transfer 1547 11.6 0.086 0.098 8.729 8.785 fft_wrap_pw1pw2 1291 12.7 0.011 0.012 8.504 8.564 wfi_extrapolate 11 7.9 0.001 0.001 8.359 8.359 rs_pw_transfer 1046 11.9 0.015 0.017 5.597 8.095 fft_wrap_pw1pw2_140 523 13.2 1.334 1.354 7.533 7.611 grid_integrate_task_list 128 12.3 7.177 7.603 7.177 7.603 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.262 7.113 calculate_dm_sparse 128 9.5 0.001 0.001 6.926 7.086 cp_fm_cholesky_decompose 22 10.9 6.774 6.871 6.774 6.871 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.173 6.183 grid_collocate_task_list 128 9.7 4.784 5.981 4.784 5.981 fft3d_ps 1291 14.7 2.815 2.992 5.814 5.866 mp_sum_l 7950 12.9 4.031 5.729 4.031 5.729 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.297 5.469 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.355 5.430 mp_alltoall_d11v 2415 14.1 4.094 5.008 4.094 5.008 mp_waitany 11748 13.9 2.479 4.964 2.479 4.964 mp_allgather_i34 2507 14.6 1.906 4.937 1.906 4.937 potential_pw2rs 128 12.3 0.016 0.019 4.862 4.878 rs_pw_transfer_RS2PW_140 139 11.5 0.351 0.376 2.095 4.558 dbcsr_complete_redistribute 395 12.7 0.777 0.873 3.088 3.932 mp_sum_d 4470 12.1 2.574 3.878 2.574 3.878 calculate_first_density_matrix 1 7.0 0.000 0.000 3.865 3.867 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.763 3.812 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=190.140000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=802.727273, yerr=1.813631 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/72113d53c0cb11844f79f3c088677e10c62b6b2a_performance_tests/17/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 2.950886E+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 4015680 0.0% 0.0% 100.0% average stack size 0.0 0.0 1694.9 marketing flops 145.646636E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 966.037504E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1051232 MPI messages size (bytes): total size 2.737039E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.603648E+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 284160 37245419520 131072 < size <= 4194304 665408 1004401590272 4194304 < size <= 16777216 66080 937889582784 16777216 < size 28896 757491302400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4030 57812. MP_Allreduce 11152 997. MP_Sync 87 MP_Alltoall 1724 9394160. MP_SendRecv 7998 75008. MP_ISendRecv 7998 75008. MP_Wait 21986 MP_ISend 11836 275177. MP_IRecv 11836 275177. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.046 0.075 177.040 177.040 qs_mol_dyn_low 1 2.0 0.003 0.003 176.212 176.227 qs_forces 11 3.9 0.004 0.007 176.120 176.124 qs_energies 11 4.9 0.002 0.009 169.476 169.486 scf_env_do_scf 11 5.9 0.001 0.002 153.518 153.519 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 119.018 119.018 velocity_verlet 10 3.0 0.002 0.002 113.286 113.288 dbcsr_multiply_generic 2527 12.6 0.183 0.187 82.903 84.048 qs_scf_new_mos 118 7.6 0.001 0.001 80.230 80.557 qs_scf_loop_do_ot 118 8.6 0.001 0.001 80.230 80.556 ot_scf_mini 118 9.6 0.003 0.004 75.987 76.370 multiply_cannon 2527 13.6 0.514 0.532 63.198 67.916 multiply_cannon_loop 2527 14.6 1.135 1.192 60.108 62.867 ot_mini 118 10.6 0.001 0.001 42.948 43.355 mp_waitall_1 171878 16.6 24.824 34.472 24.824 34.472 init_scf_loop 11 6.9 0.003 0.019 34.404 34.405 rebuild_ks_matrix 129 8.3 0.001 0.001 31.184 31.702 qs_ks_build_kohn_sham_matrix 129 9.3 0.016 0.017 31.183 31.702 prepare_preconditioner 11 7.9 0.000 0.000 30.308 30.362 make_preconditioner 11 8.9 0.000 0.001 30.307 30.362 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.846 29.338 qs_ks_update_qs_env 129 7.6 0.001 0.001 28.058 28.535 multiply_cannon_multrec 20216 15.6 13.212 16.246 22.809 25.769 multiply_cannon_metrocomm3 20216 15.6 0.062 0.066 15.476 25.134 qs_ot_get_derivative 118 11.6 0.001 0.002 23.246 23.637 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.829 20.775 apply_single 129 13.6 0.001 0.001 19.828 20.774 ot_diis_step 118 11.6 0.018 0.018 19.590 19.590 qs_ot_get_p 129 10.4 0.001 0.002 18.245 18.696 make_m2s 5054 13.6 0.080 0.087 14.592 15.967 make_images 5054 14.6 1.145 1.252 14.354 15.727 multiply_cannon_sync_h2d 20216 15.6 13.811 15.460 13.811 15.460 sum_up_and_integrate 129 10.3 0.134 0.145 14.711 14.737 integrate_v_rspace 129 11.3 0.004 0.004 14.576 14.600 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.560 14.591 calculate_rho_elec 129 8.7 0.131 0.146 14.559 14.590 cp_fm_cholesky_invert 11 10.9 14.004 14.012 14.004 14.012 qs_ot_p2m_diag 83 11.4 0.265 0.272 13.916 13.924 cp_dbcsr_syevd 83 12.4 0.005 0.006 12.873 12.874 init_scf_run 11 5.9 0.000 0.001 10.962 10.962 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.962 10.962 make_images_data 5054 15.6 0.062 0.074 8.739 10.568 hybrid_alltoall_any 5240 16.5 0.451 2.053 7.760 9.817 cp_fm_diag_elpa 83 13.4 0.000 0.001 9.745 9.765 cp_fm_diag_elpa_base 83 14.4 9.340 9.483 9.742 9.762 multiply_cannon_metrocomm4 17689 15.6 0.067 0.077 3.487 9.532 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.103 9.376 mp_irecv_dv 50625 16.2 3.357 9.273 3.357 9.273 dbcsr_mm_accdrv_process 41830 16.2 5.939 6.253 9.053 9.243 density_rs2pw 129 9.7 0.006 0.006 7.153 9.233 pw_transfer 1559 11.6 0.086 0.103 8.905 9.008 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 8.679 8.786 grid_integrate_task_list 129 12.3 7.346 7.882 7.346 7.882 fft_wrap_pw1pw2_140 527 13.2 1.425 1.447 7.724 7.843 wfi_extrapolate 11 7.9 0.001 0.001 7.492 7.492 cp_fm_upper_to_full 105 14.8 5.670 7.292 5.670 7.292 cp_fm_cholesky_decompose 22 10.9 7.127 7.171 7.127 7.171 rs_pw_transfer 1054 12.0 0.014 0.014 5.031 7.138 dbcsr_complete_redistribute 395 12.7 1.165 1.189 4.641 6.388 