=== 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: e4289968d5d1c81be94a335915a0cf308a6045a0 ################# 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 2023.05.001, # HDF5 1.12.0, 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: 28.08.2023 # # \ 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=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2023.05.001 USE_HDF5 := 1.12.0 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.2 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(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)) 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__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 += $(LIBVDWXC_LIB)/libvdwxc.a endif ifneq ($(USE_HDF5),) USE_HDF5 := $(strip $(USE_HDF5)) HDF5_INC := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/include HDF5_LIB := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/lib CFLAGS += -I$(HDF5_INC) DFLAGS += -D__HDF5 LIBS += $(HDF5_LIB)/libhdf5_fortran.a LIBS += $(HDF5_LIB)/libhdf5_hl.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/e4289968d5d1c81be94a335915a0cf308a6045a0_performance_tests/01 job id: 48830666 --- 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/e4289968d5d1c81be94a335915a0cf308a6045a0_performance_tests/02 job id: 48830667 --- 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/e4289968d5d1c81be94a335915a0cf308a6045a0_performance_tests/03 job id: 48830668 --- 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/e4289968d5d1c81be94a335915a0cf308a6045a0_performance_tests/04 job id: 48830669 --- 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/e4289968d5d1c81be94a335915a0cf308a6045a0_performance_tests/05 job id: 48830670 --- 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/e4289968d5d1c81be94a335915a0cf308a6045a0_performance_tests/06 job id: 48830671 --- 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/e4289968d5d1c81be94a335915a0cf308a6045a0_performance_tests/07 job id: 48830672 --- 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/e4289968d5d1c81be94a335915a0cf308a6045a0_performance_tests/08 job id: 48830673 --- 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/e4289968d5d1c81be94a335915a0cf308a6045a0_performance_tests/09 job id: 48830674 --- 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/e4289968d5d1c81be94a335915a0cf308a6045a0_performance_tests/10 job id: 48830675 --- 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/e4289968d5d1c81be94a335915a0cf308a6045a0_performance_tests/11 job id: 48830677 --- 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/e4289968d5d1c81be94a335915a0cf308a6045a0_performance_tests/12 job id: 48830678 --- 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/e4289968d5d1c81be94a335915a0cf308a6045a0_performance_tests/13 job id: 48830679 --- 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/e4289968d5d1c81be94a335915a0cf308a6045a0_performance_tests/14 job id: 48830680 --- 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/e4289968d5d1c81be94a335915a0cf308a6045a0_performance_tests/15 job id: 48830682 --- 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/e4289968d5d1c81be94a335915a0cf308a6045a0_performance_tests/16 job id: 48830683 --- 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/e4289968d5d1c81be94a335915a0cf308a6045a0_performance_tests/17 job id: 48830684 --- 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/e4289968d5d1c81be94a335915a0cf308a6045a0_performance_tests/18 job id: 48830685 --- 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/e4289968d5d1c81be94a335915a0cf308a6045a0_performance_tests/19 job id: 48830686 --- 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/e4289968d5d1c81be94a335915a0cf308a6045a0_performance_tests/20 job id: 48830687 --- 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/e4289968d5d1c81be94a335915a0cf308a6045a0_performance_tests/21 job id: 48830688 --- 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/e4289968d5d1c81be94a335915a0cf308a6045a0_performance_tests/22 job id: 48830689 --- 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/e4289968d5d1c81be94a335915a0cf308a6045a0_performance_tests/23 job id: 48830690 --- 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/e4289968d5d1c81be94a335915a0cf308a6045a0_performance_tests/24 job id: 48830691 --- 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/e4289968d5d1c81be94a335915a0cf308a6045a0_performance_tests/25 job id: 48830692 --- 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/e4289968d5d1c81be94a335915a0cf308a6045a0_performance_tests/26 job id: 48830693 --- 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/e4289968d5d1c81be94a335915a0cf308a6045a0_performance_tests/27 job id: 48830694 --- 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/e4289968d5d1c81be94a335915a0cf308a6045a0_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 0.018 0.032 134.696 134.697 farming_run 1 2.0 134.161 134.163 134.663 134.666 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.502552E+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 0.007 0.028 115.107 115.108 qs_energies 1 2.0 0.000 0.000 114.912 114.916 mp2_main 1 3.0 0.000 0.000 112.492 112.497 mp2_gpw_main 1 4.0 0.028 0.034 111.375 111.380 mp2_ri_gpw_compute_in 1 5.0 0.172 0.175 92.164 92.575 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 54.683 55.095 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.165 41.087 46.184 get_2c_integrals 1 6.0 0.008 0.009 36.783 37.310 integrate_v_rspace 273 8.0 0.440 0.457 24.632 29.519 pw_transfer 6555 10.6 0.378 0.399 26.771 27.436 fft_wrap_pw1pw2 5465 11.4 0.045 0.048 25.412 26.154 grid_integrate_task_list 273 9.0 20.532 25.875 20.532 25.875 fft_wrap_pw1pw2_100 2178 12.4 1.104 1.305 22.957 23.689 rpa_ri_compute_en 1 5.0 0.020 0.021 19.093 19.333 compute_2c_integrals 1 7.0 0.003 0.004 18.942 18.944 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.537 18.640 mp2_eri_2c_integrate_gpw 1 9.0 2.401 2.437 18.534 18.638 cp_fm_cholesky_decompose 12 8.2 17.958 18.510 17.958 18.510 cholesky_decomp 1 7.0 0.000 0.000 16.684 17.230 fft3d_s 5443 13.4 16.135 16.583 16.157 16.605 ao_to_mo_and_store_B_mult_1 272 7.0 10.767 15.305 10.767 15.305 calculate_wavefunction 272 8.0 5.446 5.614 12.378 13.022 rpa_num_int 1 6.0 0.000 0.005 10.861 10.864 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.817 10.833 calc_mat_Q 8 8.0 0.000 0.000 9.499 9.605 contract_S_to_Q 8 9.0 0.000 0.000 8.919 9.026 calc_potential_gpw 544 9.5 0.005 0.005 8.192 8.674 parallel_gemm_fm 14 9.1 0.000 0.000 8.529 8.617 parallel_gemm_fm_cosma 14 10.1 8.529 8.617 8.529 8.617 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.034 8.267 potential_pw2rs 545 10.0 0.107 0.109 7.477 8.086 create_integ_mat 1 6.0 0.022 0.028 7.726 7.726 collocate_single_gaussian 272 10.0 0.039 0.041 7.295 7.545 array2fm 1 7.0 0.000 0.000 6.677 7.043 pw_scatter_s 2720 13.7 4.340 4.571 4.340 4.571 pw_gather_s 2722 13.2 3.426 3.830 3.426 3.830 array2fm_buffer_send 1 8.0 2.854 3.134 2.854 3.134 pw_poisson_solve 545 10.5 1.111 1.172 2.175 2.306 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.379802, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2819.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e4289968d5d1c81be94a335915a0cf308a6045a0_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 0.041 0.122 422.495 422.496 farming_run 1 2.0 421.409 421.467 422.392 422.405 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827128576 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788821 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.242837E+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 608231. MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.035 217.531 217.532 qs_energies 1 2.0 0.004 0.017 217.238 217.258 scf_env_do_scf 1 3.0 0.000 0.000 115.058 115.058 qs_ks_update_qs_env 5 5.0 0.000 0.000 114.002 114.008 rebuild_ks_matrix 4 6.0 0.000 0.000 114.001 114.007 qs_ks_build_kohn_sham_matrix 4 7.0 0.055 0.061 114.001 114.007 hfx_ks_matrix 4 8.0 0.001 0.001 113.630 113.635 integrate_four_center 4 9.0 0.153 0.470 113.630 113.634 integrate_four_center_main 4 10.0 0.133 0.663 101.781 105.185 integrate_four_center_bin 264 11.0 101.648 105.059 101.648 105.059 mp2_main 1 3.0 0.005 0.027 101.876 101.888 mp2_gpw_main 1 4.0 0.048 0.068 100.826 100.842 init_scf_loop 1 4.0 0.000 0.000 96.686 96.686 mp2_ri_gpw_compute_in 1 5.0 0.066 0.102 73.831 74.815 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.003 53.704 54.691 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.165 41.478 46.439 integrate_v_rspace 95 8.0 0.399 0.571 27.969 32.746 pw_transfer 2240 10.6 0.145 0.165 29.521 29.857 ao_to_mo_and_store_B_mult_1 91 7.0 10.558 28.876 10.558 28.876 fft_wrap_pw1pw2 1868 11.4 0.018 0.020 28.533 28.839 mp2_ri_gpw_compute_en 1 5.0 0.061 0.091 26.809 28.735 grid_integrate_task_list 95 9.0 23.372 28.341 23.372 28.341 fft_wrap_pw1pw2_100 730 12.4 1.283 1.411 26.250 26.587 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.844 1.886 25.151 25.161 get_2c_integrals 1 6.0 0.000 0.009 19.996 20.061 compute_2c_integrals 1 7.0 0.003 0.003 18.971 18.977 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 18.646 18.808 mp2_eri_2c_integrate_gpw 1 9.0 1.740 1.820 18.645 18.807 fft3d_s 1823 13.4 18.474 18.744 18.488 18.758 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 18.370 18.370 calculate_wavefunction 91 8.0 2.021 2.050 9.620 9.816 mp2_ri_gpw_compute_en_expansio 172 7.0 0.557 0.581 8.782 9.328 potential_pw2rs 186 10.0 0.033 0.034 8.405 8.993 local_gemm 172 8.0 8.225 8.762 8.225 8.762 mp2_ri_gpw_compute_en_comm 22 7.0 0.497 0.521 8.143 8.605 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.167 8.480 calc_potential_gpw 182 9.5 0.002 0.002 7.938 8.129 collocate_single_gaussian 91 10.0 0.017 0.026 7.754 7.905 mp_sync 37 10.5 3.833 7.060 3.833 7.060 integrate_four_center_load 4 10.0 0.000 0.000 6.776 6.780 hfx_load_balance 1 11.0 0.000 0.000 6.776 6.780 mp_sendrecv_dm3 2068 8.0 6.179 6.585 6.179 6.585 mp2_ri_gpw_compute_en_ener 172 7.0 6.336 6.403 6.336 6.403 pw_gather_s 912 13.2 4.461 4.907 4.461 4.907 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=100.837136, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1503.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e4289968d5d1c81be94a335915a0cf308a6045a0_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 452.644864E+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 MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.044 57.314 57.317 qs_mol_dyn_low 1 2.0 0.003 0.003 56.913 56.921 qs_forces 11 3.9 0.002 0.002 56.825 56.829 qs_energies 11 4.9 0.011 0.030 55.231 55.259 scf_env_do_scf 11 5.9 0.000 0.001 48.779 48.779 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 46.511 46.512 qs_scf_new_mos 108 7.5 0.000 0.001 35.768 36.086 qs_scf_loop_do_ot 108 8.5 0.000 0.001 35.767 36.086 dbcsr_multiply_generic 2286 12.5 0.095 0.100 35.082 35.554 ot_scf_mini 108 9.5 0.002 0.002 34.044 34.200 velocity_verlet 10 3.0 0.001 0.003 28.342 28.343 multiply_cannon 2286 13.5 0.188 0.197 26.829 28.294 multiply_cannon_loop 2286 14.5 1.823 1.946 25.964 27.516 ot_mini 108 10.5 0.001 0.001 20.537 20.785 qs_ot_get_derivative 108 11.5 0.001 0.001 17.400 17.587 mp_waitall_1 245248 16.5 9.224 15.190 9.224 15.190 multiply_cannon_metrocomm3 54864 15.5 0.075 0.081 6.283 13.011 multiply_cannon_multrec 54864 15.5 3.628 5.794 7.551 10.936 qs_ot_get_p 119 10.4 0.001 0.001 8.632 8.948 rebuild_ks_matrix 119 8.3 0.000 0.000 8.280 8.444 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.012 8.279 8.444 mp_sum_l 7287 12.8 5.891 7.622 5.891 7.622 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.284 7.432 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 6.406 6.849 multiply_cannon_sync_h2d 54864 15.5 5.086 6.308 5.086 6.308 qs_ot_p2m_diag 50 11.0 0.004 0.007 5.559 5.616 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.422 5.511 dbcsr_mm_accdrv_process 76910 16.1 1.825 2.873 3.837 5.489 init_scf_run 11 5.9 0.000 0.001 4.968 4.969 scf_env_initial_rho_setup 11 6.9 0.001 0.004 4.968 4.968 sum_up_and_integrate 119 10.3 0.001 0.002 4.772 4.778 integrate_v_rspace 119 11.3 0.002 0.003 4.760 4.768 cp_dbcsr_syevd 50 12.0 0.002 0.003 4.685 4.685 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.515 4.642 calculate_rho_elec 119 8.7 0.011 0.017 4.514 4.641 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.420 4.421 cp_fm_redistribute_end 50 14.0 2.257 4.388 2.267 4.393 cp_fm_diag_elpa_base 50 14.0 2.119 4.264 2.123 4.268 multiply_cannon_metrocomm1 54864 15.5 0.057 0.063 2.044 3.301 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.001 3.206 apply_single 119 13.6 0.000 0.000 3.001 3.206 calculate_dm_sparse 119 9.5 0.000 0.000 2.951 3.144 ot_diis_step 108 11.5 0.006 0.006 2.872 2.873 acc_transpose_blocks 54864 15.5 0.229 0.255 2.246 2.804 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.632 2.725 calculate_first_density_matrix 1 7.0 0.000 0.002 2.636 2.638 jit_kernel_multiply 13 15.8 1.948 2.588 1.948 2.588 density_rs2pw 119 9.7 0.004 0.004 2.453 2.562 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.434 2.436 wfi_extrapolate 11 7.9 0.001 0.001 2.254 2.254 init_scf_loop 11 6.9 0.000 0.001 2.249 2.250 make_m2s 4572 13.5 0.054 0.056 2.073 2.135 grid_integrate_task_list 119 12.3 2.032 2.134 2.032 2.134 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.043 2.108 mp_sum_d 4135 12.0 1.506 2.091 1.506 2.091 potential_pw2rs 119 12.3 0.004 0.004 2.055 2.075 make_images 4572 14.5 0.133 0.139 1.991 2.052 pw_transfer 1439 11.6 0.052 0.056 1.960 2.037 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.885 1.963 mp_alltoall_d11v 2130 13.8 1.612 1.771 1.612 1.771 transfer_rs2pw 487 10.6 0.005 0.006 1.561 1.706 fft3d_ps 1201 14.6 0.375 0.483 1.601 1.675 mp_waitany 12084 13.8 1.364 1.589 1.364 1.589 fft_wrap_pw1pw2_140 487 13.2 0.139 0.157 1.459 1.538 transfer_pw2rs 487 13.2 0.006 0.006 1.495 1.514 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.424 1.464 acc_transpose_blocks_sync 164592 16.5 1.207 1.449 1.207 1.449 grid_collocate_task_list 119 9.7 1.348 1.421 1.348 1.421 dbcsr_dot_sd 1205 11.9 0.048 0.060 0.858 1.255 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=57.317000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=431.000000, yerr=1.414214 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e4289968d5d1c81be94a335915a0cf308a6045a0_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 488.718336E+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 1031941. 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 0.023 0.054 41.824 41.827 qs_mol_dyn_low 1 2.0 0.003 0.004 41.452 41.482 qs_forces 11 3.9 0.015 0.033 41.382 41.383 qs_energies 11 4.9 0.004 0.009 39.647 39.656 scf_env_do_scf 11 5.9 0.001 0.002 34.231 34.232 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 31.473 31.475 dbcsr_multiply_generic 2286 12.5 0.101 0.104 22.999 23.469 qs_scf_new_mos 108 7.5 0.001 0.001 22.167 22.388 qs_scf_loop_do_ot 108 8.5 0.001 0.001 22.166 22.387 ot_scf_mini 108 9.5 0.003 0.005 21.205 21.367 velocity_verlet 10 3.0 0.003 0.004 19.866 19.867 multiply_cannon 2286 13.5 0.209 0.217 17.333 19.092 multiply_cannon_loop 2286 14.5 1.194 1.260 16.056 17.891 ot_mini 108 10.5 0.001 0.001 13.040 13.276 mp_waitall_1 200699 16.5 6.235 11.398 6.235 11.398 qs_ot_get_derivative 108 11.5 0.001 0.001 10.590 10.752 multiply_cannon_metrocomm3 27432 15.5 0.073 0.076 4.380 9.884 multiply_cannon_multrec 27432 15.5 1.808 4.252 6.124 8.907 rebuild_ks_matrix 119 8.3 0.000 0.000 7.140 7.306 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.015 7.139 7.306 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.293 6.443 dbcsr_mm_accdrv_process 47894 16.0 3.533 5.784 4.236 6.362 qs_ot_get_p 119 10.4 0.001 0.002 5.156 5.375 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 4.096 4.938 mp_sum_l 7287 12.8 2.305 4.451 2.305 4.451 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.072 4.117 calculate_rho_elec 119 8.7 0.021 0.024 4.071 4.117 sum_up_and_integrate 119 10.3 0.001 0.003 4.009 4.016 integrate_v_rspace 119 11.3 0.003 0.004 3.995 4.002 init_scf_run 11 5.9 0.000 0.001 3.990 3.991 scf_env_initial_rho_setup 11 6.9 0.003 0.006 3.990 3.991 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.022 3.708 apply_single 119 13.6 0.000 0.000 3.022 3.707 qs_ot_p2m_diag 50 11.0 0.009 0.014 3.315 3.341 make_m2s 4572 13.5 0.052 0.053 2.921 3.293 make_images 4572 14.5 0.206 0.244 2.832 3.208 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.888 2.889 init_scf_loop 11 6.9 0.001 0.004 2.735 2.740 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.528 2.615 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.477 2.479 density_rs2pw 119 9.7 0.004 0.004 2.343 2.473 cp_fm_redistribute_end 50 14.0 1.257 2.444 1.263 2.447 ot_diis_step 108 11.5 0.011 0.011 2.373 2.374 cp_fm_diag_elpa_base 50 14.0 1.147 2.330 1.177 2.366 calculate_first_density_matrix 1 7.0 0.001 0.004 2.361 2.365 multiply_cannon_sync_h2d 27432 15.5 1.660 2.191 1.660 2.191 calculate_dm_sparse 119 9.5 0.000 0.001 2.108 2.189 pw_transfer 1439 11.6 0.065 0.070 2.032 2.080 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.940 1.991 acc_transpose_blocks 27432 15.5 0.114 0.118 1.620 1.966 grid_integrate_task_list 119 12.3 1.835 1.950 1.835 1.950 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.911 1.914 prepare_preconditioner 11 7.9 0.000 0.000 1.751 1.777 make_preconditioner 11 8.9 0.001 0.003 1.751 1.777 make_images_data 4572 15.5 0.048 0.054 1.354 1.771 jit_kernel_multiply 9 16.2 0.645 1.766 0.645 1.766 potential_pw2rs 119 12.3 0.006 0.007 1.730 1.745 transfer_rs2pw 487 10.6 0.005 0.005 1.566 1.730 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.676 1.718 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.637 1.689 hybrid_alltoall_any 4725 16.4 0.054 0.116 1.204 1.650 fft3d_ps 1201 14.6 0.519 0.575 1.576 1.619 wfi_extrapolate 11 7.9 0.001 0.001 1.559 1.559 fft_wrap_pw1pw2_140 487 13.2 0.160 0.169 1.469 1.516 mp_alltoall_d11v 2130 13.8 1.382 1.487 1.382 1.487 grid_collocate_task_list 119 9.7 1.286 1.354 1.286 1.354 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.273 1.317 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.278 1.292 mp_allgather_i34 2286 14.5 0.698 1.169 0.698 1.169 mp_sum_d 4135 12.0 0.656 1.136 0.656 1.136 transfer_pw2rs 487 13.2 0.004 0.005 1.115 1.125 qs_energies_init_hamiltonians 11 5.9 0.001 0.005 1.109 1.111 make_images_sizes 4572 15.5 0.005 0.005 0.700 0.979 mp_alltoall_i44 4572 16.5 0.695 0.974 0.695 0.974 acc_transpose_blocks_sync 82296 16.5 0.817 0.947 0.817 0.947 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.906 0.921 mp_alltoall_z22v 1201 16.6 0.807 0.882 0.807 0.882 acc_transpose_blocks_kernels 27432 16.5 0.188 0.278 0.664 0.877 transfer_rs2pw_140 130 11.5 0.177 0.250 0.704 0.864 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=41.827000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=464.454545, yerr=1.671343 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e4289968d5d1c81be94a335915a0cf308a6045a0_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 521.146368E+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 1607811. 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 0.007 0.028 35.471 35.473 qs_mol_dyn_low 1 2.0 0.003 0.009 35.223 35.246 qs_forces 11 3.9 0.003 0.004 35.122 35.123 qs_energies 11 4.9 0.001 0.001 33.507 33.510 scf_env_do_scf 11 5.9 0.001 0.004 28.547 28.548 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 25.750 25.750 dbcsr_multiply_generic 2286 12.5 0.095 0.097 18.419 18.669 qs_scf_new_mos 108 7.5 0.001 0.001 17.415 17.427 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.415 17.426 velocity_verlet 10 3.0 0.001 0.002 17.042 17.045 ot_scf_mini 108 9.5 0.002 0.002 16.561 16.577 multiply_cannon 2286 13.5 0.194 0.197 14.310 15.271 multiply_cannon_loop 2286 14.5 0.856 0.924 13.387 14.532 ot_mini 108 10.5 0.001 0.001 10.227 10.252 qs_ot_get_derivative 108 11.5 0.001 0.001 8.554 8.568 multiply_cannon_multrec 18288 15.5 1.869 2.954 6.930 7.297 rebuild_ks_matrix 119 8.3 0.000 0.000 6.374 6.425 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.373 6.424 dbcsr_mm_accdrv_process 38222 16.0 4.892 5.633 4.969 5.699 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.621 5.666 mp_waitall_1 158411 16.6 3.563 5.299 3.563 5.299 qs_ot_get_p 119 10.4 0.001 0.001 3.902 3.938 sum_up_and_integrate 119 10.3 0.001 0.001 3.880 3.885 integrate_v_rspace 119 11.3 0.003 0.003 3.867 3.874 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.225 3.783 init_scf_run 11 5.9 0.000 0.001 3.733 3.734 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.733 3.733 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.683 3.698 calculate_rho_elec 119 8.7 0.031 0.031 3.682 3.697 multiply_cannon_metrocomm3 18288 15.5 0.048 0.050 1.855 3.462 init_scf_loop 11 6.9 0.000 0.000 2.766 2.766 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.504 2.513 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.046 2.488 apply_single 119 13.6 0.000 0.000 2.046 2.487 make_m2s 4572 13.5 0.044 0.045 2.263 2.469 make_images 4572 14.5 0.192 0.205 2.177 2.384 density_rs2pw 119 9.7 0.004 0.004 2.194 2.271 calculate_first_density_matrix 1 7.0 0.000 0.000 2.246 2.248 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.165 2.166 pw_transfer 1439 11.6 0.065 0.069 2.081 2.095 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.067 2.077 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.989 2.004 calculate_dm_sparse 119 9.5 0.000 0.001 1.965 1.986 prepare_preconditioner 11 7.9 0.000 0.000 1.961 1.967 make_preconditioner 11 8.9 0.000 0.000 1.961 1.967 mp_sum_l 7287 12.8 1.454 1.939 1.454 1.939 grid_integrate_task_list 119 12.3 1.804 1.910 1.804 1.910 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.811 1.902 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.865 1.873 cp_fm_diag_elpa_base 50 14.0 1.839 1.847 1.863 1.871 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.791 1.793 acc_transpose_blocks 18288 15.5 0.079 0.081 1.646 1.755 potential_pw2rs 119 12.3 0.007 0.009 1.629 1.637 ot_diis_step 108 11.5 0.011 0.012 1.625 1.627 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.565 1.583 fft3d_ps 1201 14.6 0.527 0.546 1.559 1.574 fft_wrap_pw1pw2_140 487 13.2 0.211 0.218 1.502 1.520 wfi_extrapolate 11 7.9 0.001 0.001 1.400 1.400 grid_collocate_task_list 119 9.7 1.232 1.330 1.232 1.330 transfer_rs2pw 487 10.6 0.005 0.005 1.229 1.323 make_images_data 4572 15.5 0.048 0.052 1.015 1.263 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.214 1.222 multiply_cannon_sync_h2d 18288 15.5 1.008 1.173 1.008 1.173 hybrid_alltoall_any 4725 16.4 0.058 0.117 0.871 1.117 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.068 1.081 transfer_pw2rs 487 13.2 0.004 0.004 1.004 1.009 mp_alltoall_d11v 2130 13.8 0.863 0.963 0.863 0.963 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 0.959 0.961 mp_alltoall_z22v 1201 16.6 0.855 0.928 0.855 0.928 cp_fm_cholesky_invert 11 10.9 0.865 0.870 0.865 0.870 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.828 0.834 acc_transpose_blocks_sync 54864 16.5 0.736 0.829 0.736 0.829 multiply_cannon_metrocomm1 18288 15.5 0.030 0.031 0.558 0.828 acc_transpose_blocks_kernels 18288 16.5 0.217 0.227 0.812 0.826 mp_allgather_i34 2286 14.5 0.372 0.782 0.372 0.782 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.676 0.735 make_images_sizes 4572 15.5 0.005 0.005 0.496 0.723 mp_alltoall_i44 4572 16.5 0.491 0.718 0.491 0.718 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=35.473000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.636364, yerr=2.226809 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e4289968d5d1c81be94a335915a0cf308a6045a0_performance_tests/06/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 114.044384E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3805952 0.0% 0.0% 100.0% average stack size 0.0 0.0 38.6 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 558.673920E+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 0.013 0.032 39.963 39.964 qs_mol_dyn_low 1 2.0 0.003 0.003 39.735 39.745 qs_forces 11 3.9 0.002 0.003 39.649 39.650 qs_energies 11 4.9 0.002 0.006 37.882 37.887 scf_env_do_scf 11 5.9 0.001 0.001 32.344 32.345 scf_env_do_scf_inner_loop 108 6.5 0.007 0.039 28.474 28.475 dbcsr_multiply_generic 2286 12.5 0.107 0.136 21.392 21.510 velocity_verlet 10 3.0 0.002 0.002 20.177 20.180 qs_scf_new_mos 108 7.5 0.001 0.001 19.688 19.759 qs_scf_loop_do_ot 108 8.5 0.001 0.001 19.688 19.759 ot_scf_mini 108 9.5 0.002 0.003 18.606 18.661 multiply_cannon 2286 13.5 0.219 0.225 16.710 17.240 multiply_cannon_loop 2286 14.5 1.547 1.623 15.560 15.910 ot_mini 108 10.5 0.001 0.001 11.304 11.375 qs_ot_get_derivative 108 11.5 0.001 0.001 9.341 9.399 multiply_cannon_multrec 27432 15.5 2.453 3.146 8.994 9.250 dbcsr_mm_accdrv_process 47916 15.9 6.040 7.671 6.438 7.770 rebuild_ks_matrix 119 8.3 0.000 0.000 6.754 6.806 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.753 6.806 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.001 6.048 qs_ot_get_p 119 10.4 0.001 0.002 4.264 4.341 init_scf_run 11 5.9 0.000 0.002 4.113 4.113 scf_env_initial_rho_setup 11 6.9 0.001 0.002 4.113 4.113 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.358 3.869 init_scf_loop 11 6.9 0.001 0.005 3.842 3.843 sum_up_and_integrate 119 10.3 0.001 0.001 3.828 3.836 integrate_v_rspace 119 11.3 0.003 0.003 3.817 3.825 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.742 3.798 calculate_rho_elec 119 8.7 0.040 0.046 3.742 3.798 mp_waitall_1 137007 16.6 2.575 3.088 2.575 3.088 make_m2s 4572 13.5 0.054 0.058 2.839 3.038 prepare_preconditioner 11 7.9 0.000 0.000 2.943 2.950 make_preconditioner 11 8.9 0.000 0.001 2.943 2.950 make_images 4572 14.5 0.273 0.335 2.731 2.929 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.518 2.859 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.284 2.769 apply_single 119 13.6 0.000 0.000 2.284 2.768 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.659 2.692 acc_transpose_blocks 27432 15.5 0.119 0.124 2.508 2.675 qs_ot_p2m_diag 50 11.0 0.016 0.024 2.613 2.626 calculate_first_density_matrix 1 7.0 0.001 0.012 2.566 2.568 calculate_dm_sparse 119 9.5 0.000 0.001 2.350 2.402 density_rs2pw 119 9.7 0.004 0.004 2.155 2.235 pw_transfer 1439 11.6 0.065 0.069 2.153 2.194 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.171 2.172 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.061 2.104 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.040 2.043 ot_diis_step 108 11.5 0.012 0.012 1.915 1.916 grid_integrate_task_list 119 12.3 1.835 1.909 1.835 1.909 mp_sum_l 7287 12.8 1.425 1.903 1.425 1.903 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.848 1.857 cp_fm_diag_elpa_base 50 14.0 1.811 1.829 1.846 1.856 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.786 1.802 fft_wrap_pw1pw2_140 487 13.2 0.245 0.259 1.651 1.696 fft3d_ps 1201 14.6 0.560 0.612 1.581 1.607 potential_pw2rs 119 12.3 0.009 0.010 1.598 1.603 acc_transpose_blocks_sync 82296 16.5 1.462 1.563 1.462 1.563 wfi_extrapolate 11 7.9 0.001 0.001 1.484 1.485 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.434 1.452 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.402 1.411 multiply_cannon_metrocomm3 27432 15.5 0.041 0.042 0.903 1.399 make_images_data 4572 15.5 0.048 0.055 1.135 1.366 grid_collocate_task_list 119 9.7 1.252 1.318 1.252 1.318 transfer_rs2pw 487 10.6 0.004 0.005 1.194 1.280 dbcsr_complete_redistribute 329 12.2 0.131 0.179 0.931 1.223 hybrid_alltoall_any 4725 16.4 0.066 0.154 0.969 1.218 cp_fm_upper_to_full 72 14.2 0.841 1.207 0.841 1.207 qs_energies_init_hamiltonians 11 5.9 0.006 0.010 1.144 1.144 jit_kernel_multiply 6 16.0 0.332 1.070 0.332 1.070 mp_alltoall_d11v 2130 13.8 0.925 1.067 0.925 1.067 acc_transpose_blocks_kernels 27432 16.5 0.272 0.282 0.897 0.994 transfer_pw2rs 487 13.2 0.004 0.004 0.970 0.975 cp_fm_cholesky_invert 11 10.9 0.954 0.958 0.954 0.958 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.674 0.950 mp_alltoall_z22v 1201 