calculate_dm_sparse 129 9.5 0.001 0.001 6.046 6.160 grid_collocate_task_list 129 9.7 5.033 5.869 5.033 5.869 fft3d_ps 1301 14.7 2.765 2.995 5.781 5.863 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.499 5.503 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 4.676 5.375 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.504 5.252 mp_alltoall_d11v 2423 14.1 4.511 5.238 4.511 5.238 mp_sum_l 8010 12.9 3.268 5.003 3.268 5.003 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.613 4.760 potential_pw2rs 129 12.3 0.020 0.023 4.649 4.660 mp_allgather_i34 2527 14.6 1.580 4.506 1.580 4.506 mp_waitany 11836 13.9 2.316 4.454 2.316 4.454 transfer_fm_to_dbcsr 11 9.9 0.016 0.021 2.442 4.176 rs_pw_transfer_RS2PW_140 140 11.5 0.335 0.357 1.906 3.998 mp_alltoall_i22 718 14.1 2.063 3.962 2.063 3.962 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.927 3.953 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 3.733 3.734 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.631 3.675 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=177.040000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=913.545455, yerr=7.935692 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/72113d53c0cb11844f79f3c088677e10c62b6b2a_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410023282688 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444707676160 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796579E+12 0.0% 0.0% 100.0% flops max/rank 4.320339E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500928 0.0% 0.0% 100.0% number of processed stacks 5927808 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.2 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.165672E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1133160 MPI messages size (bytes): total size 2.008142E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.772161E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 315952 35695099904 131072 < size <= 4194304 709496 778939400192 4194304 < size <= 16777216 69840 660837789680 16777216 < size 30480 532676608000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4003 58189. MP_Allreduce 11085 1083. MP_Sync 86 MP_Alltoall 1700 12496381. 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 0.046 188.673 188.675 qs_mol_dyn_low 1 2.0 0.003 0.004 188.242 188.256 qs_forces 11 3.9 0.004 0.004 188.092 188.097 qs_energies 11 4.9 0.001 0.002 180.868 180.878 scf_env_do_scf 11 5.9 0.001 0.001 163.169 163.177 velocity_verlet 10 3.0 0.002 0.002 124.145 124.150 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 117.147 117.150 dbcsr_multiply_generic 2485 12.5 0.186 0.191 80.642 81.458 qs_scf_new_mos 116 7.6 0.001 0.001 79.908 80.228 qs_scf_loop_do_ot 116 8.6 0.001 0.001 79.907 80.227 ot_scf_mini 116 9.6 0.003 0.003 75.281 75.613 multiply_cannon 2485 13.5 0.551 0.581 56.379 59.255 multiply_cannon_loop 2485 14.5 1.811 1.922 52.403 54.162 init_scf_loop 11 6.9 0.000 0.001 45.903 45.909 ot_mini 116 10.6 0.001 0.001 42.740 43.052 prepare_preconditioner 11 7.9 0.000 0.000 41.788 41.814 make_preconditioner 11 8.9 0.000 0.000 41.788 41.814 make_full_inverse_cholesky 11 9.9 0.011 0.020 35.387 40.433 multiply_cannon_multrec 29820 15.5 13.520 18.363 26.662 31.191 rebuild_ks_matrix 127 8.3 0.001 0.001 29.601 29.908 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.019 29.601 29.907 mp_waitall_1 146592 16.7 16.751 27.023 16.751 27.023 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.608 26.882 qs_ot_get_derivative 116 11.6 0.001 0.002 23.186 23.517 make_m2s 4970 13.5 0.095 0.099 19.888 21.006 make_images 4970 14.5 1.916 2.206 19.580 20.699 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 18.996 19.673 apply_single 127 13.6 0.001 0.001 18.996 19.672 ot_diis_step 116 11.6 0.017 0.020 19.438 19.439 qs_ot_get_p 127 10.4 0.002 0.002 18.851 19.183 cp_fm_upper_to_full 104 14.8 11.375 16.811 11.375 16.811 cp_fm_cholesky_invert 11 10.9 15.873 15.882 15.873 15.882 multiply_cannon_metrocomm3 29820 15.5 0.050 0.054 6.327 15.078 qs_ot_p2m_diag 82 11.4 0.339 0.386 14.692 14.742 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.673 14.714 calculate_rho_elec 127 8.7 0.172 0.188 14.672 14.714 sum_up_and_integrate 127 10.3 0.139 0.150 14.556 14.583 integrate_v_rspace 127 11.3 0.004 0.004 14.417 14.448 cp_dbcsr_syevd 82 12.4 0.005 0.005 13.312 13.314 dbcsr_mm_accdrv_process 61748 16.2 8.781 9.521 12.713 13.251 dbcsr_complete_redistribute 393 12.7 1.497 1.610 9.076 12.835 make_images_data 4970 15.5 0.064 0.071 10.608 12.476 init_scf_run 11 5.9 0.000 0.001 11.774 11.775 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.774 11.775 multiply_cannon_sync_h2d 29820 15.5 10.383 11.669 10.383 11.669 copy_fm_to_dbcsr 208 11.6 0.001 0.002 7.731 11.523 hybrid_alltoall_any 5155 16.4 0.526 2.195 9.594 11.183 cp_fm_diag_elpa 82 13.4 0.000 0.000 10.157 10.174 cp_fm_diag_elpa_base 82 14.4 9.210 9.529 10.150 10.166 transfer_fm_to_dbcsr 11 9.9 0.001 0.006 6.379 10.097 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.794 10.036 mp_alltoall_i22 712 14.1 5.603 9.336 5.603 9.336 pw_transfer 1535 11.6 0.085 0.101 9.028 9.105 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.804 8.886 density_rs2pw 127 9.7 0.006 0.006 6.898 8.507 fft_wrap_pw1pw2_140 519 13.2 1.544 1.581 7.854 7.953 grid_integrate_task_list 127 12.3 7.445 7.919 7.445 7.919 wfi_extrapolate 11 7.9 0.001 0.001 7.657 7.657 cp_fm_cholesky_decompose 22 10.9 7.368 7.455 7.368 7.455 calculate_dm_sparse 127 9.5 0.001 0.001 7.000 7.095 multiply_cannon_metrocomm4 24850 15.5 0.084 0.096 2.805 7.044 mp_irecv_dv 75445 16.2 2.649 6.766 2.649 6.766 rs_pw_transfer 1038 11.9 0.013 0.014 4.567 6.258 