16.6 0.902 0.931 0.902 0.931 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.799 0.875 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.852 0.857 mp_alltoall_i22 627 13.8 0.498 0.812 0.498 0.812 multiply_cannon_metrocomm1 27432 15.5 0.036 0.037 0.349 0.800 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=39.964000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=528.454545, yerr=4.008256 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e4289968d5d1c81be94a335915a0cf308a6045a0_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 630.587392E+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 0.010 0.031 32.860 32.861 qs_mol_dyn_low 1 2.0 0.003 0.003 32.343 32.356 qs_forces 11 3.9 0.018 0.021 32.274 32.275 qs_energies 11 4.9 0.002 0.003 30.501 30.507 scf_env_do_scf 11 5.9 0.001 0.003 24.428 24.428 scf_env_do_scf_inner_loop 108 6.5 0.015 0.018 21.601 21.602 velocity_verlet 10 3.0 0.002 0.002 16.211 16.214 dbcsr_multiply_generic 2286 12.5 0.092 0.097 14.333 14.414 qs_scf_new_mos 108 7.5 0.001 0.001 13.268 13.298 qs_scf_loop_do_ot 108 8.5 0.001 0.001 13.267 13.297 ot_scf_mini 108 9.5 0.002 0.002 12.503 12.533 multiply_cannon 2286 13.5 0.223 0.226 11.216 11.840 multiply_cannon_loop 2286 14.5 0.645 0.665 10.190 10.394 ot_mini 108 10.5 0.001 0.001 6.998 7.034 multiply_cannon_multrec 9144 15.5 1.709 1.899 6.191 6.385 rebuild_ks_matrix 119 8.3 0.000 0.000 6.220 6.243 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 6.219 6.243 qs_ot_get_derivative 108 11.5 0.001 0.001 5.593 5.622 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.552 5.574 dbcsr_mm_accdrv_process 12550 15.8 3.706 4.400 4.372 4.444 init_scf_run 11 5.9 0.000 0.002 4.388 4.389 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.388 4.388 sum_up_and_integrate 119 10.3 0.001 0.001 3.794 3.798 integrate_v_rspace 119 11.3 0.003 0.003 3.783 3.788 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.695 3.706 calculate_rho_elec 119 8.7 0.060 0.061 3.694 3.706 qs_ot_get_p 119 10.4 0.001 0.001 3.480 3.518 calculate_first_density_matrix 1 7.0 0.000 0.000 2.991 2.992 init_scf_loop 11 6.9 0.000 0.000 2.803 2.805 mp_waitall_1 115863 16.7 2.061 2.635 2.061 2.635 make_m2s 4572 13.5 0.034 0.035 2.167 2.389 make_images 4572 14.5 0.268 0.300 2.078 2.299 pw_transfer 1439 11.6 0.065 0.068 2.243 2.257 qs_ot_p2m_diag 50 11.0 0.022 0.023 2.245 2.248 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.149 2.165 density_rs2pw 119 9.7 0.003 0.004 2.028 2.093 prepare_preconditioner 11 7.9 0.000 0.000 2.004 2.009 make_preconditioner 11 8.9 0.000 0.000 2.004 2.009 grid_integrate_task_list 119 12.3 1.877 1.974 1.877 1.974 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.971 1.973 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.924 1.945 calculate_dm_sparse 119 9.5 0.000 0.000 1.915 1.934 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.887 1.919 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.844 1.847 fft_wrap_pw1pw2_140 487 13.2 0.322 0.332 1.713 1.731 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.631 1.639 cp_fm_diag_elpa_base 50 14.0 1.603 1.619 1.629 1.636 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.617 1.628 acc_transpose_blocks 9144 15.5 0.041 0.042 1.561 1.590 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.573 1.588 fft3d_ps 1201 14.6 0.569 0.582 1.563 1.573 potential_pw2rs 119 12.3 0.010 0.011 1.535 1.542 qs_energies_init_hamiltonians 11 5.9 0.040 0.046 1.434 1.438 grid_collocate_task_list 119 9.7 1.313 1.401 1.313 1.401 ot_diis_step 108 11.5 0.013 0.013 1.382 1.385 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.302 1.346 apply_single 119 13.6 0.000 0.000 1.302 1.346 wfi_extrapolate 11 7.9 0.001 0.001 1.339 1.340 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.258 1.263 make_images_data 4572 15.5 0.042 0.046 0.974 1.257 hybrid_alltoall_any 4725 16.4 0.065 0.175 0.947 1.245 jit_kernel_multiply 6 15.4 0.625 1.178 0.625 1.178 mp_alltoall_d11v 2130 13.8 0.973 1.097 0.973 1.097 transfer_rs2pw 487 10.6 0.004 0.004 1.015 1.093 parallel_gemm_fm 81 9.0 0.000 0.000 1.005 1.007 parallel_gemm_fm_cosma 81 10.0 1.005 1.007 1.005 1.007 cp_fm_cholesky_invert 11 10.9 1.000 1.004 1.000 1.004 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.876 0.926 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.908 0.917 mp_alltoall_z22v 1201 16.6 0.867 0.902 0.867 0.902 make_basis_sm 11 9.8 0.000 0.000 0.877 0.879 transfer_pw2rs 487 13.2 0.003 0.004 0.844 0.849 mp_allgather_i34 2286 14.5 0.339 0.849 0.339 0.849 multiply_cannon_metrocomm3 9144 15.5 0.021 0.021 0.381 0.840 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.786 0.832 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.798 0.800 acc_transpose_blocks_kernels 9144 16.5 0.118 0.120 0.780 0.782 acc_transpose_blocks_sync 27432 16.5 0.724 0.750 0.724 0.750 dbcsr_complete_redistribute 329 12.2 0.170 0.184 0.639 0.674 multiply_cannon_metrocomm1 9144 15.5 0.023 0.024 0.561 0.672 jit_kernel_transpose 5 15.6 0.662 0.665 0.662 0.665 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=32.861000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=596.090909, yerr=7.948179 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e4289968d5d1c81be94a335915a0cf308a6045a0_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 778.047488E+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 0.011 0.029 47.202 47.204 qs_mol_dyn_low 1 2.0 0.003 0.003 46.980 46.990 qs_forces 11 3.9 0.002 0.003 46.917 46.918 qs_energies 11 4.9 0.002 0.002 44.819 44.824 scf_env_do_scf 11 5.9 0.001 0.001 38.754 38.754 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 30.223 30.226 velocity_verlet 10 3.0 0.002 0.002 26.506 26.512 dbcsr_multiply_generic 2286 12.5 0.101 0.103 21.009 21.163 qs_scf_new_mos 108 7.5 0.001 0.001 19.997 20.098 qs_scf_loop_do_ot 108 8.5 0.001 0.001 19.996 20.098 ot_scf_mini 108 9.5 0.002 0.002 18.747 18.844 multiply_cannon 2286 13.5 0.301 0.305 15.943 17.017 multiply_cannon_loop 2286 14.5 0.879 0.904 14.515 15.226 ot_mini 108 10.5 0.001 0.001 11.360 11.496 qs_ot_get_derivative 108 11.5 0.001 0.001 9.122 9.218 multiply_cannon_multrec 9144 15.5 3.471 4.830 8.848 8.929 init_scf_loop 11 6.9 0.000 0.000 8.496 8.501 rebuild_ks_matrix 119 8.3 0.000 0.000 7.646 7.804 qs_ks_build_kohn_sham_matrix 119 9.3 0.014 0.014 7.646 7.804 prepare_preconditioner 11 7.9 0.000 0.000 7.462 7.476 make_preconditioner 11 8.9 0.000 0.000 7.462 7.476 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.974 7.351 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.872 7.012 dbcsr_mm_accdrv_process 12550 15.8 4.592 6.525 5.236 6.569 cp_fm_upper_to_full 72 14.2 3.198 4.630 3.198 4.630 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.499 4.553 calculate_rho_elec 119 8.7 0.118 0.121 4.498 4.553 qs_ot_get_p 119 10.4 0.001 0.002 4.173 4.327 mp_waitall_1 94719 16.7 3.542 4.229 3.542 4.229 sum_up_and_integrate 119 10.3 0.001 0.001 4.154 4.159 integrate_v_rspace 119 11.3 0.003 0.003 4.143 4.149 init_scf_run 11 5.9 0.000 0.001 3.961 3.961 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.961 3.961 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.293 3.716 make_m2s 4572 13.5 0.037 0.038 3.084 3.334 make_images 4572 14.5 0.351 0.384 2.964 3.213 dbcsr_complete_redistribute 329 12.2 0.283 0.287 2.139 2.993 pw_transfer 1439 11.6 0.068 0.069 2.914 2.923 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.816 2.825 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.351 2.672 apply_single 119 13.6 0.000 0.000 2.351 2.671 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.772 2.611 multiply_cannon_metrocomm3 9144 15.5 0.021 0.022 1.941 2.597 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.469 2.473 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.402 2.453 calculate_dm_sparse 119 9.5 0.000 0.001 2.399 2.453 mp_sum_l 7287 12.8 1.434 2.446 1.434 2.446 density_rs2pw 119 9.7 0.003 0.004 2.365 2.377 mp_alltoall_i22 627 13.8 1.503 2.360 1.503 2.360 fft_wrap_pw1pw2_140 487 13.2 0.576 0.581 2.322 2.333 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.483 2.315 calculate_first_density_matrix 1 7.0 0.000 0.000 2.173 2.229 ot_diis_step 108 11.5 0.014 0.015 2.187 2.188 grid_integrate_task_list 119 12.3 2.173 2.186 2.173 2.186 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.070 2.072 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.946 1.946 fft3d_ps 1201 14.6 0.596 0.607 1.913 1.920 make_images_data 4572 15.5 0.046 0.050 1.509 1.890 hybrid_alltoall_any 4725 16.4 0.090 0.148 1.490 1.885 acc_transpose_blocks 9144 15.5 0.043 0.044 1.852 1.883 cp_fm_cholesky_invert 11 10.9 1.855 1.858 1.855 1.858 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.792 1.793 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.721 1.721 cp_fm_diag_elpa_base 50 14.0 1.569 1.625 1.719 1.719 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.621 1.678 wfi_extrapolate 11 7.9 0.001 0.001 1.664 1.664 grid_collocate_task_list 119 9.7 1.585 1.595 1.585 1.595 potential_pw2rs 119 12.3 0.014 0.014 1.529 1.532 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.487 1.523 mp_alltoall_d11v 2130 13.8 1.417 1.480 1.417 1.480 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.328 1.377 mp_alltoall_z22v 1201 16.6 1.179 1.205 1.179 1.205 acc_transpose_blocks_sync 27432 16.5 1.121 1.167 1.121 1.167 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.125 1.141 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 0.991 1.046 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.990 1.008 qs_create_task_list 11 7.9 0.000 0.000 0.940 0.952 generate_qs_task_list 11 8.9 0.368 0.389 0.940 0.952 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=47.204000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=734.818182, yerr=12.783641 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e4289968d5d1c81be94a335915a0cf308a6045a0_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 501.321728E+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 1383689. 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 0.009 0.030 87.107 87.108 qs_mol_dyn_low 1 2.0 0.003 0.004 86.816 86.830 qs_forces 11 3.9 0.003 0.004 86.736 86.737 qs_energies 11 4.9 0.002 0.003 83.760 83.778 scf_env_do_scf 11 5.9 0.000 0.001 74.645 74.648 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 68.292 68.293 dbcsr_multiply_generic 2055 12.4 0.107 0.110 52.707 53.053 qs_scf_new_mos 99 7.5 0.000 0.001 50.085 50.179 qs_scf_loop_do_ot 99 8.5 0.000 0.001 50.084 50.178 ot_scf_mini 99 9.5 0.002 0.002 47.551 47.648 velocity_verlet 10 3.0 0.002 0.002 45.600 45.601 multiply_cannon 2055 13.4 0.183 0.188 43.217 44.239 multiply_cannon_loop 2055 14.4 1.791 1.827 41.998 43.099 ot_mini 99 10.5 0.001 0.001 27.592 27.675 qs_ot_get_derivative 99 11.5 0.001 0.001 20.722 20.807 multiply_cannon_multrec 49320 15.4 11.255 12.004 17.248 18.004 rebuild_ks_matrix 110 8.3 0.000 0.000 15.268 15.423 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.015 15.268 15.422 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.408 13.534 mp_waitall_1 220248 16.4 11.930 13.042 11.930 13.042 qs_ot_get_p 110 10.4 0.001 0.002 11.203 11.316 multiply_cannon_sync_h2d 49320 15.4 9.419 10.074 9.419 10.074 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.833 8.372 multiply_cannon_metrocomm3 49320 15.4 0.084 0.088 6.733 8.044 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.236 7.697 apply_single 110 13.6 0.000 0.000 7.236 7.697 qs_ot_p2m_diag 48 11.0 0.012 0.019 7.440 7.497 sum_up_and_integrate 110 10.3 0.001 0.003 7.482 7.496 integrate_v_rspace 110 11.3 0.003 0.003 7.456 7.476 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.935 7.092 calculate_rho_elec 110 8.6 0.020 0.024 6.935 7.091 init_scf_run 11 5.9 0.000 0.002 7.018 7.018 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.017 7.018 ot_diis_step 99 11.5 0.006 0.006 6.636 6.637 cp_dbcsr_syevd 48 12.0 0.002 0.003 6.494 6.495 init_scf_loop 11 6.9 0.000 0.000 6.313 6.314 dbcsr_mm_accdrv_process 87628 16.1 3.024 3.105 5.865 6.137 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.938 5.993 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.768 5.788 cp_fm_diag_elpa_base 48 14.0 5.745 5.767 5.765 5.786 mp_sum_l 6594 12.7 4.448 5.431 4.448 5.431 make_m2s 4110 13.4 0.061 0.065 4.591 4.701 make_images 4110 14.4 0.177 0.190 4.493 4.606 wfi_extrapolate 11 7.9 0.001 0.001 4.262 4.262 calculate_dm_sparse 110 9.5 0.001 0.001 3.912 4.033 density_rs2pw 110 9.6 0.004 0.005 3.852 4.000 prepare_preconditioner 11 7.9 0.000 0.000 3.981 3.999 make_preconditioner 11 8.9 0.000 0.000 3.981 3.999 multiply_cannon_metrocomm1 49320 15.4 0.065 0.068 2.902 3.843 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.758 3.800 pw_transfer 1331 11.6 0.055 0.064 3.627 3.716 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.644 3.650 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 3.539 3.629 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.448 3.496 grid_integrate_task_list 110 12.3 3.275 3.456 3.275 3.456 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.237 3.299 potential_pw2rs 110 12.3 0.006 0.006 2.997 3.020 fft_wrap_pw1pw2_140 451 13.1 0.387 0.428 2.891 2.984 fft3d_ps 1111 14.6 0.798 0.890 2.836 2.907 mp_alltoall_d11v 2046 13.8 2.372 2.828 2.372 2.828 calculate_first_density_matrix 1 7.0 0.000 0.000 2.646 2.650 acc_transpose_blocks 49320 15.4 0.214 0.224 2.579 2.628 jit_kernel_multiply 13 15.9 2.562 2.580 2.562 2.580 grid_collocate_task_list 110 9.6 2.155 2.284 2.155 2.284 transfer_rs2pw 451 10.6 0.005 0.006 2.168 2.282 cp_fm_cholesky_invert 11 10.9 2.271 2.275 2.271 2.275 make_images_data 4110 15.4 0.044 0.047 2.030 2.194 mp_waitany 14300 13.8 1.934 2.148 1.934 2.148 hybrid_alltoall_any 4261 16.3 0.084 0.484 1.743 2.035 mp_sum_d 3889 11.9 1.517 2.035 1.517 2.035 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.997 2.024 transfer_pw2rs 451 13.1 0.006 0.007 1.997 2.018 mp_alltoall_z22v 1111 16.6 1.683 1.836 1.683 1.836 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=87.108000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=475.818182, yerr=3.157046 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e4289968d5d1c81be94a335915a0cf308a6045a0_performance_tests/10/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 390.715586E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 5019072 0.0% 0.0% 100.0% average stack size 0.0 0.0 196.1 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 592.384000E+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 2090479. 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 0.029 0.057 72.533 72.535 qs_mol_dyn_low 1 2.0 0.003 0.003 72.239 72.251 qs_forces 11 3.9 0.003 0.005 72.164 72.165 qs_energies 11 4.9 0.002 0.008 68.770 68.775 scf_env_do_scf 11 5.9 0.000 0.001 59.792 59.796 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 51.406 51.408 dbcsr_multiply_generic 2055 12.4 0.115 0.119 39.410 39.709 velocity_verlet 10 3.0 0.002 0.003 37.970 37.971 qs_scf_new_mos 99 7.5 0.001 0.001 34.935 35.050 qs_scf_loop_do_ot 99 8.5 0.001 0.001 34.935 35.049 ot_scf_mini 99 9.5 0.003 0.003 33.205 33.325 multiply_cannon 2055 13.4 0.223 0.244 32.124 33.206 multiply_cannon_loop 2055 14.4 1.170 1.199 30.718 31.634 ot_mini 99 10.5 0.001 0.001 19.278 19.414 multiply_cannon_multrec 24660 15.4 6.941 8.410 13.976 15.628 rebuild_ks_matrix 110 8.3 0.000 0.000 13.981 14.074 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.015 13.981 14.074 qs_ot_get_derivative 99 11.5 0.001 0.001 13.382 13.506 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.289 12.369 mp_waitall_1 176588 16.5 8.521 11.157 8.521 11.157 init_scf_loop 11 6.9 0.001 0.006 8.351 8.354 multiply_cannon_metrocomm3 24660 15.4 0.074 0.077 5.629 8.341 multiply_cannon_sync_h2d 24660 15.4 6.380 7.487 6.380 7.487 qs_ot_get_p 110 10.4 0.002 0.002 7.094 7.217 dbcsr_mm_accdrv_process 52282 16.1 5.616 6.326 6.867 7.180 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.541 7.155 apply_single 110 13.6 0.000 0.001 6.541 7.155 sum_up_and_integrate 110 10.3 0.001 0.003 6.702 6.714 integrate_v_rspace 110 11.3 0.003 0.003 6.675 6.689 init_scf_run 11 5.9 0.000 0.001 6.564 6.565 scf_env_initial_rho_setup 11 6.9 0.001 0.002 6.564 6.565 prepare_preconditioner 11 7.9 0.000 0.001 6.245 6.261 make_preconditioner 11 8.9 0.000 0.002 6.245 6.261 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.201 6.223 calculate_rho_elec 110 8.6 0.039 0.047 6.200 6.222 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.793 5.967 ot_diis_step 99 11.5 0.010 0.010 5.823 5.824 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.945 5.638 make_m2s 4110 13.4 0.057 0.060 4.585 5.041 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.974 4.999 make_images 4110 14.4 0.407 0.467 4.476 4.929 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.450 4.452 pw_transfer 1331 11.6 0.066 0.074 3.793 3.936 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.817 3.829 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.686 3.829 cp_fm_diag_elpa_base 48 14.0 3.766 3.785 3.815 3.826 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.603 3.672 wfi_extrapolate 11 7.9 0.001 0.001 3.647 3.647 density_rs2pw 110 9.6 0.004 0.005 3.404 3.576 calculate_dm_sparse 110 9.5 0.001 0.001 3.292 3.326 grid_integrate_task_list 110 12.3 3.160 3.325 3.160 3.325 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.296 3.299 fft_wrap_pw1pw2_140 451 13.1 0.458 0.471 3.037 3.180 cp_fm_cholesky_invert 11 10.9 3.011 3.021 3.011 3.021 make_images_data 4110 15.4 0.049 0.053 2.512 2.996 hybrid_alltoall_any 4261 16.3 0.107 0.454 2.196 2.984 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.925 2.967 fft3d_ps 1111 14.6 1.110 1.332 2.784 2.941 mp_sum_l 6594 12.7 2.074 2.853 2.074 2.853 calculate_first_density_matrix 1 7.0 0.001 0.010 2.816 2.820 potential_pw2rs 110 12.3 0.009 0.009 2.534 2.552 grid_collocate_task_list 110 9.6 2.166 2.284 2.166 2.284 acc_transpose_blocks 24660 15.4 0.114 0.117 2.057 2.186 mp_alltoall_d11v 2046 13.8 1.924 2.136 1.924 2.136 qs_ot_get_orbitals 99 10.5 0.001 0.001 2.100 2.115 jit_kernel_multiply 10 16.3 0.887 1.928 0.887 1.928 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.882 1.884 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.853 1.866 mp_allgather_i34 2055 14.4 0.710 1.827 0.710 1.827 multiply_cannon_metrocomm4 22605 15.4 0.078 0.082 0.776 1.816 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.799 1.811 transfer_rs2pw 451 10.6 0.006 0.007 1.595 1.788 mp_irecv_dv 57340 16.2 0.649 1.593 0.649 1.593 cp_fm_cholesky_decompose 22 10.9 1.546 1.551 1.546 1.551 dbcsr_complete_redistribute 325 12.2 0.244 0.296 1.237 1.530 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.002 1.386 1.492 mp_alltoall_z22v 1111 16.6 1.403 1.480 1.403 1.480 transfer_pw2rs 451 13.1 0.005 0.007 1.455 1.471 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=72.535000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=560.000000, yerr=7.032393 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e4289968d5d1c81be94a335915a0cf308a6045a0_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 664.711168E+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 0.012 0.029 63.290 63.291 qs_mol_dyn_low 1 2.0 0.003 0.003 62.947 62.958 qs_forces 11 3.9 0.003 0.003 62.863 62.864 qs_energies 11 4.9 0.003 0.011 59.644 59.648 scf_env_do_scf 11 5.9 0.001 0.001 51.414 51.414 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 42.167 42.168 velocity_verlet 10 3.0 0.004 0.006 34.350 34.352 dbcsr_multiply_generic 2055 12.4 0.108 0.116 30.041 30.426 qs_scf_new_mos 99 7.5 0.001 0.001 27.158 27.258 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.158 27.257 ot_scf_mini 99 9.5 0.002 0.004 25.850 25.955 multiply_cannon 2055 13.4 0.215 0.228 23.044 24.369 multiply_cannon_loop 2055 14.4 0.814 0.845 21.733 23.221 ot_mini 99 10.5 0.001 0.001 14.598 14.715 rebuild_ks_matrix 110 8.3 0.000 0.000 12.388 12.561 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.388 12.561 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.921 11.078 mp_waitall_1 139946 16.5 7.746 10.776 7.746 10.776 multiply_cannon_multrec 16440 15.4 3.624 4.536 9.778 10.583 qs_ot_get_derivative 99 11.5 0.001 0.001 9.992 10.094 init_scf_loop 11 6.9 0.001 0.004 9.207 9.208 multiply_cannon_metrocomm3 16440 15.4 0.047 0.048 4.787 7.848 prepare_preconditioner 11 7.9 0.000 0.000 7.399 7.418 make_preconditioner 11 8.9 0.000 0.001 7.399 7.418 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.701 7.061 sum_up_and_integrate 110 10.3 0.001 0.002 6.434 6.450 integrate_v_rspace 110 11.3 0.003 0.003 6.408 6.424 qs_ot_get_p 110 10.4 0.002 0.003 6.224 6.391 dbcsr_mm_accdrv_process 34862 16.1 5.341 5.663 6.000 6.155 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.943 5.959 calculate_rho_elec 110 8.6 0.058 0.059 5.942 5.959 init_scf_run 11 5.9 0.000 0.001 5.745 5.745 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.745 5.745 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.059 5.544 apply_single 110 13.6 0.000 0.000 5.058 5.544 make_m2s 4110 13.4 0.050 0.056 4.578 4.932 make_images 4110 14.4 0.396 0.516 4.463 4.810 ot_diis_step 99 11.5 0.011 0.011 4.553 4.554 qs_ot_p2m_diag 48 11.0 0.042 0.044 4.373 4.383 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.396 4.038 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.959 3.960 multiply_cannon_sync_h2d 16440 15.4 3.193 3.782 3.193 3.782 pw_transfer 1331 11.6 0.065 0.073 3.711 3.721 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.604 3.616 grid_integrate_task_list 110 12.3 3.181 3.388 3.181 3.388 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.340 3.350 density_rs2pw 110 9.6 0.004 0.005 3.152 3.349 cp_fm_diag_elpa_base 48 14.0 3.271 3.307 3.338 3.348 wfi_extrapolate 11 7.9 0.001 0.001 3.138 3.139 cp_fm_cholesky_invert 11 10.9 3.102 3.109 3.102 3.109 make_images_data 4110 15.4 0.046 0.050 2.606 3.093 fft_wrap_pw1pw2_140 451 13.1 0.579 0.586 3.075 3.090 hybrid_alltoall_any 4261 16.3 0.109 0.381 2.303 3.025 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.932 2.935 qs_ot_get_derivative_diag 47 12.0 0.001 0.002 2.719 2.779 calculate_dm_sparse 110 9.5 0.001 0.001 2.619 2.656 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.526 2.590 mp_sum_l 6594 12.7 1.804 2.583 1.804 2.583 fft3d_ps 1111 14.6 1.089 1.100 2.546 2.556 calculate_first_density_matrix 1 7.0 0.001 0.003 2.509 2.513 multiply_cannon_metrocomm4 14385 15.4 0.048 0.052 0.891 2.461 grid_collocate_task_list 110 9.6 2.211 2.416 2.211 2.416 mp_irecv_dv 48980 15.7 0.817 2.329 0.817 2.329 potential_pw2rs 110 12.3 0.011 0.012 2.247 2.256 mp_alltoall_d11v 2046 13.8 1.936 2.220 1.936 2.220 dbcsr_complete_redistribute 325 12.2 0.338 0.381 1.588 2.039 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 2.001 2.002 mp_allgather_i34 2055 14.4 0.620 1.783 0.620 1.783 cp_fm_upper_to_full 70 14.2 1.381 1.777 1.381 1.777 cp_fm_cholesky_decompose 22 10.9 1.749 1.774 1.749 1.774 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.664 1.681 acc_transpose_blocks 16440 15.4 0.076 0.078 1.600 1.678 transfer_rs2pw 451 10.6 0.005 0.006 1.414 1.608 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.057 1.523 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.500 1.511 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.361 1.480 mp_waitany 17072 13.8 1.231 1.474 1.231 1.474 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.396 1.406 transfer_pw2rs 451 13.1 0.005 0.005 1.263 1.270 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=63.291000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=630.000000, yerr=7.942979 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e4289968d5d1c81be94a335915a0cf308a6045a0_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 743.059456E+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 0.010 0.030 68.988 68.989 qs_mol_dyn_low 1 2.0 0.003 0.004 68.747 68.758 qs_forces 11 3.9 0.003 0.004 68.672 68.674 qs_energies 11 4.9 0.001 0.002 65.262 65.266 scf_env_do_scf 11 5.9 0.000 0.001 56.579 56.582 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 44.100 44.100 velocity_verlet 10 3.0 0.002 0.002 38.998 39.001 dbcsr_multiply_generic 2055 12.4 0.113 0.119 31.765 32.007 qs_scf_new_mos 99 7.5 0.001 0.001 28.948 29.064 qs_scf_loop_do_ot 99 8.5 0.001 0.001 28.947 29.064 ot_scf_mini 99 9.5 0.002 0.003 27.206 27.299 multiply_cannon 2055 13.4 0.239 0.260 24.194 25.419 multiply_cannon_loop 2055 14.4 1.402 1.485 22.557 23.350 ot_mini 99 10.5 0.001 0.001 15.510 15.622 multiply_cannon_multrec 24660 15.4 4.095 6.863 13.121 14.359 init_scf_loop 11 6.9 0.000 0.000 12.433 12.434 rebuild_ks_matrix 110 8.3 0.000 0.000 12.247 12.360 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.247 12.360 qs_ot_get_derivative 99 11.5 0.001 0.001 11.271 11.364 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.835 10.937 prepare_preconditioner 11 7.9 0.000 0.000 10.652 10.665 make_preconditioner 11 8.9 0.000 0.000 10.652 10.664 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.859 10.322 dbcsr_mm_accdrv_process 52304 16.0 7.746 9.348 8.868 9.845 qs_ot_get_p 110 10.4 0.058 0.095 6.474 6.624 mp_waitall_1 121746 16.5 4.761 6.619 4.761 6.619 sum_up_and_integrate 110 10.3 0.001 0.002 6.453 6.466 integrate_v_rspace 110 11.3 0.003 0.003 6.427 6.439 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.164 6.177 calculate_rho_elec 110 8.6 0.078 0.081 6.164 6.177 make_m2s 4110 13.4 0.060 0.062 5.716 6.119 make_images 4110 14.4 0.576 0.700 5.574 5.973 init_scf_run 11 5.9 0.000 0.002 5.887 5.888 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.887 5.888 cp_fm_upper_to_full 70 14.2 3.402 4.919 3.402 4.919 qs_ot_p2m_diag 48 11.0 0.055 0.064 4.426 4.501 ot_diis_step 99 11.5 0.011 0.011 4.190 4.190 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.097 4.176 apply_single 110 13.6 0.000 0.000 4.097 4.176 pw_transfer 1331 11.6 0.065 0.074 3.887 3.925 dbcsr_complete_redistribute 325 12.2 0.420 0.463 2.763 3.906 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.828 3.884 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.780 3.822 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.666 3.715 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.565 3.619 multiply_cannon_metrocomm3 24660 15.4 0.038 0.040 1.602 3.617 grid_integrate_task_list 110 12.3 3.287 3.423 3.287 3.423 density_rs2pw 110 9.6 0.004 0.004 3.235 3.346 calculate_dm_sparse 110 9.5 0.001 0.001 3.297 3.340 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.182 3.318 make_images_data 4110 15.4 0.049 0.053 2.852 3.283 hybrid_alltoall_any 4261 16.3 0.123 0.460 2.469 3.267 fft_wrap_pw1pw2_140 451 13.1 0.610 0.631 3.174 3.221 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.175 3.184 cp_fm_diag_elpa_base 48 14.0 3.020 3.081 3.172 3.182 wfi_extrapolate 11 7.9 0.001 0.001 3.108 3.108 cp_fm_cholesky_invert 11 10.9 3.046 3.057 3.046 3.057 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.783 2.894 mp_alltoall_i22 605 13.7 1.686 2.866 1.686 2.866 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.842 2.843 calculate_first_density_matrix 1 7.0 0.000 0.000 2.679 2.682 fft3d_ps 1111 14.6 1.087 1.116 2.661 2.676 acc_transpose_blocks 24660 15.4 0.112 0.116 2.401 2.551 multiply_cannon_sync_h2d 24660 15.4 2.379 2.491 2.379 2.491 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.454 2.487 grid_collocate_task_list 110 9.6 2.264 2.376 2.264 2.376 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.241 2.242 potential_pw2rs 110 12.3 0.013 0.014 2.208 2.213 mp_alltoall_d11v 2046 13.8 1.950 2.152 1.950 2.152 jit_kernel_multiply 10 15.6 0.774 2.089 0.774 2.089 cp_fm_cholesky_decompose 22 10.9 1.988 2.037 1.988 2.037 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.889 1.924 mp_sum_l 6594 12.7 1.166 1.862 1.166 1.862 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.807 1.824 mp_allgather_i34 2055 14.4 0.715 1.736 0.715 1.736 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.614 1.707 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.655 1.666 acc_transpose_blocks_sync 73980 16.4 1.385 1.529 1.385 1.529 transfer_rs2pw 