mp_alltoall_d11v 2401 14.1 5.261 6.149 5.261 6.149 grid_collocate_task_list 127 9.7 5.101 5.842 5.101 5.842 fft3d_ps 1281 14.7 2.822 2.907 5.759 5.820 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.462 5.510 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.530 4.623 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.405 4.507 potential_pw2rs 127 12.3 0.023 0.023 4.446 4.461 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 4.423 4.423 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.222 4.276 calculate_first_density_matrix 1 7.0 0.000 0.000 3.906 3.913 mp_waitany 14840 13.9 2.203 3.868 2.203 3.868 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=188.675000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1105.909091, yerr=14.859674 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/72113d53c0cb11844f79f3c088677e10c62b6b2a_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456039424 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613072052224 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176077312 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176077312 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.233020E+12 0.0% 0.0% 100.0% flops max/rank 5.910120E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383904 0.0% 0.0% 100.0% number of processed stacks 1976928 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.9 marketing flops 145.650931E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 1.532510E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 242784 MPI messages size (bytes): total size 1.341806E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.526748E+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 115008 60297314304 4194304 < size <= 16777216 105840 554906419200 16777216 < size 20352 726592466352 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 9010 51. MP_Alltoall 9724 794507. MP_ISend 40420 2100460. MP_IRecv 40420 2099564. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4039 57681. MP_Allreduce 11174 1163. MP_Sync 88 MP_Alltoall 1724 18848034. MP_SendRecv 3870 122880. MP_ISendRecv 3870 122880. MP_Wait 16244 MP_ISend 10760 423501. MP_IRecv 10760 423501. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.058 173.504 173.520 qs_mol_dyn_low 1 2.0 0.003 0.003 172.706 172.748 qs_forces 11 3.9 0.004 0.004 172.590 172.593 qs_energies 11 4.9 0.002 0.002 165.021 165.027 scf_env_do_scf 11 5.9 0.001 0.001 146.788 146.791 velocity_verlet 10 3.0 0.027 0.029 113.666 113.671 scf_env_do_scf_inner_loop 118 6.6 0.003 0.009 110.963 110.964 dbcsr_multiply_generic 2529 12.6 0.198 0.210 72.966 73.396 qs_scf_new_mos 118 7.6 0.001 0.001 72.143 72.253 qs_scf_loop_do_ot 118 8.6 0.001 0.001 72.142 72.253 ot_scf_mini 118 9.6 0.003 0.004 67.705 67.818 multiply_cannon 2529 13.6 0.567 0.586 54.404 57.682 multiply_cannon_loop 2529 14.6 0.815 0.850 51.247 52.296 ot_mini 118 10.6 0.001 0.001 37.953 38.059 init_scf_loop 11 6.9 0.001 0.001 35.681 35.682 prepare_preconditioner 11 7.9 0.000 0.000 31.568 31.588 make_preconditioner 11 8.9 0.000 0.000 31.568 31.588 mp_waitall_1 126876 16.7 24.867 31.498 24.867 31.498 rebuild_ks_matrix 129 8.3 0.001 0.001 30.008 30.136 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.018 30.007 30.135 make_full_inverse_cholesky 11 9.9 0.018 0.027 29.524 29.782 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.258 27.376 multiply_cannon_multrec 10116 15.6 10.419 15.905 18.360 22.764 qs_ot_get_derivative 118 11.6 0.001 0.002 20.536 20.643 multiply_cannon_metrocomm3 10116 15.6 0.026 0.027 13.101 20.433 cp_fm_cholesky_invert 11 10.9 18.059 18.065 18.059 18.065 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 17.448 17.662 apply_single 129 13.6 0.001 0.001 17.447 17.661 ot_diis_step 118 11.6 0.020 0.020 17.344 17.345 qs_ot_get_p 129 10.4 0.230 0.230 17.182 17.322 make_m2s 5058 13.6 0.067 0.071 14.698 15.707 qs_rho_update_rho_low 129 7.7 0.001 0.001 15.483 15.513 calculate_rho_elec 129 8.7 0.258 0.269 15.483 15.512 make_images 5058 14.6 2.182 2.623 14.388 15.391 sum_up_and_integrate 129 10.3 0.181 0.190 15.050 15.097 integrate_v_rspace 129 11.3 0.004 0.004 14.869 14.925 qs_ot_p2m_diag 84 11.4 0.502 0.508 13.222 13.238 cp_dbcsr_syevd 84 12.4 0.005 0.006 12.055 12.056 multiply_cannon_sync_h2d 10116 15.6 10.823 11.205 10.823 11.205 init_scf_run 11 5.9 0.000 0.001 11.190 11.190 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.190 11.190 make_images_data 5058 15.6 0.056 0.066 8.417 9.892 hybrid_alltoall_any 5245 16.5 0.849 3.803 8.264 9.536 pw_transfer 1559 11.6 0.086 0.095 9.490 9.509 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 9.265 9.290 cp_fm_diag_elpa 84 13.4 0.000 0.000 9.174 9.186 cp_fm_diag_elpa_base 84 14.4 8.939 9.014 9.171 9.183 grid_integrate_task_list 129 12.3 7.812 8.300 7.812 8.300 dbcsr_mm_accdrv_process 20934 16.1 3.379 4.355 7.588 8.283 fft_wrap_pw1pw2_140 527 13.2 1.927 1.962 8.211 8.230 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 8.083 8.170 cp_fm_cholesky_decompose 22 10.9 7.857 7.990 7.857 7.990 density_rs2pw 129 9.7 0.006 0.006 6.897 7.841 wfi_extrapolate 11 7.9 0.001 0.001 7.428 7.428 multiply_cannon_metrocomm1 10116 15.6 0.030 0.031 4.155 7.093 calculate_dm_sparse 129 9.5 0.001 0.001 6.291 6.372 grid_collocate_task_list 129 9.7 5.521 6.203 5.521 6.203 fft3d_ps 1301 14.7 2.781 2.860 5.717 5.757 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.544 5.562 mp_alltoall_d11v 2429 14.1 4.840 5.536 4.840 5.536 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.505 5.505 dbcsr_complete_redistribute 397 12.7 2.231 2.278 5.160 5.459 multiply_cannon_metrocomm4 7587 15.6 0.027 0.030 1.877 5.419 