451 10.6 0.005 0.006 1.345 1.472 multiply_cannon_metrocomm4 20550 15.4 0.062 0.065 0.853 1.434 mp_alltoall_z22v 1111 16.6 1.357 1.419 1.357 1.419 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=68.989000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=701.545455, yerr=12.115832 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e4289968d5d1c81be94a335915a0cf308a6045a0_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 867.868672E+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 0.010 0.029 58.450 58.452 qs_mol_dyn_low 1 2.0 0.003 0.003 58.179 58.189 qs_forces 11 3.9 0.003 0.003 58.107 58.109 qs_energies 11 4.9 0.001 0.001 54.342 54.348 scf_env_do_scf 11 5.9 0.000 0.001 45.677 45.677 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 37.057 37.058 velocity_verlet 10 3.0 0.002 0.003 33.002 33.005 dbcsr_multiply_generic 2055 12.4 0.105 0.110 24.429 24.591 qs_scf_new_mos 99 7.5 0.001 0.001 22.161 22.219 qs_scf_loop_do_ot 99 8.5 0.001 0.001 22.161 22.218 ot_scf_mini 99 9.5 0.002 0.002 20.854 20.892 multiply_cannon 2055 13.4 0.238 0.252 18.400 19.596 multiply_cannon_loop 2055 14.4 0.605 0.637 17.042 17.379 rebuild_ks_matrix 110 8.3 0.000 0.000 11.901 11.956 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 11.900 11.955 ot_mini 99 10.5 0.001 0.001 11.473 11.509 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.538 10.589 multiply_cannon_multrec 8220 15.4 3.190 4.426 7.644 8.718 mp_waitall_1 103326 16.6 6.759 8.646 6.759 8.646 init_scf_loop 11 6.9 0.000 0.000 8.565 8.566 qs_ot_get_derivative 99 11.5 0.001 0.001 7.659 7.698 prepare_preconditioner 11 7.9 0.000 0.000 6.866 6.871 make_preconditioner 11 8.9 0.000 0.000 6.866 6.871 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.421 6.496 sum_up_and_integrate 110 10.3 0.001 0.002 6.402 6.416 integrate_v_rspace 110 11.3 0.003 0.003 6.375 6.389 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.191 6.210 calculate_rho_elec 110 8.6 0.114 0.114 6.190 6.209 init_scf_run 11 5.9 0.000 0.002 5.500 5.500 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.499 5.500 qs_ot_get_p 110 10.4 0.002 0.002 5.288 5.330 dbcsr_mm_accdrv_process 17442 15.9 3.155 3.949 4.312 5.303 make_m2s 4110 13.4 0.038 0.039 4.551 4.845 multiply_cannon_metrocomm3 8220 15.4 0.019 0.020 3.282 4.823 make_images 4110 14.4 0.649 0.712 4.421 4.713 pw_transfer 1331 11.6 0.066 0.070 4.022 4.032 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.914 3.927 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.841 3.883 apply_single 110 13.6 0.000 0.000 3.841 3.882 ot_diis_step 99 11.5 0.012 0.012 3.790 3.790 qs_ot_p2m_diag 48 11.0 0.082 0.094 3.755 3.769 grid_integrate_task_list 110 12.3 3.380 3.492 3.380 3.492 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.391 3.391 cp_fm_cholesky_invert 11 10.9 3.384 3.388 3.384 3.388 fft_wrap_pw1pw2_140 451 13.1 0.779 0.790 3.349 3.361 density_rs2pw 110 9.6 0.004 0.004 3.121 3.272 wfi_extrapolate 11 7.9 0.001 0.001 2.916 2.916 hybrid_alltoall_any 4261 16.3 0.203 0.850 2.473 2.881 make_images_data 4110 15.4 0.041 0.047 2.482 2.869 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.798 2.806 cp_fm_diag_elpa_base 48 14.0 2.742 2.765 2.797 2.804 calculate_dm_sparse 110 9.5 0.001 0.001 2.722 2.761 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.687 2.689 fft3d_ps 1111 14.6 1.138 1.157 2.576 2.598 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.566 2.568 grid_collocate_task_list 110 9.6 2.365 2.556 2.365 2.556 calculate_first_density_matrix 1 7.0 0.000 0.000 2.473 2.476 multiply_cannon_sync_h2d 8220 15.4 2.354 2.475 2.354 2.475 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.236 2.268 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.162 2.179 potential_pw2rs 110 12.3 0.015 0.016 2.126 2.130 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.106 2.129 mp_alltoall_d11v 2046 13.8 1.943 2.065 1.943 2.065 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.791 2.014 cp_fm_cholesky_decompose 22 10.9 1.965 1.996 1.965 1.996 mp_allgather_i34 2055 14.4 0.526 1.726 0.526 1.726 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.697 1.701 dbcsr_complete_redistribute 325 12.2 0.557 0.580 1.556 1.644 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.623 1.634 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.515 1.630 multiply_cannon_metrocomm1 8220 15.4 0.022 0.023 1.084 1.561 acc_transpose_blocks 8220 15.4 0.039 0.040 1.383 1.437 transfer_rs2pw 451 10.6 0.005 0.005 1.239 1.417 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.299 1.338 mp_waitany 9240 13.8 1.127 1.335 1.127 1.335 qs_create_task_list 11 7.9 0.000 0.001 1.224 1.322 generate_qs_task_list 11 8.9 0.373 0.441 1.224 1.321 multiply_cannon_metrocomm4 6165 15.4 0.019 0.020 0.474 1.278 mp_irecv_dv 24056 15.7 0.449 1.233 0.449 1.233 mp_alltoall_z22v 1111 16.6 1.203 1.227 1.203 1.227 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=58.452000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=816.363636, yerr=12.571777 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e4289968d5d1c81be94a335915a0cf308a6045a0_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 1.416024E+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 0.061 0.080 90.310 90.311 qs_mol_dyn_low 1 2.0 0.011 0.067 89.922 89.933 qs_forces 11 3.9 0.003 0.003 89.835 89.844 qs_energies 11 4.9 0.002 0.002 85.704 85.714 scf_env_do_scf 11 5.9 0.001 0.001 75.510 75.511 velocity_verlet 10 3.0 0.002 0.002 57.453 57.460 scf_env_do_scf_inner_loop 99 6.5 0.004 0.007 46.367 46.369 dbcsr_multiply_generic 2055 12.4 0.119 0.124 30.742 30.885 init_scf_loop 11 6.9 0.000 0.000 29.066 29.068 qs_scf_new_mos 99 7.5 0.001 0.001 28.489 28.572 qs_scf_loop_do_ot 99 8.5 0.001 0.001 28.488 28.571 prepare_preconditioner 11 7.9 0.000 0.000 27.019 27.032 make_preconditioner 11 8.9 0.000 0.000 27.019 27.032 ot_scf_mini 99 9.5 0.002 0.002 26.630 26.703 make_full_inverse_cholesky 11 9.9 0.000 0.000 21.390 26.494 multiply_cannon 2055 13.4 0.334 0.369 23.043 23.768 multiply_cannon_loop 2055 14.4 0.833 0.844 21.133 21.534 cp_fm_upper_to_full 70 14.2 12.654 18.010 12.654 18.010 ot_mini 99 10.5 0.001 0.001 15.014 15.087 rebuild_ks_matrix 110 8.3 0.000 0.000 14.031 14.121 qs_ks_build_kohn_sham_matrix 110 9.3 0.016 0.016 14.030 14.120 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.742 12.822 dbcsr_complete_redistribute 325 12.2 1.014 1.033 7.289 10.327 qs_ot_get_derivative 99 11.5 0.001 0.001 10.190 10.274 multiply_cannon_multrec 8220 15.4 4.060 4.218 9.753 9.877 mp_waitall_1 84994 16.7 8.333 9.512 8.333 9.512 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.198 9.231 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.614 8.616 mp_alltoall_i22 605 13.7 5.274 8.347 5.274 8.347 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.327 7.361 calculate_rho_elec 110 8.6 0.225 0.225 7.326 7.360 sum_up_and_integrate 110 10.3 0.002 0.002 7.031 7.047 integrate_v_rspace 110 11.3 0.004 0.004 7.004 7.019 make_m2s 4110 13.4 0.043 0.044 5.697 6.241 qs_ot_get_p 110 10.4 0.003 0.008 6.121 6.213 make_images 4110 14.4 0.882 0.932 5.509 6.051 cp_fm_cholesky_invert 11 10.9 5.886 5.892 5.886 5.892 init_scf_run 11 5.9 0.000 0.002 5.854 5.854 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.853 5.853 dbcsr_mm_accdrv_process 11614 15.7 3.895 4.148 5.544 5.763 multiply_cannon_metrocomm3 8220 15.4 0.019 0.020 5.319 5.685 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.983 5.503 apply_single 110 13.6 0.000 0.000 4.983 5.503 pw_transfer 1331 11.6 0.074 0.075 5.173 5.181 fft_wrap_pw1pw2 1111 12.6 0.009 0.010 5.056 5.065 ot_diis_step 99 11.5 0.015 0.016 4.795 4.795 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.395 4.403 fft_wrap_pw1pw2_140 451 13.1 1.281 1.286 4.377 4.386 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.855 3.856 hybrid_alltoall_any 4261 16.3 0.264 0.566 3.017 3.784 make_images_data 4110 15.4 0.045 0.048 3.063 3.777 grid_integrate_task_list 110 12.3 3.698 3.755 3.698 3.755 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.211 3.751 density_rs2pw 110 9.6 0.004 0.004 3.708 3.737 qs_energies_init_hamiltonians 11 5.9 0.002 0.003 3.710 3.719 wfi_extrapolate 11 7.9 0.001 0.001 3.431 3.431 calculate_dm_sparse 110 9.5 0.001 0.001 3.230 3.259 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.224 3.224 cp_fm_diag_elpa_base 48 14.0 2.675 2.882 3.221 3.221 multiply_cannon_sync_h2d 8220 15.4 3.115 3.135 3.115 3.135 fft3d_ps 1111 14.6 1.304 1.317 3.108 3.121 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.960 2.963 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.876 2.919 grid_collocate_task_list 110 9.6 2.684 2.706 2.684 2.706 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.424 2.458 potential_pw2rs 110 12.3 0.022 0.022 2.447 2.454 mp_alltoall_d11v 2046 13.8 2.296 2.358 2.296 2.358 calculate_first_density_matrix 1 7.0 0.000 0.000 2.305 2.306 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.233 2.297 cp_fm_cholesky_decompose 22 10.9 2.272 2.296 2.272 2.296 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.098 2.191 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.877 1.939 qs_create_task_list 11 7.9 0.001 0.001 1.884 1.928 generate_qs_task_list 11 8.9 0.732 0.785 1.883 1.927 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.889 1.898 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=90.311000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1269.454545, yerr=68.024546 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e4289968d5d1c81be94a335915a0cf308a6045a0_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430458527744 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613083000832 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228655E+12 0.0% 0.0% 100.0% flops max/rank 1.103326E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806314816 0.0% 0.0% 100.0% number of processed stacks 12044928 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.1 marketing flops 145.647559E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 632.053760E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10431456 MPI messages size (bytes): total size 4.526012E+12 min size 0.000000E+00 max size 4.537280E+06 average size 433.881094E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3605352 96403587072 32768 < size <= 131072 1305088 74666999808 131072 < size <= 4194304 5190114 3200148065288 4194304 < size <= 16777216 263934 1154745839544 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4030 57365. MP_Allreduce 11154 793. MP_Sync 87 MP_Alltoall 2242 3164610. MP_SendRecv 24510 18752. MP_ISendRecv 24510 18752. MP_Wait 42802 MP_ISend 16140 108019. MP_IRecv 16140 108019. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.028 213.938 213.940 qs_mol_dyn_low 1 2.0 0.003 0.004 213.503 213.518 qs_forces 11 3.9 0.005 0.006 213.394 213.396 qs_energies 11 4.9 0.002 0.005 207.730 207.749 scf_env_do_scf 11 5.9 0.001 0.002 191.000 191.004 scf_env_do_scf_inner_loop 118 6.6 0.010 0.038 169.457 169.459 qs_scf_new_mos 118 7.6 0.001 0.001 128.748 129.031 qs_scf_loop_do_ot 118 8.6 0.001 0.001 128.747 129.030 dbcsr_multiply_generic 2527 12.6 0.181 0.185 127.753 128.740 velocity_verlet 10 3.0 0.002 0.002 128.391 128.392 ot_scf_mini 118 9.6 0.003 0.003 121.926 122.160 multiply_cannon 2527 13.6 0.241 0.249 102.831 104.704 multiply_cannon_loop 2527 14.6 2.418 2.478 100.486 102.576 ot_mini 118 10.6 0.001 0.001 67.740 67.996 multiply_cannon_multrec 60648 15.6 32.133 33.827 41.913 43.830 qs_ot_get_derivative 118 11.6 0.001 0.001 42.364 42.614 rebuild_ks_matrix 129 8.3 0.001 0.001 34.382 34.773 qs_ks_build_kohn_sham_matrix 129 9.3 0.015 0.017 34.382 34.772 mp_waitall_1 269254 16.5 29.854 32.989 29.854 32.989 qs_ot_get_p 129 10.4 0.001 0.002 31.563 31.966 qs_ks_update_qs_env 129 7.6 0.001 0.001 30.926 31.285 multiply_cannon_sync_h2d 60648 15.6 26.511 28.984 26.511 28.984 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 24.756 25.868 apply_single 129 13.6 0.001 0.001 24.756 25.868 ot_diis_step 118 11.6 0.008 0.008 25.081 25.083 qs_ot_p2m_diag 83 11.4 0.079 0.091 24.630 24.683 cp_dbcsr_syevd 83 12.4 0.005 0.005 21.758 21.759 init_scf_loop 11 6.9 0.000 0.000 21.460 21.462 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.961 20.153 multiply_cannon_metrocomm3 60648 15.6 0.120 0.127 16.432 18.717 cp_fm_diag_elpa 83 13.4 0.000 0.000 18.576 18.624 cp_fm_diag_elpa_base 83 14.4 18.492 18.541 18.572 18.621 prepare_preconditioner 11 7.9 0.000 0.000 16.822 16.879 make_preconditioner 11 8.9 0.000 0.000 16.822 16.879 make_full_inverse_cholesky 11 9.9 0.000 0.000 16.015 16.206 make_m2s 5054 13.6 0.105 0.113 14.378 14.651 sum_up_and_integrate 129 10.3 0.002 0.004 14.524 14.540 integrate_v_rspace 129 11.3 0.004 0.004 14.465 14.483 make_images 5054 14.6 0.400 0.420 14.193 14.481 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.695 13.821 calculate_rho_elec 129 8.7 0.045 0.064 13.695 13.821 init_scf_run 11 5.9 0.000 0.003 12.537 12.537 scf_env_initial_rho_setup 11 6.9 0.001 0.003 12.537 12.537 mp_sum_l 8010 12.9 9.277 10.583 9.277 10.583 dbcsr_mm_accdrv_process 125468 16.2 4.711 4.885 9.350 9.886 cp_fm_cholesky_invert 11 10.9 9.573 9.580 9.573 9.580 wfi_extrapolate 11 7.9 0.001 0.001 9.336 9.336 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 8.624 8.775 calculate_dm_sparse 129 9.5 0.001 0.001 8.635 8.726 qs_ot_get_orbitals 118 10.6 0.001 0.001 8.276 8.369 pw_transfer 1559 11.6 0.075 0.097 7.952 8.150 multiply_cannon_metrocomm1 60648 15.6 0.094 0.097 6.446 8.026 make_images_data 5054 15.6 0.068 0.073 7.052 7.976 fft_wrap_pw1pw2 1301 12.7 0.010 0.013 7.748 7.942 density_rs2pw 129 9.7 0.006 0.007 7.225 7.764 grid_integrate_task_list 129 12.3 7.092 7.575 7.092 7.575 hybrid_alltoall_any 5240 16.5 0.297 2.284 6.191 7.540 fft_wrap_pw1pw2_140 527 13.2 1.130 1.179 6.681 6.863 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.671 6.683 fft3d_ps 1301 14.7 2.211 2.862 5.742 6.094 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.756 5.855 mp_alltoall_d11v 2423 14.1 4.669 5.769 4.669 5.769 grid_collocate_task_list 129 9.7 4.867 5.277 4.867 5.277 cp_fm_cholesky_decompose 22 10.9 4.947 4.962 4.947 4.962 potential_pw2rs 129 12.3 0.009 0.010 4.872 4.914 mp_sum_d 4502 12.1 3.614 4.398 3.614 4.398 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=213.940000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=597.818182, yerr=6.421812 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e4289968d5d1c81be94a335915a0cf308a6045a0_performance_tests/16/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 2.199914E+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 6022464 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.2 marketing flops 145.647559E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 846.905344E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2425920 MPI messages size (bytes): total size 4.132350E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703416E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 71436 2336489472 32768 < size <= 131072 728832 55956209664 131072 < size <= 4194304 1386864 1409906900992 4194304 < size <= 16777216 155760 1473826487232 16777216 < size 68112 1190343475200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4030 57818. MP_Allreduce 11153 957. MP_Sync 87 MP_Alltoall 1983 5058697. MP_SendRecv 12126 47072. MP_ISendRecv 12126 47072. MP_Wait 26114 MP_ISend 11836 212447. MP_IRecv 11836 212447. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.020 0.044 226.459 226.462 qs_mol_dyn_low 1 2.0 0.004 0.006 225.965 225.984 qs_forces 11 3.9 0.007 0.016 225.858 225.862 qs_energies 11 4.9 0.004 0.014 218.629 218.643 scf_env_do_scf 11 5.9 0.001 0.002 200.425 200.435 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 156.958 156.961 velocity_verlet 10 3.0 0.002 0.002 143.112 143.114 qs_scf_new_mos 118 7.6 0.001 0.001 111.724 112.442 qs_scf_loop_do_ot 118 8.6 0.001 0.001 111.723 112.441 ot_scf_mini 118 9.6 0.004 0.005 106.649 107.274 dbcsr_multiply_generic 2527 12.6 0.192 0.199 104.814 106.235 multiply_cannon 2527 13.6 0.509 0.565 79.575 85.940 multiply_cannon_loop 2527 14.6 1.587 1.661 75.383 78.951 ot_mini 118 10.6 0.001 0.001 55.856 56.583 mp_waitall_1 216438 16.6 30.031 43.995 30.031 43.995 init_scf_loop 11 6.9 0.009 0.033 43.341 43.345 prepare_preconditioner 11 7.9 0.000 0.001 38.437 38.516 make_preconditioner 11 8.9 0.001 0.003 38.437 38.516 make_full_inverse_cholesky 11 9.9 0.000 0.001 36.882 37.623 rebuild_ks_matrix 129 8.3 0.001 0.001 36.531 37.186 qs_ks_build_kohn_sham_matrix 129 9.3 0.035 0.071 36.530 37.185 multiply_cannon_multrec 30324 15.6 20.601 24.688 31.223 35.802 qs_ot_get_derivative 118 11.6 0.001 0.002 33.108 33.762 qs_ks_update_qs_env 129 7.6 0.001 0.001 33.001 33.592 qs_ot_get_p 129 10.4 0.003 0.006 31.866 32.562 multiply_cannon_metrocomm3 30324 15.6 0.099 0.106 17.208 31.294 qs_ot_p2m_diag 83 11.4 0.189 0.222 25.518 25.576 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 22.666 23.928 apply_single 129 13.6 0.001 0.001 22.666 23.928 cp_dbcsr_syevd 83 12.4 0.006 0.011 23.911 23.916 ot_diis_step 118 11.6 0.015 0.015 22.526 22.529 cp_fm_cholesky_invert 11 10.9 22.491 22.509 22.491 22.509 cp_fm_diag_elpa 83 13.4 0.001 0.002 19.873 19.900 cp_fm_diag_elpa_base 83 14.4 19.487 19.670 19.858 19.881 make_m2s 5054 13.6 0.090 0.094 17.967 19.710 multiply_cannon_sync_h2d 30324 15.6 17.594 19.521 17.594 19.521 make_images 5054 14.6 1.181 1.394 17.755 19.495 sum_up_and_integrate 129 10.3 0.002 0.004 17.402 17.435 integrate_v_rspace 129 11.3 0.003 0.005 17.340 17.374 qs_rho_update_rho_low 129 7.7 0.001 0.004 16.895 16.925 calculate_rho_elec 129 8.7 0.088 0.106 16.894 16.924 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 13.650 14.116 init_scf_run 11 5.9 0.000 0.004 13.029 13.031 scf_env_initial_rho_setup 11 6.9 0.002 0.004 13.029 13.031 make_images_data 5054 15.6 0.067 0.077 10.343 12.799 hybrid_alltoall_any 5240 16.5 0.354 1.544 8.667 11.826 multiply_cannon_metrocomm4 27797 15.6 0.105 0.121 3.984 10.956 pw_transfer 1559 11.6 0.086 0.103 10.676 10.750 cp_fm_cholesky_decompose 22 10.9 10.629 10.720 10.629 10.720 dbcsr_mm_accdrv_process 62734 16.2 5.553 6.449 10.093 10.709 density_rs2pw 129 9.7 0.006 0.008 10.298 10.644 mp_irecv_dv 70031 16.3 3.782 10.551 3.782 10.551 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 10.451 10.525 wfi_extrapolate 11 7.9 0.001 0.002 9.716 9.717 fft_wrap_pw1pw2_140 527 13.2 1.218 1.246 8.348 8.426 fft3d_ps 1301 14.7 2.864 3.047 7.997 8.065 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 7.094 7.897 mp_sum_l 8010 12.9 5.476 7.697 5.476 7.697 grid_integrate_task_list 129 12.3 7.168 7.646 7.168 7.646 potential_pw2rs 129 12.3 0.016 0.019 7.532 7.569 calculate_dm_sparse 129 9.5 0.001 0.002 6.859 6.992 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.326 6.335 mp_allgather_i34 2527 14.6 2.701 6.155 2.701 6.155 mp_alltoall_d11v 2423 14.1 5.449 6.053 5.449 6.053 qs_ot_get_orbitals 118 10.6 0.001 0.001 5.972 6.048 transfer_rs2pw 527 10.6 0.007 0.008 5.177 5.524 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.252 5.466 grid_collocate_task_list 129 9.7 5.041 5.442 5.041 5.442 dbcsr_complete_redistribute 395 12.7 0.769 0.874 4.049 5.096 make_images_sizes 5054 15.6 0.007 0.007 2.897 5.008 mp_alltoall_i44 5054 16.6 2.890 5.001 2.890 5.001 mp_sum_d 4502 12.1 3.141 4.675 3.141 4.675 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=226.462000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=802.636364, yerr=4.922540 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e4289968d5d1c81be94a335915a0cf308a6045a0_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 2.928533E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 3984192 0.0% 0.0% 100.0% average stack size 0.0 0.0 1695.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 959.041536E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1042912 MPI messages size (bytes): total size 2.716210E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.604448E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 281856 36943429632 131072 < size <= 4194304 660064 996105256960 4194304 < size <= 16777216 65632 931531265168 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 58208. MP_Allreduce 11082 999. MP_Sync 87 MP_Alltoall 1712 9388896. MP_SendRecv 7936 75008. MP_ISendRecv 7936 75008. MP_Wait 21820 MP_ISend 11748 275205. MP_IRecv 11748 275205. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.030 182.862 182.863 qs_mol_dyn_low 1 2.0 0.003 0.003 182.385 182.399 qs_forces 11 3.9 0.004 0.005 182.288 182.290 qs_energies 11 4.9 0.007 0.008 175.591 175.603 scf_env_do_scf 11 5.9 0.001 0.004 159.490 159.499 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 122.682 122.684 velocity_verlet 10 3.0 0.002 0.002 117.184 117.187 dbcsr_multiply_generic 2507 12.6 0.180 0.189 84.003 85.398 qs_scf_new_mos 117 7.6 0.001 0.001 84.686 85.089 qs_scf_loop_do_ot 117 8.6 0.001 0.001 84.685 85.088 ot_scf_mini 117 9.6 0.003 0.004 80.367 80.785 multiply_cannon 2507 13.6 0.508 0.526 63.564 67.297 multiply_cannon_loop 2507 14.6 1.133 1.199 60.227 63.807 ot_mini 117 10.6 0.001 0.001 43.873 44.294 init_scf_loop 11 6.9 0.000 0.000 36.695 36.696 mp_waitall_1 170520 16.6 26.229 35.673 26.229 35.673 prepare_preconditioner 11 7.9 0.000 0.000 32.585 32.646 make_preconditioner 11 8.9 0.000 0.000 32.585 32.646 make_full_inverse_cholesky 11 9.9 0.000 0.000 30.221 31.595 rebuild_ks_matrix 128 8.3 0.001 0.001 30.929 31.458 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.018 30.929 31.457 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.824 28.306 multiply_cannon_metrocomm3 20056 15.6 0.063 0.067 16.432 26.378 multiply_cannon_multrec 20056 15.6 12.822 16.011 22.278 25.389 qs_ot_get_derivative 117 11.6 0.002 0.002 23.846 24.250 qs_ot_get_p 128 10.4 0.002 0.003 21.537 22.008 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 20.047 21.210 apply_single 128 13.6 0.001 0.001 20.047 21.209 ot_diis_step 117 11.6 0.018 0.019 19.893 19.893 qs_ot_p2m_diag 83 11.4 0.265 0.273 17.094 17.124 make_m2s 5014 13.6 0.080 0.085 15.091 16.366 make_images 5014 14.6 1.153 1.241 14.858 16.134 cp_dbcsr_syevd 83 12.4 0.005 0.006 15.891 15.892 cp_fm_cholesky_invert 11 10.9 15.535 15.543 15.535 15.543 multiply_cannon_sync_h2d 20056 15.6 13.441 15.144 13.441 15.144 sum_up_and_integrate 128 10.3 0.002 0.003 14.435 14.461 integrate_v_rspace 128 11.3 0.003 0.004 14.376 14.402 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.994 14.036 calculate_rho_elec 128 8.7 0.130 0.145 13.993 14.035 cp_fm_diag_elpa 83 13.4 0.000 0.001 12.461 12.478 cp_fm_diag_elpa_base 83 14.4 12.038 12.210 12.457 12.475 init_scf_run 11 5.9 0.000 0.001 10.965 10.965 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.964 10.965 make_images_data 5014 15.6 0.063 0.072 8.979 10.669 hybrid_alltoall_any 5200 16.5 0.452 2.045 7.754 9.942 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.500 9.808 multiply_cannon_metrocomm4 17549 15.6 0.066 0.075 3.581 9.719 mp_irecv_dv 50230 16.2 3.452 9.458 3.452 9.458 pw_transfer 1547 11.6 0.085 0.103 9.149 9.264 dbcsr_mm_accdrv_process 41502 16.2 5.778 6.054 8.938 9.065 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.925 9.048 cp_fm_cholesky_decompose 22 10.9 7.935 7.957 7.935 7.957 fft_wrap_pw1pw2_140 523 13.2 1.300 1.337 7.814 7.945 wfi_extrapolate 11 7.9 0.001 0.001 7.751 7.751 grid_integrate_task_list 128 12.3 7.339 7.748 7.339 7.748 density_rs2pw 128 9.7 0.006 0.006 7.285 7.698 cp_fm_upper_to_full 105 14.8 5.689 7.211 5.689 7.211 dbcsr_complete_redistribute 395 12.7 1.172 1.202 4.748 6.486 fft3d_ps 1291 14.7 2.760 2.980 6.304 6.388 calculate_dm_sparse 128 9.5 0.001 0.001 5.977 6.080 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.685 5.691 grid_collocate_task_list 128 9.7 5.169 5.607 5.169 5.607 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.879 5.500 mp_alltoall_d11v 2415 14.1 4.534 5.457 4.534 5.457 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.507 5.239 mp_sum_l 7950 12.9 3.545 5.105 3.545 5.105 mp_allgather_i34 2507 14.6 1.831 5.064 1.831 5.064 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.730 4.878 potential_pw2rs 128 12.3 0.021 0.023 4.800 4.824 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.113 4.155 transfer_fm_to_dbcsr 11 9.9 0.019 0.024 2.345 4.026 mp_alltoall_i22 716 14.1 1.976 3.802 1.976 3.802 qs_energies_init_hamiltonians 11 5.9 0.009 0.012 3.786 3.788 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.607 3.661 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=182.863000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=896.181818, yerr=17.413778 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e4289968d5d1c81be94a335915a0cf308a6045a0_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 4.353788E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 5977344 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.151816E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1143192 MPI messages size (bytes): total size 2.023815E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.770319E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 319024 36042702848 131072 < size <= 4194304 715736 785529176064 4194304 < size <= 16777216 70320 665379241840 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58348. MP_Allreduce 11057 1083. MP_Sync 87 MP_Alltoall 1712 12503107. MP_SendRecv 5888 75008. MP_ISendRecv 5888 75008. MP_Wait 22442 MP_ISend 14952 244818. MP_IRecv 14952 244818. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.035 0.076 194.692 194.693 qs_mol_dyn_low 1 2.0 0.003 0.004 194.179 194.193 qs_forces 11 3.9 0.005 0.013 194.046 194.056 qs_energies 11 4.9 0.003 0.008 186.830 186.846 scf_env_do_scf 11 5.9 0.001 0.003 169.459 169.473 velocity_verlet 10 3.0 0.002 0.002 129.063 129.067 scf_env_do_scf_inner_loop 117 6.6 0.005 0.011 120.849 120.851 qs_scf_new_mos 117 7.6 0.001 0.001 83.891 84.207 qs_scf_loop_do_ot 117 8.6 0.001 0.002 83.890 84.206 dbcsr_multiply_generic 2507 12.6 0.189 0.205 81.965 82.698 ot_scf_mini 117 9.6 0.003 0.004 79.261 79.639 multiply_cannon 2507 13.6 0.551 0.587 56.432 59.030 multiply_cannon_loop 2507 14.6 1.824 1.903 52.421 54.186 init_scf_loop 11 6.9 0.003 0.017 48.468 48.472 ot_mini 117 10.6 0.001 0.001 44.011 44.373 prepare_preconditioner 11 7.9 0.000 0.001 44.291 44.315 make_preconditioner 11 8.9 0.000 0.002 44.291 44.315 make_full_inverse_cholesky 11 9.9 0.012 0.023 37.869 42.914 multiply_cannon_multrec 30084 15.6 13.507 19.084 26.373 31.629 rebuild_ks_matrix 128 8.3 0.001 0.001 29.735 30.018 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.019 29.735 30.017 mp_waitall_1 147882 16.7 18.017 27.431 18.017 27.431 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.786 27.040 qs_ot_get_derivative 117 11.6 0.002 0.002 24.152 24.530 make_m2s 5014 13.6 0.096 0.100 20.821 22.067 make_images 5014 14.6 1.955 2.309 20.510 21.755 qs_ot_get_p 128 10.4 0.003 0.005 21.212 21.629 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 19.275 19.869 apply_single 128 13.6 0.001 0.001 19.275 19.869 ot_diis_step 117 11.6 0.018 0.020 19.729 19.731 cp_fm_cholesky_invert 11 10.9 17.508 17.517 17.508 17.517 cp_fm_upper_to_full 105 14.8 11.573 17.037 11.573 17.037 qs_ot_p2m_diag 83 11.4 0.343 0.390 16.789 16.850 cp_dbcsr_syevd 83 12.4 0.005 0.006 15.314 15.315 multiply_cannon_metrocomm3 30084 15.6 0.050 0.053 6.708 15.078 sum_up_and_integrate 128 10.3 0.002 0.003 14.341 14.373 integrate_v_rspace 128 11.3 0.004 0.005 14.281 14.316 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.147 14.187 calculate_rho_elec 128 8.7 0.173 0.189 14.147 14.187 make_images_data 5014 15.6 0.067 0.074 11.248 13.326 dbcsr_complete_redistribute 395 12.7 1.509 1.655 9.402 13.211 dbcsr_mm_accdrv_process 62264 16.2 8.391 9.174 12.436 12.994 hybrid_alltoall_any 5200 16.5 0.529 2.190 10.163 12.334 cp_fm_diag_elpa 83 13.4 0.000 0.001 11.940 11.956 cp_fm_diag_elpa_base 83 14.4 10.917 11.265 11.933 11.948 copy_fm_to_dbcsr 209 11.7 0.002 0.002 7.900 11.704 multiply_cannon_sync_h2d 30084 15.6 10.497 11.475 10.497 11.475 init_scf_run 11 5.9 0.000 0.002 