mp_irecv_dv 29102 15.9 1.836 5.327 1.836 5.327 rs_pw_transfer 1054 12.0 0.012 0.013 4.090 5.049 mp_allgather_i34 2529 14.6 1.319 4.584 1.319 4.584 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.355 4.392 potential_pw2rs 129 12.3 0.026 0.026 4.375 4.389 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.584 3.863 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.668 3.720 copy_fm_to_dbcsr 210 11.7 0.001 0.002 3.388 3.697 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.655 3.670 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.573 3.662 calculate_first_density_matrix 1 7.0 0.000 0.000 3.499 3.500 copy_dbcsr_to_fm 187 11.8 0.004 0.004 3.397 3.495 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=173.520000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1428.545455, yerr=49.415426 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/72113d53c0cb11844f79f3c088677e10c62b6b2a_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 3.185680E+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 4057 58363. MP_Allreduce 11219 1499. 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 0.023 0.053 296.574 296.578 qs_mol_dyn_low 1 2.0 0.003 0.004 295.909 296.033 qs_forces 11 3.9 0.005 0.005 295.799 295.803 qs_energies 11 4.9 0.002 0.002 286.368 286.380 scf_env_do_scf 11 5.9 0.001 0.001 263.010 263.013 velocity_verlet 10 3.0 0.002 0.002 213.653 213.662 scf_env_do_scf_inner_loop 118 6.6 0.004 0.008 139.229 139.231 init_scf_loop 11 6.9 0.001 0.001 123.508 123.511 prepare_preconditioner 11 7.9 0.000 0.000 118.242 118.262 make_preconditioner 11 8.9 0.000 0.000 118.242 118.262 make_full_inverse_cholesky 11 9.9 0.038 0.039 94.192 115.411 qs_scf_new_mos 118 7.6 0.001 0.001 89.360 89.417 qs_scf_loop_do_ot 118 8.6 0.001 0.001 89.359 89.416 ot_scf_mini 118 9.6 0.004 0.004 84.479 84.485 dbcsr_multiply_generic 2535 12.6 0.213 0.225 82.646 83.072 cp_fm_upper_to_full 106 14.8 52.326 75.403 52.326 75.403 multiply_cannon 2535 13.6 0.687 0.728 59.542 61.049 multiply_cannon_loop 2535 14.6 1.036 1.064 55.554 56.702 ot_mini 118 10.6 0.001 0.001 44.084 44.093 dbcsr_complete_redistribute 397 12.7 3.994 4.040 29.905 42.823 copy_fm_to_dbcsr 210 11.7 0.001 0.002 26.418 39.274 rebuild_ks_matrix 129 8.3 0.001 0.001 37.733 37.745 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.018 37.733 37.744 transfer_fm_to_dbcsr 11 9.9 0.030 0.030 24.007 36.770 qs_ks_update_qs_env 129 7.6 0.001 0.001 34.691 34.713 mp_alltoall_i22 720 14.1 21.776 34.481 21.776 34.481 cp_fm_cholesky_invert 11 10.9 32.896 32.903 32.896 32.903 mp_waitall_1 104820 16.8 27.996 32.370 27.996 32.370 qs_ot_get_p 129 10.4 0.002 0.002 25.215 25.254 qs_ot_get_derivative 118 11.6 0.002 0.002 24.622 24.628 qs_ot_p2m_diag 84 11.4 0.891 0.896 21.192 21.222 qs_rho_update_rho_low 129 7.7 0.001 0.001 20.515 20.524 calculate_rho_elec 129 8.7 0.483 0.484 20.514 20.523 make_m2s 5070 13.6 0.075 0.079 18.597 20.052 multiply_cannon_metrocomm3 10140 15.6 0.025 0.028 18.831 20.013 sum_up_and_integrate 129 10.3 0.324 0.326 19.493 19.579 make_images 5070 14.6 3.077 3.263 18.112 19.568 ot_diis_step 118 11.6 0.022 0.022 19.439 19.439 cp_dbcsr_syevd 84 12.4 0.006 0.006 19.391 19.392 integrate_v_rspace 129 11.3 0.004 0.004 19.169 19.255 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 18.745 18.817 apply_single 129 13.6 0.001 0.001 18.744 18.817 multiply_cannon_multrec 10140 15.6 10.372 12.319 18.182 18.248 cp_fm_diag_elpa 84 13.4 0.000 0.000 16.226 16.227 cp_fm_diag_elpa_base 84 14.4 11.834 13.442 16.223 16.223 multiply_cannon_sync_h2d 10140 15.6 14.448 14.470 14.448 14.470 hybrid_alltoall_any 5257 16.5 1.325 3.079 10.739 13.031 init_scf_run 11 5.9 0.000 0.001 12.884 12.884 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.883 12.884 make_images_data 5070 15.6 0.064 0.069 10.479 12.607 pw_transfer 1559 11.6 0.094 0.095 11.441 11.447 fft_wrap_pw1pw2 1301 12.7 0.012 0.012 11.205 11.210 fft_wrap_pw1pw2_140 527 13.2 3.128 3.171 9.989 9.999 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.745 9.748 dbcsr_mm_accdrv_process 20958 16.1 4.273 6.064 7.558 9.459 wfi_extrapolate 11 7.9 0.001 0.001 9.144 9.144 grid_integrate_task_list 129 12.3 8.637 8.834 8.637 8.834 cp_fm_cholesky_decompose 22 10.9 8.770 8.790 8.770 8.790 mp_alltoall_d11v 2429 14.1 7.798 8.455 7.798 8.455 qs_energies_init_hamiltonians 11 5.9 0.191 0.203 8.345 8.346 density_rs2pw 129 9.7 0.005 0.005 7.925 8.087 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.877 6.903 calculate_dm_sparse 129 9.5 0.001 0.001 6.741 6.815 grid_collocate_task_list 129 9.7 6.479 6.550 6.479 6.550 rs_scatter_matrices 140 9.7 3.981 4.862 6.095 6.308 fft3d_ps 1301 14.7 2.870 2.881 6.213 6.248 copy_dbcsr_to_fm 187 11.8 0.004 0.004 6.102 6.166 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=296.578000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2900.090909, yerr=157.381098 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/72113d53c0cb11844f79f3c088677e10c62b6b2a_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 1.259495E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255646. MP_Allreduce 3139 6114. MP_Sync 4 MP_Alltoall 54 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 0.034 84.971 84.972 qs_energies 1 2.0 0.000 0.000 84.437 84.463 ls_scf 1 3.0 0.000 0.000 83.537 83.564 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.535 72.740 multiply_cannon 111 7.7 0.017 0.020 56.044 57.202 multiply_cannon_loop 111 8.7 0.227 0.239 52.610 53.874 ls_scf_main 1 4.0 0.000 0.000 52.271 52.271 density_matrix_trs4 2 5.0 0.002 0.003 46.734 46.819 ls_scf_init_scf 1 4.0 0.000 0.000 28.202 28.203 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.058 27.117 mp_waitall_1 11031 10.9 22.564 25.478 22.564 25.478 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 24.952 24.972 multiply_cannon_multrec 2664 9.7 8.140 9.023 15.526 17.331 multiply_cannon_sync_h2d 2664 9.7 13.497 15.805 13.497 15.805 make_m2s 222 7.7 0.009 0.011 12.988 13.583 make_images 222 8.7 0.100 0.110 12.965 13.561 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 9.838 12.458 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.502 8.715 make_images_data 222 9.7 0.004 0.005 7.560 8.095 dbcsr_mm_accdrv_process 4760 10.4 0.591 0.705 7.001 7.973 hybrid_alltoall_any 227 10.6 0.215 1.811 6.502 7.853 dbcsr_mm_accdrv_process_sort 4760 11.4 6.212 7.095 6.212 7.095 calculate_norms 4752 9.8 5.489 6.089 5.489 6.089 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.024 5.148 mp_sum_l 887 5.1 2.963 4.322 2.963 4.322 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.045 3.653 mp_irecv_dv 6231 10.9 2.029 3.625 2.029 3.625 make_images_sizes 222 9.7 0.000 0.000 0.660 3.235 mp_alltoall_i44 222 10.7 0.660 3.235 0.660 3.235 arnoldi_extremal 4 6.8 0.000 0.000 3.173 3.205 arnoldi_normal_ev 4 7.8 0.001 0.003 3.173 3.204 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.235 3.155 ls_scf_post 1 4.0 0.000 0.000 3.064 3.091 build_subspace 16 8.4 0.009 0.012 3.082 3.085 ls_scf_store_result 1 5.0 0.000 0.000 2.861 2.902 dbcsr_special_finalize 555 9.7 0.005 0.006 2.412 2.826 dbcsr_merge_single_wm 555 10.7 0.463 0.587 2.403 2.818 make_images_pack 222 9.7 2.208 2.632 2.209 2.633 dbcsr_sort_data 658 11.4 2.198 2.551 2.198 2.551 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.310 2.545 dbcsr_matrix_vector_mult_local 304 10.0 2.069 2.455 2.071 2.457 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.311 2.391 buffer_matrices_ensure_size 222 8.7 1.745 2.079 1.745 2.079 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.769 1.770 rebuild_ks_matrix 3 7.3 0.000 0.000 1.759 1.761 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.759 1.761 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=84.972000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1140.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/72113d53c0cb11844f79f3c088677e10c62b6b2a_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 2.191651E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266673. MP_Allreduce 3138 10075. MP_Sync 4 MP_Alltoall 47 15335933. MP_SendRecv 141 57600. MP_ISendRecv 141 57600. MP_Wait 687 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.040 0.089 91.289 91.290 qs_energies 1 2.0 0.000 0.000 90.541 90.569 ls_scf 1 3.0 0.000 0.001 88.671 88.698 dbcsr_multiply_generic 111 6.7 0.016 0.019 74.219 74.564 multiply_cannon 111 7.7 0.027 0.034 52.528 56.306 ls_scf_main 1 4.0 0.000 0.005 54.388 54.389 multiply_cannon_loop 111 8.7 0.136 0.147 49.757 53.196 density_matrix_trs4 2 5.0 0.002 0.005 48.673 48.880 ls_scf_init_scf 1 4.0 0.000 0.002 30.480 30.482 mp_waitall_1 9105 10.9 20.614 29.769 20.614 29.769 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.066 29.176 multiply_cannon_multrec 1332 9.7 13.352 17.645 22.650 28.101 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 26.350 26.363 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.307 20.651 make_m2s 222 7.7 0.007 0.008 15.328 15.934 make_images 222 8.7 1.368 1.695 15.297 15.904 dbcsr_mm_accdrv_process 4041 10.4 0.337 0.506 8.892 10.437 make_images_data 222 9.7 0.004 0.005 8.803 9.938 dbcsr_mm_accdrv_process_sort 4041 11.4 8.412 9.911 8.412 9.911 hybrid_alltoall_any 227 10.6 0.542 2.589 8.183 9.442 mp_sum_l 887 5.1 4.915 7.994 4.915 7.994 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.184 7.907 mp_irecv_dv 3311 11.0 3.164 7.851 3.164 7.851 calculate_norms 2376 9.8 6.079 6.803 6.079 6.803 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.661 6.386 multiply_cannon_sync_h2d 1332 9.7 4.867 6.371 4.867 6.371 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.269 5.519 arnoldi_extremal 4 6.8 0.000 0.002 4.876 4.901 arnoldi_normal_ev 4 7.8 0.004 0.023 4.876 4.900 build_subspace 16 8.4 0.014 0.021 4.379 4.382 ls_scf_post 1 4.0 0.004 0.030 3.803 3.832 ls_scf_store_result 1 5.0 0.000 0.000 3.409 3.514 dbcsr_matrix_vector_mult 304 9.0 0.010 0.023 3.199 3.445 mp_allgather_i34 111 8.7 0.765 3.234 0.765 3.234 dbcsr_matrix_vector_mult_local 304 10.0 2.744 3.227 2.746 3.229 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.154 2.846 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.651 2.768 make_images_pack 222 9.7 2.040 2.434 2.043 2.436 dbcsr_data_new 4174 10.1 2.111 2.412 2.111 2.412 dbcsr_sort_data 436 11.2 1.821 2.104 1.821 2.104 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.864 1.866 qs_energies_init_hamiltonians 1 3.0 0.000 0.004 1.865 1.866 rebuild_ks_matrix 3 7.3 0.000 0.000 1.851 1.853 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.851 1.853 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.290000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1779.