11.258 11.259 scf_env_initial_rho_setup 11 6.9 0.002 0.003 11.257 11.259 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 10.162 10.444 transfer_fm_to_dbcsr 11 9.9 0.001 0.010 6.399 10.118 mp_alltoall_i22 716 14.1 5.632 9.504 5.632 9.504 pw_transfer 1547 11.6 0.085 0.102 9.319 9.397 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 9.093 9.178 cp_fm_cholesky_decompose 22 10.9 8.122 8.220 8.122 8.220 fft_wrap_pw1pw2_140 523 13.2 1.445 1.474 8.027 8.123 wfi_extrapolate 11 7.9 0.001 0.001 8.088 8.089 grid_integrate_task_list 128 12.3 7.563 7.822 7.563 7.822 density_rs2pw 128 9.7 0.006 0.006 7.070 7.474 multiply_cannon_metrocomm4 25070 15.6 0.084 0.095 2.874 7.214 mp_irecv_dv 76098 16.2 2.719 6.935 2.719 6.935 calculate_dm_sparse 128 9.5 0.001 0.001 6.346 6.429 fft3d_ps 1291 14.7 2.853 2.941 6.269 6.321 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.689 5.742 grid_collocate_task_list 128 9.7 5.286 5.713 5.286 5.713 mp_alltoall_d11v 2415 14.1 5.330 5.711 5.330 5.711 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.632 4.729 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.483 4.587 potential_pw2rs 128 12.3 0.023 0.023 4.558 4.573 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 4.539 4.541 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.382 4.443 mp_sum_l 7950 12.9 2.783 4.005 2.783 4.005 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=194.693000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1086.272727, yerr=16.186925 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e4289968d5d1c81be94a335915a0cf308a6045a0_performance_tests/19/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 5.909858E+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 1976192 0.0% 0.0% 100.0% average stack size 0.0 0.0 3444.2 marketing flops 145.646636E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 1.539674E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 242592 MPI messages size (bytes): total size 1.341705E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.530704E+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 114816 60196651008 4194304 < size <= 16777216 105840 554906419200 16777216 < size 20352 726591807936 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 9003 51. MP_Alltoall 9708 795382. MP_ISend 40388 2101966. MP_IRecv 40388 2101070. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4030 57807. MP_Allreduce 11152 1164. MP_Sync 87 MP_Alltoall 1724 18848449. 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 0.015 0.036 177.356 177.358 qs_mol_dyn_low 1 2.0 0.003 0.004 176.886 176.900 qs_forces 11 3.9 0.004 0.005 176.787 176.789 qs_energies 11 4.9 0.002 0.002 169.318 169.327 scf_env_do_scf 11 5.9 0.001 0.005 151.768 151.778 velocity_verlet 10 3.0 0.003 0.003 116.603 116.607 scf_env_do_scf_inner_loop 118 6.6 0.006 0.011 113.934 113.936 qs_scf_new_mos 118 7.6 0.001 0.001 75.989 76.137 qs_scf_loop_do_ot 118 8.6 0.001 0.001 75.988 76.136 dbcsr_multiply_generic 2527 12.6 0.182 0.193 74.709 75.174 ot_scf_mini 118 9.6 0.003 0.004 71.431 71.559 multiply_cannon 2527 13.6 0.564 0.593 54.886 58.222 multiply_cannon_loop 2527 14.6 0.818 0.853 51.693 52.668 ot_mini 118 10.6 0.001 0.001 39.258 39.408 init_scf_loop 11 6.9 0.000 0.000 37.664 37.666 prepare_preconditioner 11 7.9 0.000 0.000 33.732 33.755 make_preconditioner 11 8.9 0.000 0.000 33.732 33.755 mp_waitall_1 126780 16.7 26.671 33.520 26.671 33.520 make_full_inverse_cholesky 11 9.9 0.017 0.026 31.635 31.925 rebuild_ks_matrix 129 8.3 0.001 0.001 29.874 30.001 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.018 29.874 30.000 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.264 27.378 multiply_cannon_multrec 10108 15.6 10.399 14.904 18.116 23.046 qs_ot_get_derivative 118 11.6 0.001 0.002 21.665 21.807 multiply_cannon_metrocomm3 10108 15.6 0.026 0.027 13.404 21.170 cp_fm_cholesky_invert 11 10.9 19.636 19.642 19.636 19.642 qs_ot_get_p 129 10.4 0.002 0.002 19.022 19.184 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 17.564 17.718 apply_single 129 13.6 0.001 0.001 17.564 17.718 ot_diis_step 118 11.6 0.020 0.021 17.503 17.505 make_m2s 5054 13.6 0.067 0.070 15.737 16.694 make_images 5054 14.6 2.183 2.609 15.429 16.388 qs_ot_p2m_diag 83 11.4 0.495 0.501 14.946 14.964 sum_up_and_integrate 129 10.3 0.002 0.002 14.605 14.653 integrate_v_rspace 129 11.3 0.004 0.004 14.544 14.594 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.516 14.554 calculate_rho_elec 129 8.7 0.258 0.269 14.515 14.554 cp_dbcsr_syevd 83 12.4 0.005 0.006 13.708 13.709 multiply_cannon_sync_h2d 10108 15.6 10.758 11.354 10.758 11.354 cp_fm_diag_elpa 83 13.4 0.000 0.000 10.668 10.679 cp_fm_diag_elpa_base 83 14.4 10.411 10.503 10.663 10.675 init_scf_run 11 5.9 0.000 0.002 10.673 10.673 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.673 10.673 make_images_data 5054 15.6 0.056 0.065 9.201 10.627 hybrid_alltoall_any 5240 16.5 0.849 3.800 8.899 10.382 pw_transfer 1559 11.6 0.086 0.094 9.732 9.760 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 9.507 9.542 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 8.522 8.630 cp_fm_cholesky_decompose 22 10.9 8.385 8.512 8.385 8.512 fft_wrap_pw1pw2_140 527 13.2 1.815 1.856 8.322 8.359 grid_integrate_task_list 129 12.3 7.812 8.202 7.812 8.202 dbcsr_mm_accdrv_process 20926 16.1 3.154 4.224 7.367 8.100 multiply_cannon_metrocomm1 10108 15.6 0.030 0.031 4.697 8.002 wfi_extrapolate 11 7.9 0.001 0.001 7.552 7.552 density_rs2pw 129 9.7 0.006 0.006 7.031 7.365 calculate_dm_sparse 129 9.5 0.001 0.001 6.302 6.386 fft3d_ps 1301 14.7 2.790 2.865 6.215 6.249 grid_collocate_task_list 129 9.7 5.611 5.865 5.611 5.865 multiply_cannon_metrocomm4 7581 15.6 0.027 0.030 1.873 5.755 dbcsr_complete_redistribute 395 12.7 2.122 2.166 5.323 5.716 mp_irecv_dv 29086 15.9 1.832 5.655 1.832 5.655 mp_alltoall_d11v 2423 14.1 4.957 5.485 4.957 5.485 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.246 5.257 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 5.250 5.252 mp_allgather_i34 2527 14.6 1.369 5.139 1.369 5.139 potential_pw2rs 129 12.3 0.027 0.028 4.595 4.600 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.101 4.154 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 3.870 3.945 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.590 3.875 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.801 3.851 copy_fm_to_dbcsr 209 11.7 0.001 0.002 3.410 3.740 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.569 3.669 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=177.358000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1451.090909, yerr=30.224778 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e4289968d5d1c81be94a335915a0cf308a6045a0_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 1973537472512 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 94.243766E+12 0.0% 0.0% 100.0% flops max/rank 11.787405E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806547424 0.0% 0.0% 100.0% number of processed stacks 1982128 0.0% 0.0% 100.0% average stack size 0.0 0.0 3434.0 marketing flops 145.661668E+12 ------------------------------------------------------------------------------- # multiplications 2534 max memory usage/rank 3.216507E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101360 MPI messages size (bytes): total size 1.145137E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.297725E+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 45848 35601252352 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592540656 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4041 58586. MP_Allreduce 11179 1500. 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 0.047 0.069 301.733 301.739 qs_mol_dyn_low 1 2.0 0.003 0.004 301.162 301.178 qs_forces 11 3.9 0.004 0.005 301.065 301.067 qs_energies 11 4.9 0.002 0.003 292.144 292.152 scf_env_do_scf 11 5.9 0.001 0.002 269.442 269.449 velocity_verlet 10 3.0 0.006 0.008 217.610 217.618 scf_env_do_scf_inner_loop 118 6.6 0.004 0.009 141.670 141.671 init_scf_loop 11 6.9 0.000 0.000 127.484 127.485 prepare_preconditioner 11 7.9 0.000 0.000 122.616 122.638 make_preconditioner 11 8.9 0.000 0.000 122.616 122.638 make_full_inverse_cholesky 11 9.9 0.038 0.039 98.356 119.683 qs_scf_new_mos 118 7.6 0.001 0.001 96.316 96.390 qs_scf_loop_do_ot 118 8.6 0.001 0.001 96.315 96.390 ot_scf_mini 118 9.6 0.004 0.004 91.257 91.298 dbcsr_multiply_generic 2534 12.6 0.214 0.228 85.706 86.334 cp_fm_upper_to_full 106 14.8 52.856 76.038 52.856 76.038 multiply_cannon 2534 13.6 0.677 0.709 60.831 62.280 multiply_cannon_loop 2534 14.6 1.063 1.079 56.508 58.259 ot_mini 118 10.6 0.001 0.001 47.010 47.062 dbcsr_complete_redistribute 397 12.7 4.012 4.059 30.447 43.568 copy_fm_to_dbcsr 210 11.7 0.001 0.002 26.801 39.867 transfer_fm_to_dbcsr 11 9.9 0.030 0.030 24.217 37.152 cp_fm_cholesky_invert 11 10.9 35.584 35.590 35.584 35.590 rebuild_ks_matrix 129 8.3 0.001 0.001 35.205 35.277 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.018 35.204 35.277 mp_waitall_1 104780 16.8 30.536 35.006 30.536 35.006 mp_alltoall_i22 720 14.1 21.946 34.841 21.946 34.841 qs_ks_update_qs_env 129 7.6 0.001 0.001 32.698 32.776 qs_ot_get_p 129 10.4 0.003 0.004 28.745 28.838 qs_ot_get_derivative 118 11.6 0.002 0.002 26.329 26.365 qs_ot_p2m_diag 84 11.4 0.891 0.896 24.238 24.275 cp_dbcsr_syevd 84 12.4 0.006 0.006 22.347 22.350 multiply_cannon_metrocomm3 10136 15.6 0.025 0.027 20.181 21.942 make_m2s 5068 13.6 0.077 0.081 19.899 21.611 make_images 5068 14.6 3.085 3.272 19.415 21.126 ot_diis_step 118 11.6 0.022 0.022 20.639 20.640 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.907 20.085 apply_single 129 13.6 0.001 0.001 19.907 20.085 cp_fm_diag_elpa 84 13.4 0.000 0.000 18.911 18.912 cp_fm_diag_elpa_base 84 14.4 14.384 15.947 18.904 18.904 multiply_cannon_multrec 10136 15.6 10.391 12.242 18.213 18.285 qs_rho_update_rho_low 129 7.7 0.001 0.001 17.313 17.326 calculate_rho_elec 129 8.7 0.485 0.485 17.312 17.325 sum_up_and_integrate 129 10.3 0.002 0.002 16.751 16.843 integrate_v_rspace 129 11.3 0.004 0.004 16.688 16.780 multiply_cannon_sync_h2d 10136 15.6 14.413 14.438 14.413 14.438 make_images_data 5068 15.6 0.065 0.071 11.503 14.062 hybrid_alltoall_any 5255 16.5 1.322 3.057 11.544 14.026 init_scf_run 11 5.9 0.000 0.001 12.358 12.360 scf_env_initial_rho_setup 11 6.9 0.001 0.002 12.358 12.359 pw_transfer 1559 11.6 0.096 0.096 12.126 12.131 fft_wrap_pw1pw2 1301 12.7 0.012 0.012 11.886 11.893 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 10.596 10.631 fft_wrap_pw1pw2_140 527 13.2 3.070 3.131 10.518 10.523 cp_fm_cholesky_decompose 22 10.9 9.793 9.820 9.793 9.820 dbcsr_mm_accdrv_process 20954 16.1 4.284 6.162 7.568 9.443 wfi_extrapolate 11 7.9 0.001 0.001 9.189 9.189 grid_integrate_task_list 129 12.3 8.620 8.803 8.620 8.803 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 8.157 8.157 density_rs2pw 129 9.7 0.005 0.006 8.077 8.124 mp_alltoall_d11v 2429 14.1 7.597 7.726 7.597 7.726 fft3d_ps 1301 14.7 2.872 2.880 7.100 7.162 calculate_dm_sparse 129 9.5 0.001 0.001 6.942 7.007 grid_collocate_task_list 129 9.7 6.466 6.537 6.466 6.537 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.457 6.499 copy_dbcsr_to_fm 187 11.8 0.004 0.004 6.357 6.429 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=301.739000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2908.545455, yerr=154.665248 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e4289968d5d1c81be94a335915a0cf308a6045a0_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 1.260204E+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 418231. 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 0.013 0.032 85.444 85.445 qs_energies 1 2.0 0.000 0.000 84.972 84.979 ls_scf 1 3.0 0.000 0.000 84.069 84.076 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.864 73.056 multiply_cannon 111 7.7 0.017 0.020 55.896 56.928 multiply_cannon_loop 111 8.7 0.226 0.240 52.486 53.660 ls_scf_main 1 4.0 0.000 0.000 52.622 52.623 density_matrix_trs4 2 5.0 0.002 0.003 47.055 47.119 ls_scf_init_scf 1 4.0 0.000 0.000 28.400 28.403 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.252 27.301 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.129 25.149 mp_waitall_1 11031 10.9 22.660 24.945 22.660 24.945 multiply_cannon_multrec 2664 9.7 8.151 8.945 15.648 17.435 multiply_cannon_sync_h2d 2664 9.7 13.548 15.031 13.548 15.031 make_m2s 222 7.7 0.009 0.012 13.369 13.892 make_images 222 8.7 0.099 0.108 13.347 13.872 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.659 12.933 make_images_data 222 9.7 0.004 0.005 7.935 8.570 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.398 8.297 dbcsr_mm_accdrv_process 4760 10.4 0.593 0.718 7.113 8.106 hybrid_alltoall_any 227 10.6 0.216 1.835 6.790 8.002 dbcsr_mm_accdrv_process_sort 4760 11.4 6.320 7.234 6.320 7.234 calculate_norms 4752 9.8 5.514 6.226 5.514 6.226 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.047 5.184 mp_sum_l 887 5.1 3.065 4.743 3.065 4.743 make_images_sizes 222 9.7 0.000 0.000 0.752 3.686 mp_alltoall_i44 222 10.7 0.752 3.686 0.752 3.686 multiply_cannon_metrocomm4 2442 9.7 0.012 0.014 2.060 3.499 mp_irecv_dv 6231 10.9 2.043 3.469 2.043 3.469 arnoldi_extremal 4 6.8 0.000 0.000 3.375 3.404 arnoldi_normal_ev 4 7.8 0.001 0.002 3.375 3.404 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.292 3.342 build_subspace 16 8.4 0.009 0.012 3.274 3.277 ls_scf_post 1 4.0 0.000 0.000 3.047 3.053 ls_scf_store_result 1 5.0 0.000 0.000 2.867 2.905 dbcsr_special_finalize 555 9.7 0.005 0.006 2.333 2.776 dbcsr_merge_single_wm 555 10.7 0.456 0.598 2.325 2.767 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.373 2.616 make_images_pack 222 9.7 2.202 2.589 2.204 2.591 dbcsr_sort_data 658 11.4 2.127 2.517 2.127 2.517 dbcsr_matrix_vector_mult_local 304 10.0 2.067 2.466 2.069 2.468 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.323 2.388 buffer_matrices_ensure_size 222 8.7 1.748 2.076 1.748 2.076 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.787 1.789 rebuild_ks_matrix 3 7.3 0.000 0.000 1.778 1.780 