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/72113d53c0cb11844f79f3c088677e10c62b6b2a_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 2.815140E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265448. MP_Allreduce 3138 10896. MP_Sync 4 MP_Alltoall 47 23526250. MP_SendRecv 93 57600. MP_ISendRecv 93 57600. MP_Wait 639 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.036 0.049 93.423 93.424 qs_energies 1 2.0 0.000 0.000 92.370 92.701 ls_scf 1 3.0 0.000 0.002 90.662 90.994 dbcsr_multiply_generic 111 6.7 0.016 0.019 75.027 75.371 multiply_cannon 111 7.7 0.039 0.131 52.146 56.374 ls_scf_main 1 4.0 0.001 0.019 56.005 56.010 multiply_cannon_loop 111 8.7 0.117 0.130 49.412 53.816 density_matrix_trs4 2 5.0 0.002 0.005 50.057 50.258 mp_waitall_1 7281 11.0 23.090 33.766 23.090 33.766 ls_scf_init_scf 1 4.0 0.000 0.002 30.843 30.847 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.552 29.637 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 27.180 27.190 multiply_cannon_multrec 888 9.7 12.767 15.624 21.352 24.874 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.736 23.385 make_m2s 222 7.7 0.006 0.008 16.065 16.805 make_images 222 8.7 1.580 1.847 16.027 16.764 make_images_data 222 9.7 0.004 0.004 9.373 10.351 hybrid_alltoall_any 227 10.6 0.640 2.936 8.978 10.057 dbcsr_mm_accdrv_process 3754 10.4 0.313 0.498 8.114 9.337 mp_sum_l 887 5.1 5.132 8.933 5.132 8.933 dbcsr_mm_accdrv_process_sort 3754 11.4 7.676 8.839 7.676 8.839 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.624 7.142 multiply_cannon_sync_h2d 888 9.7 6.027 7.128 6.027 7.128 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.472 7.032 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.955 7.005 mp_irecv_dv 2335 11.1 2.456 6.972 2.456 6.972 arnoldi_extremal 4 6.8 0.000 0.000 5.140 5.163 arnoldi_normal_ev 4 7.8 0.001 0.005 5.140 5.163 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.774 5.116 build_subspace 16 8.4 0.014 0.020 4.841 4.849 calculate_norms 1584 9.8 4.340 4.743 4.340 4.743 ls_scf_post 1 4.0 0.053 0.423 3.813 4.146 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.426 3.760 mp_allgather_i34 111 8.7 0.854 3.638 0.854 3.638 dbcsr_matrix_vector_mult_local 304 10.0 3.026 3.605 3.028 3.607 ls_scf_store_result 1 5.0 0.000 0.000 3.444 3.525 ls_scf_dm_to_ks 2 5.0 0.001 0.008 2.897 3.017 dbcsr_data_new 4116 9.9 2.096 2.464 2.096 2.464 dbcsr_sort_data 325 11.1 1.865 2.061 1.865 2.061 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.979 1.981 make_images_sizes 222 9.7 0.000 0.000 0.868 1.970 mp_alltoall_i44 222 10.7 0.868 1.970 0.868 1.970 rebuild_ks_matrix 3 7.3 0.000 0.000 1.961 1.963 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.961 1.963 make_images_pack 222 9.7 1.624 1.875 1.626 1.878 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=93.424000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2233.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/72113d53c0cb11844f79f3c088677e10c62b6b2a_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 3.382510E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266673. MP_Allreduce 3138 13030. MP_Sync 4 MP_Alltoall 47 30278988. MP_SendRecv 69 86400. MP_ISendRecv 69 86400. MP_Wait 531 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.036 0.053 97.952 97.953 qs_energies 1 2.0 0.000 0.000 97.273 97.278 ls_scf 1 3.0 0.044 0.355 95.402 95.408 dbcsr_multiply_generic 111 6.7 0.017 0.021 78.400 78.637 ls_scf_main 1 4.0 0.000 0.004 59.118 59.163 multiply_cannon 111 7.7 0.054 0.105 51.760 55.873 density_matrix_trs4 2 5.0 0.002 0.009 52.685 52.811 multiply_cannon_loop 111 8.7 0.152 0.167 46.757 49.676 ls_scf_init_scf 1 4.0 0.001 0.005 33.007 33.009 ls_scf_init_matrix_S 1 5.0 0.000 0.001 31.770 31.860 mp_waitall_1 6369 11.0 22.830 30.051 22.830 30.051 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 29.208 29.220 multiply_cannon_multrec 1332 9.7 14.133 16.987 22.053 25.458 make_m2s 222 7.7 0.007 0.009 21.171 22.550 make_images 222 8.7 3.134 3.597 21.120 22.501 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.286 16.550 make_images_data 222 9.7 0.004 0.004 11.848 13.515 hybrid_alltoall_any 227 10.6 0.796 3.758 11.190 12.847 dbcsr_mm_accdrv_process 3641 10.4 0.287 0.488 7.563 9.080 dbcsr_mm_accdrv_process_sort 3641 11.4 7.121 8.591 7.121 8.591 mp_sum_l 887 5.1 3.890 6.996 3.890 6.996 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.105 6.059 mp_irecv_dv 3229 10.9 2.079 5.978 2.079 5.978 multiply_cannon_sync_h2d 1332 9.7 5.483 5.938 5.483 5.938 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.021 5.476 arnoldi_extremal 4 6.8 0.000 0.000 5.158 5.172 arnoldi_normal_ev 4 7.8 0.001 0.005 5.157 5.172 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.526 4.963 build_subspace 16 8.4 0.014 0.021 4.826 4.833 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.535 4.681 calculate_norms 2376 9.8 4.178 4.563 4.178 4.563 mp_allgather_i34 111 8.7 2.116 4.350 2.116 4.350 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.567 3.865 dbcsr_matrix_vector_mult_local 304 10.0 3.199 3.688 3.201 3.690 dbcsr_sort_data 658 11.4 3.059 3.470 3.059 3.470 ls_scf_post 1 4.0 0.000 0.000 3.232 3.238 dbcsr_special_finalize 555 9.7 0.006 0.007 2.801 3.195 dbcsr_merge_single_wm 555 10.7 0.533 0.654 2.792 3.187 ls_scf_dm_to_ks 2 5.0 0.000 0.008 3.033 3.090 ls_scf_store_result 1 5.0 0.000 0.000 2.935 2.996 dbcsr_data_release 10477 10.7 1.586 2.432 1.586 2.432 dbcsr_finalize 304 7.8 0.049 0.061 1.801 1.994 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.953000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2724.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/72113d53c0cb11844f79f3c088677e10c62b6b2a_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 4.772274E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265536. MP_Allreduce 3129 15263. MP_Sync 4 MP_Alltoall 47 46208988. MP_SendRecv 45 115200. MP_ISendRecv 45 115200. MP_Wait 528 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.050 0.229 94.039 94.039 qs_energies 1 2.0 0.000 0.000 93.046 93.050 ls_scf 1 3.0 0.000 0.000 90.825 90.830 dbcsr_multiply_generic 111 6.7 0.017 0.018 70.903 71.158 ls_scf_main 1 4.0 0.000 0.000 56.690 56.691 multiply_cannon 111 7.7 0.085 0.162 52.941 56.180 multiply_cannon_loop 111 8.7 0.090 0.110 50.351 51.660 density_matrix_trs4 2 5.0 0.002 0.003 49.708 49.767 ls_scf_init_scf 1 4.0 0.000 0.000 29.506 29.507 mp_waitall_1 5436 11.0 24.607 28.435 24.607 28.435 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.177 28.208 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.081 26.096 multiply_cannon_multrec 444 9.7 13.677 16.460 20.761 21.992 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.909 15.603 make_m2s 222 7.7 0.005 0.005 13.375 14.325 make_images 222 8.7 2.038 2.485 13.308 14.257 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.352 13.327 make_images_data 222 9.7 0.003 0.004 8.161 9.676 hybrid_alltoall_any 227 10.6 0.800 3.791 8.070 9.602 dbcsr_mm_accdrv_process 3003 10.4 0.355 0.426 6.780 7.904 multiply_cannon_sync_h2d 444 9.7 6.742 7.855 6.742 7.855 dbcsr_mm_accdrv_process_sort 3003 11.4 6.412 7.498 6.412 7.498 arnoldi_extremal 4 6.8 0.000 0.000 5.846 5.863 arnoldi_normal_ev 4 7.8 0.002 0.005 5.846 5.863 build_subspace 16 8.4 0.015 0.020 5.456 5.467 ls_scf_post 1 4.0 0.000 0.000 4.630 4.633 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.359 4.593 mp_sum_l 887 5.1 2.739 4.587 2.739 4.587 dbcsr_matrix_vector_mult 304 9.0 0.011 0.022 4.155 4.349 dbcsr_matrix_vector_mult_local 304 10.0 3.697 4.163 3.699 4.165 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.568 3.905 mp_irecv_dv 1241 11.2 1.554 3.870 1.554 3.870 mp_allgather_i34 111 8.7 1.172 3.736 1.172 3.736 calculate_norms 792 9.8 3.622 3.732 3.622 3.732 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.571 3.635 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.836 3.411 ls_scf_store_result 1 5.0 0.000 0.000 3.178 3.208 make_images_sizes 222 9.7 0.000 0.000 0.834 3.104 mp_alltoall_i44 222 10.7 0.833 3.103 0.833 3.103 dbcsr_data_new 4608 9.7 1.782 2.283 1.782 2.283 dbcsr_finalize 304 7.8 0.062 0.077 2.196 2.276 qs_energies_init_hamiltonians 1 3.0 0.000 0.001 2.205 2.205 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.131 2.132 dbcsr_merge_all 275 8.9 0.478 0.522 2.055 2.125 rebuild_ks_matrix 3 7.3 0.000 0.000 2.098 2.099 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.098 2.099 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=94.039000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3751.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/72113d53c0cb11844f79f3c088677e10c62b6b2a_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 8.935932E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284089. MP_Allreduce 3123 21388. MP_Sync 4 MP_Alltoall 47 88727262. MP_SendRecv 42 732600. MP_ISendRecv 42 732600. MP_Wait 267 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.074 0.090 106.010 106.010 qs_energies 1 2.0 0.000 0.000 104.606 104.612 ls_scf 1 3.0 0.000 0.000 101.337 101.342 dbcsr_multiply_generic 111 6.7 0.023 0.028 74.977 75.145 ls_scf_main 1 4.0 0.000 0.000 63.680 63.681 density_matrix_trs4 2 5.0 0.002 0.003 54.657 54.732 multiply_cannon 111 7.7 0.202 0.272 48.708 50.691 multiply_cannon_loop 111 8.7 0.098 0.100 45.730 46.391 ls_scf_init_scf 1 4.0 0.000 0.000 33.867 33.868 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.286 32.304 matrix_sqrt_Newton_Schulz 2 6.5 0.002 0.007 29.584 29.594 mp_waitall_1 4527 11.1 21.911 25.795 21.911 25.795 make_m2s 222 7.7 0.005 0.005 22.542 23.625 make_images 222 8.7 3.573 3.871 22.434 23.517 multiply_cannon_multrec 444 9.7 17.826 18.442 22.626 23.419 hybrid_alltoall_any 227 10.6 1.657 3.627 12.669 15.642 make_images_data 222 9.7 0.003 0.004 12.940 15.092 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.307 11.244 multiply_cannon_sync_h2d 444 9.7 8.791 8.842 8.791 8.842 arnoldi_extremal 4 6.8 0.000 0.000 7.359 7.374 arnoldi_normal_ev 4 7.8 0.004 0.010 7.359 7.373 build_subspace 16 8.4 0.026 0.036 6.805 6.814 dbcsr_matrix_vector_mult 304 9.0 0.017 0.035 5.452 5.607 dbcsr_matrix_vector_mult_local 304 10.0 5.053 5.345 5.056 5.348 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.146 5.238 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.849 5.109 dbcsr_mm_accdrv_process 1814 10.4 0.425 0.675 4.615 4.883 dbcsr_mm_accdrv_process_sort 1814 11.4 4.122 4.254 4.122 4.254 ls_scf_post 1 4.0 0.000 0.000 3.789 3.795 make_images_sizes 222 9.7 0.000 0.000 1.471 3.663 mp_alltoall_i44 222 10.7 1.470 3.662 1.470 3.662 mp_allgather_i34 111 8.7 1.105 3.586 1.105 3.586 ls_scf_store_result 1 5.0 0.000 0.000 3.506 3.542 calculate_norms 792 9.8 3.229 3.273 3.229 3.273 qs_energies_init_hamiltonians 1 3.0 0.330 0.343 3.238 3.238 dbcsr_finalize 304 7.8 0.082 0.089 3.069 3.112 dbcsr_merge_all 275 8.9 0.884 0.914 2.854 2.889 dbcsr_complete_redistribute 5 7.6 1.423 1.459 2.726 2.831 matrix_ls_to_qs 2 6.0 0.000 0.000 2.397 2.509 dbcsr_sort_data 325 11.1 2.438 2.478 2.438 2.478 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.473 2.475 rebuild_ks_matrix 3 7.3 0.000 0.000 2.407 2.409 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.407 2.409 dbcsr_new_transposed 4 7.5 0.241 0.252 2.295 2.316 dbcsr_frobenius_norm 74 6.6 2.057 2.126 2.176 2.209 dbcsr_data_new 6591 9.6 1.776 2.209 1.776 2.209 dbcsr_add_d 103 6.2 0.000 0.000 2.125 2.200 dbcsr_add_anytype 103 7.2 0.859 0.891 2.124 2.200 dbcsr_data_release 12724 10.6 1.970 2.184 1.970 2.184 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=106.010000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6981.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/72113d53c0cb11844f79f3c088677e10c62b6b2a_performance_tests/27/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 7009386627072 0.0% 0.0% 100.0% flops 9 x 9 x 32 7335108845568 0.0% 0.0% 100.0% flops 9 x 22 x 32 9866241589248 0.0% 0.0% 100.0% flops 22 x 9 x 32 9884108906496 0.0% 0.0% 100.0% flops 22 x 22 x 32 13354440523776 0.0% 0.0% 100.0% flops 32 x 32 x 9 20607185977344 0.0% 0.0% 100.0% flops 32 x 32 x 22 25186560638976 0.0% 0.0% 100.0% flops 9 x 32 x 32 28458319085568 0.0% 0.0% 100.0% flops 22 x 32 x 32 34782389993472 0.0% 0.0% 100.0% flops 9 x 32 x 9 42881542373376 0.0% 0.0% 100.0% flops 22 x 32 x 9 55680402235392 0.0% 0.0% 100.0% flops 9 x 32 x 22 55680402235392 0.0% 0.0% 100.0% flops 22 x 32 x 22 72328573419520 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 383.054662E+12 0.0% 0.0% 100.0% flops max/rank 733.641090E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 26899403712 0.0% 0.0% 100.0% number of processed stacks 118860288 0.0% 0.0% 100.0% average stack size 0.0 0.0 226.3 marketing flops 780.439111E+12 ------------------------------------------------------------------------------- # multiplications 1445 max memory usage/rank 592.859136E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 102097920 MPI messages size (bytes): total size 37.227590E+12 min size 0.000000E+00 max size 4.551360E+06 average size 364.626312E+03 MPI breakdown and total messages size (bytes): size <= 128 731472 0 128 < size <= 8192 11922720 97670922240 8192 < size <= 32768 24718992 614677610496 32768 < size <= 131072 20000256 1970081366016 131072 < size <= 4194304 42515668 24886801223040 4194304 < size <= 16777216 2208812 9656099886720 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4640 78072. MP_Allreduce 13232 2081. MP_Sync 1064 MP_Alltoall 2588 947189645. 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 0.044 0.095 229.862 229.864 qs_mol_dyn_low 1 2.0 0.008 0.349 228.787 228.809 qs_forces 5 3.8 0.004 0.005 228.355 228.356 qs_energies 5 4.8 0.003 0.045 225.346 225.381 scf_env_do_scf 5 5.8 0.000 0.002 207.563 207.566 scf_env_do_scf_inner_loop 105 6.6 0.002 0.006 180.961 180.963 qs_scf_new_mos 105 7.6 0.000 0.001 140.880 141.089 qs_scf_loop_do_ot 105 8.6 0.001 0.001 140.879 141.089 dbcsr_multiply_generic 1445 12.2 0.127 0.134 136.439 136.844 ot_scf_mini 105 9.6 0.003 0.004 131.025 131.190 multiply_cannon 1445 13.2 0.274 0.286 117.493 119.391 multiply_cannon_loop 1445 14.2 2.849 2.995 115.877 117.004 velocity_verlet 4 3.0 0.001 0.009 106.406 106.406 ot_mini 105 10.6 0.001 0.001 60.238 60.351 multiply_cannon_multrec 69360 15.2 29.754 34.960 39.749 44.714 mp_waitall_1 488190 16.1 34.494 41.813 34.494 41.813 qs_ot_get_p 112 10.4 0.001 0.001 40.979 41.244 qs_ot_get_derivative 55 11.6 0.001 0.001 38.568 38.700 multiply_cannon_sync_h2d 69360 15.2 29.235 33.794 29.235 33.794 multiply_cannon_metrocomm3 69360 15.2 0.203 0.214 25.759 32.993 qs_ot_p2m_diag 40 11.0 0.020 0.030 30.037 30.116 rebuild_ks_matrix 110 8.4 0.000 0.000 28.776 28.982 qs_ks_build_kohn_sham_matrix 110 9.4 0.011 0.013 28.775 28.982 cp_dbcsr_syevd 40 12.0 0.002 0.002 26.796 26.797 init_scf_loop 7 6.6 0.000 0.004 26.572 26.573 qs_ks_update_qs_env 112 7.6 0.001 0.001 26.374 26.558 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 22.981 23.302 apply_single 62 13.6 0.000 0.000 22.981 23.302 cp_fm_syevd 40 13.0 0.000 0.001 21.739 21.929 prepare_preconditioner 7 7.6 0.000 0.000 21.746 21.782 make_preconditioner 7 8.6 0.000 0.002 21.746 21.782 ot_new_cg_direction 55 11.6 0.001 0.001 20.994 20.994 qs_rho_update_rho_low 110 7.6 0.001 0.001 17.197 17.616 calculate_rho_elec 110 8.6 0.030 0.034 17.196 17.615 cp_fm_redistribute_end 40 14.0 8.563 17.089 8.567 17.090 cp_fm_syevd_base 40 14.0 8.516 17.044 8.516 17.044 init_scf_run 5 5.8 0.000 0.000 15.164 15.165 scf_env_initial_rho_setup 5 6.8 0.002 0.003 15.163 15.164 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.658 14.860 make_full_inverse_cholesky 7 9.6 0.000 0.000 14.536 14.599 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 13.896 14.010 mp_sum_l 4764 12.2 12.411 13.227 12.411 13.227 rs_pw_transfer 690 11.5 0.011 0.013 11.358 12.508 density_rs2pw 110 9.6 0.006 0.008 11.269 12.407 calculate_dm_sparse 110 9.5 0.000 0.001 11.855 12.093 pw_transfer 1645 12.4 0.082 0.099 11.816 11.989 fft_wrap_pw1pw2 1425 13.5 0.013 0.016 11.674 11.851 dbcsr_mm_accdrv_process 154766 15.8 6.180 6.354 9.862 10.961 acc_transpose_blocks 69360 15.2 0.359 0.377 10.561 10.895 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 10.362 10.452 calculate_first_density_matrix 1 7.0 0.000 0.003 10.308 10.329 qs_vxc_create 110 10.4 0.002 0.004 10.267 10.299 cp_fm_cholesky_invert 7 10.6 10.272 10.280 10.272 10.280 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.002 10.251 10.267 fft_wrap_pw1pw2_240 915 15.0 1.198 1.275 9.900 10.097 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 9.834 9.898 check_diag 80 13.5 8.599 8.919 9.350 9.500 sum_up_and_integrate 60 10.3 0.028 0.031 8.552 8.563 integrate_v_rspace 60 11.3 0.001 0.002 8.524 8.536 fft3d_pb 915 16.0 2.392 2.561 7.946 8.102 acc_transpose_blocks_kernels 69360 16.2 0.843 0.892 7.791 8.011 xc_rho_set_and_dset_create 110 12.4 0.077 0.098 7.231 7.473 jit_kernel_transpose 5 15.0 6.949 7.138 6.949 7.138 multiply_cannon_metrocomm1 69360 15.2 0.096 0.102 4.618 7.070 make_full_single_inverse 7 9.6 0.001 0.005 6.953 6.989 xc_vxc_pw_create 60 11.3 0.039 0.051 6.870 6.901 make_m2s 2890 13.2 0.080 0.090 6.072 6.679 make_images 2890 14.2 0.239 0.259 5.963 6.568 xc_pw_derive 510 13.4 0.006 0.007 6.054 6.137 mp_waitany 7680 13.5 4.294 5.513 4.294 5.513 mp_alltoall_z22v 2340 17.7 4.982 5.296 4.982 5.296 potential_pw2rs 60 12.3 0.003 0.003 4.764 4.780 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=229.864000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=562.000000, yerr=3.633180 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 72113d53c0cb11844f79f3c088677e10c62b6b2a Summary: empty Status: OK