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.003 1.778 1.780 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.445000, 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/e4289968d5d1c81be94a335915a0cf308a6045a0_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 2.145268E+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 0.025 0.042 90.767 90.769 qs_energies 1 2.0 0.000 0.000 90.274 90.278 ls_scf 1 3.0 0.000 0.000 88.937 88.942 dbcsr_multiply_generic 111 6.7 0.015 0.016 74.766 75.117 multiply_cannon 111 7.7 0.028 0.041 53.016 57.478 ls_scf_main 1 4.0 0.000 0.000 55.130 55.137 multiply_cannon_loop 111 8.7 0.135 0.146 50.298 53.782 density_matrix_trs4 2 5.0 0.002 0.003 49.310 49.519 mp_waitall_1 9105 10.9 21.282 30.722 21.282 30.722 ls_scf_init_scf 1 4.0 0.000 0.000 30.047 30.048 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.872 28.950 multiply_cannon_multrec 1332 9.7 13.263 17.168 22.658 27.678 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.498 26.506 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.831 21.134 make_m2s 222 7.7 0.006 0.008 15.187 15.738 make_images 222 8.7 1.369 1.691 15.156 15.708 dbcsr_mm_accdrv_process 4041 10.4 0.331 0.528 8.992 10.589 dbcsr_mm_accdrv_process_sort 4041 11.4 8.513 10.061 8.513 10.061 make_images_data 222 9.7 0.004 0.005 8.901 9.699 hybrid_alltoall_any 227 10.6 0.541 2.547 8.292 9.419 mp_sum_l 887 5.1 5.100 8.820 5.100 8.820 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.201 7.733 mp_irecv_dv 3311 11.0 3.181 7.669 3.181 7.669 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.808 6.974 calculate_norms 2376 9.8 6.034 6.705 6.034 6.705 multiply_cannon_sync_h2d 1332 9.7 4.877 6.366 4.877 6.366 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.029 5.213 arnoldi_extremal 4 6.8 0.000 0.000 4.968 4.981 arnoldi_normal_ev 4 7.8 0.001 0.005 4.968 4.980 build_subspace 16 8.4 0.014 0.021 4.711 4.715 ls_scf_post 1 4.0 0.000 0.000 3.760 3.764 ls_scf_store_result 1 5.0 0.000 0.000 3.481 3.584 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.256 3.520 dbcsr_matrix_vector_mult_local 304 10.0 2.784 3.271 2.787 3.272 mp_allgather_i34 111 8.7 0.708 2.924 0.708 2.924 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.633 2.738 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.193 2.552 dbcsr_data_new 4174 10.1 2.115 2.418 2.115 2.418 make_images_pack 222 9.7 1.813 2.133 1.815 2.135 dbcsr_sort_data 436 11.2 1.827 2.052 1.827 2.052 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.864 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=90.769000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1773.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e4289968d5d1c81be94a335915a0cf308a6045a0_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 2.831106E+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 0.044 0.066 92.300 92.303 qs_energies 1 2.0 0.000 0.000 91.538 91.564 ls_scf 1 3.0 0.000 0.001 90.115 90.139 dbcsr_multiply_generic 111 6.7 0.016 0.016 74.517 74.865 multiply_cannon 111 7.7 0.030 0.072 51.878 56.444 ls_scf_main 1 4.0 0.000 0.003 55.965 55.968 multiply_cannon_loop 111 8.7 0.117 0.129 49.153 52.517 density_matrix_trs4 2 5.0 0.002 0.004 50.065 50.231 mp_waitall_1 7281 11.0 23.420 32.616 23.420 32.616 ls_scf_init_scf 1 4.0 0.000 0.002 30.419 30.422 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.158 29.280 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.796 26.810 multiply_cannon_multrec 888 9.7 12.617 15.269 21.292 24.593 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.870 21.736 make_m2s 222 7.7 0.006 0.007 16.182 16.911 make_images 222 8.7 1.583 1.852 16.144 16.871 make_images_data 222 9.7 0.004 0.005 9.531 10.622 hybrid_alltoall_any 227 10.6 0.641 2.936 9.176 10.270 dbcsr_mm_accdrv_process 3754 10.4 0.291 0.496 8.212 9.507 dbcsr_mm_accdrv_process_sort 3754 11.4 7.778 9.011 7.778 9.011 mp_sum_l 887 5.1 4.784 7.926 4.784 7.926 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.482 7.469 mp_irecv_dv 2335 11.1 2.467 7.423 2.467 7.423 multiply_cannon_sync_h2d 888 9.7 6.052 7.423 6.052 7.423 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.664 6.760 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.684 6.262 arnoldi_extremal 4 6.8 0.000 0.000 5.324 5.342 arnoldi_normal_ev 4 7.8 0.001 0.005 5.324 5.342 build_subspace 16 8.4 0.014 0.020 5.017 5.022 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.747 4.972 calculate_norms 1584 9.8 4.325 4.775 4.325 4.775 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.577 3.862 ls_scf_post 1 4.0 0.003 0.026 3.730 3.758 mp_allgather_i34 111 8.7 0.838 3.662 0.838 3.662 dbcsr_matrix_vector_mult_local 304 10.0 3.064 3.643 3.066 3.645 ls_scf_store_result 1 5.0 0.000 0.000 3.476 3.551 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.861 2.977 dbcsr_data_new 4116 9.9 2.113 2.461 2.113 2.461 make_images_sizes 222 9.7 0.000 0.000 0.807 2.111 mp_alltoall_i44 222 10.7 0.806 2.111 0.806 2.111 dbcsr_sort_data 325 11.1 1.890 2.083 1.890 2.083 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.909 1.911 rebuild_ks_matrix 3 7.3 0.000 0.000 1.891 1.893 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.891 1.893 make_images_pack 222 9.7 1.621 1.874 1.624 1.877 dbcsr_finalize 304 7.8 0.026 0.032 1.601 1.853 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=92.303000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2209.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e4289968d5d1c81be94a335915a0cf308a6045a0_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 3.326484E+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 0.081 97.324 97.326 qs_energies 1 2.0 0.000 0.000 96.613 96.625 ls_scf 1 3.0 0.000 0.001 94.955 94.969 dbcsr_multiply_generic 111 6.7 0.017 0.021 78.527 78.796 ls_scf_main 1 4.0 0.000 0.004 58.941 58.942 multiply_cannon 111 7.7 0.046 0.135 51.739 56.388 density_matrix_trs4 2 5.0 0.002 0.006 52.786 52.906 multiply_cannon_loop 111 8.7 0.153 0.172 46.655 50.021 ls_scf_init_scf 1 4.0 0.000 0.004 32.774 32.776 ls_scf_init_matrix_S 1 5.0 0.000 0.001 31.560 31.625 mp_waitall_1 6369 11.0 22.608 29.502 22.608 29.502 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.083 29.099 multiply_cannon_multrec 1332 9.7 14.165 17.358 22.181 25.470 make_m2s 222 7.7 0.007 0.008 21.166 22.558 make_images 222 8.7 3.140 3.608 21.116 22.510 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.112 17.045 make_images_data 222 9.7 0.004 0.005 11.805 13.597 hybrid_alltoall_any 227 10.6 0.799 3.795 11.102 12.854 dbcsr_mm_accdrv_process 3641 10.4 0.301 0.481 7.654 9.231 dbcsr_mm_accdrv_process_sort 3641 11.4 7.214 8.732 7.214 8.732 mp_sum_l 887 5.1 4.057 7.387 4.057 7.387 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.103 6.108 multiply_cannon_sync_h2d 1332 9.7 5.463 6.087 5.463 6.087 mp_irecv_dv 3229 10.9 2.078 6.054 2.078 6.054 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.521 5.800 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.107 5.783 arnoldi_extremal 4 6.8 0.000 0.000 5.362 5.376 arnoldi_normal_ev 4 7.8 0.001 0.004 5.362 5.376 build_subspace 16 8.4 0.014 0.021 5.026 5.034 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.559 4.759 mp_allgather_i34 111 8.7 2.188 4.531 2.188 4.531 calculate_norms 2376 9.8 4.169 4.514 4.169 4.514 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.715 3.991 dbcsr_matrix_vector_mult_local 304 10.0 3.250 3.754 3.253 3.756 dbcsr_sort_data 658 11.4 3.022 3.407 3.022 3.407 ls_scf_post 1 4.0 0.000 0.001 3.241 3.252 dbcsr_special_finalize 555 9.7 0.006 0.007 2.771 3.247 dbcsr_merge_single_wm 555 10.7 0.532 0.668 2.762 3.239 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.021 3.073 ls_scf_store_result 1 5.0 0.000 0.000 2.966 3.031 dbcsr_data_release 10477 10.7 1.586 2.415 1.586 2.415 dbcsr_finalize 304 7.8 0.049 0.061 1.794 1.978 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.326000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2730.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e4289968d5d1c81be94a335915a0cf308a6045a0_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 4.760695E+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 0.035 0.054 92.529 92.530 qs_energies 1 2.0 0.000 0.000 91.788 91.793 ls_scf 1 3.0 0.000 0.000 89.855 89.859 dbcsr_multiply_generic 111 6.7 0.017 0.019 71.110 71.315 ls_scf_main 1 4.0 0.000 0.000 56.948 56.948 multiply_cannon 111 7.7 0.081 0.160 52.792 56.407 multiply_cannon_loop 111 8.7 0.090 0.097 50.198 51.638 density_matrix_trs4 2 5.0 0.002 0.003 49.912 49.972 ls_scf_init_scf 1 4.0 0.000 0.000 29.495 29.498 mp_waitall_1 5436 11.0 24.719 28.889 24.719 28.889 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.255 28.277 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.144 26.153 multiply_cannon_multrec 444 9.7 13.675 16.319 20.856 23.327 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.669 15.701 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.457 15.263 make_m2s 222 7.7 0.005 0.005 13.687 14.620 make_images 222 8.7 2.039 2.474 13.620 14.552 hybrid_alltoall_any 227 10.6 0.800 3.825 8.215 9.860 make_images_data 222 9.7 0.003 0.004 8.407 9.705 multiply_cannon_sync_h2d 444 9.7 6.716 8.103 6.716 8.103 dbcsr_mm_accdrv_process 3003 10.4 0.329 0.403 6.879 7.995 dbcsr_mm_accdrv_process_sort 3003 11.4 6.516 7.597 6.516 7.597 arnoldi_extremal 4 6.8 0.000 0.000 6.082 6.091 arnoldi_normal_ev 4 7.8 0.001 0.004 6.082 6.091 build_subspace 16 8.4 0.015 0.019 5.683 5.692 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.441 4.652 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.334 4.588 mp_sum_l 887 5.1 2.627 4.331 2.627 4.331 dbcsr_matrix_vector_mult_local 304 10.0 3.778 4.257 3.780 4.259 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.530 3.839 mp_irecv_dv 1241 11.2 1.517 3.813 1.517 3.813 calculate_norms 792 9.8 3.625 3.792 3.625 3.792 mp_allgather_i34 111 8.7 1.174 3.708 1.174 3.708 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.583 3.661 ls_scf_post 1 4.0 0.000 0.000 3.412 3.417 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.834 3.409 make_images_sizes 222 9.7 0.000 0.000 0.897 3.403 mp_alltoall_i44 222 10.7 0.897 3.403 0.897 3.403 ls_scf_store_result 1 5.0 0.000 0.000 3.206 3.239 dbcsr_finalize 304 7.8 0.062 0.077 2.195 2.277 dbcsr_data_new 4608 9.7 1.785 2.226 1.785 2.226 dbcsr_merge_all 275 8.9 0.481 0.532 2.053 2.108 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.046 2.047 rebuild_ks_matrix 3 7.3 0.000 0.000 2.013 2.014 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.013 2.014 qs_energies_init_hamiltonians 1 3.0 0.001 0.004 1.918 1.918 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=92.530000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3804.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e4289968d5d1c81be94a335915a0cf308a6045a0_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 8.857453E+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 0.071 0.091 106.301 106.303 qs_energies 1 2.0 0.000 0.000 104.962 104.967 ls_scf 1 3.0 0.000 0.000 102.052 102.056 dbcsr_multiply_generic 111 6.7 0.024 0.026 75.426 75.569 ls_scf_main 1 4.0 0.000 0.000 64.430 64.431 density_matrix_trs4 2 5.0 0.002 0.003 55.428 55.496 multiply_cannon 111 7.7 0.129 0.195 49.041 51.211 multiply_cannon_loop 111 8.7 0.098 0.100 46.039 46.432 ls_scf_init_scf 1 4.0 0.000 0.000 33.815 33.815 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.309 32.330 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.579 29.586 mp_waitall_1 4527 11.1 22.541 26.256 22.541 26.256 make_m2s 222 7.7 0.005 0.005 22.779 23.872 make_images 222 8.7 3.568 3.877 22.672 23.763 multiply_cannon_multrec 444 9.7 17.826 18.389 22.537 23.052 hybrid_alltoall_any 227 10.6 1.652 3.608 12.899 15.866 make_images_data 222 9.7 0.003 0.004 13.165 15.301 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.777 11.348 multiply_cannon_sync_h2d 444 9.7 8.793 8.845 8.793 8.845 arnoldi_extremal 4 6.8 0.000 0.000 7.677 7.686 arnoldi_normal_ev 4 7.8 0.002 0.008 7.677 7.685 build_subspace 16 8.4 0.026 0.036 7.116 7.126 dbcsr_matrix_vector_mult 304 9.0 0.017 0.033 5.714 5.865 dbcsr_matrix_vector_mult_local 304 10.0 5.163 5.469 5.165 5.471 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.097 5.188 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.914 5.171 dbcsr_mm_accdrv_process 1814 10.4 0.289 0.351 4.523 4.667 dbcsr_mm_accdrv_process_sort 1814 11.4 4.188 4.338 4.188 4.338 ls_scf_post 1 4.0 0.000 0.000 3.807 3.812 make_images_sizes 222 9.7 0.000 0.000 1.478 3.745 mp_alltoall_i44 222 10.7 1.478 3.745 1.478 3.745 mp_allgather_i34 111 8.7 1.107 3.628 1.107 3.628 ls_scf_store_result 1 5.0 0.000 0.000 3.527 3.561 calculate_norms 792 9.8 3.227 3.275 3.227 3.275 dbcsr_finalize 304 7.8 0.082 0.089 3.073 3.180 dbcsr_merge_all 275 8.9 0.883 0.917 2.855 2.956 dbcsr_complete_redistribute 5 7.6 1.438 1.480 2.780 2.899 qs_energies_init_hamiltonians 1 3.0 0.002 0.003 2.880 2.880 matrix_ls_to_qs 2 6.0 0.000 0.000 2.399 2.512 dbcsr_sort_data 325 11.1 2.438 2.509 2.438 2.509 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.393 2.395 dbcsr_data_new 6591 9.6 1.884 2.381 1.884 2.381 dbcsr_new_transposed 4 7.5 0.242 0.253 2.331 2.343 rebuild_ks_matrix 3 7.3 0.000 0.000 2.327 2.329 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.327 2.329 dbcsr_frobenius_norm 74 6.6 2.055 2.131 2.183 2.217 dbcsr_add_d 103 6.2 0.000 0.000 2.135 2.217 dbcsr_add_anytype 103 7.2 0.858 0.889 2.134 2.216 dbcsr_data_release 12724 10.6 1.984 2.208 1.984 2.208 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=106.303000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=7019.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e4289968d5d1c81be94a335915a0cf308a6045a0_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 588.718080E+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 1077377855. 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 0.043 0.097 313.369 313.371 qs_mol_dyn_low 1 2.0 0.003 0.014 311.961 311.987 qs_forces 5 3.8 0.005 0.057 311.822 311.824 qs_energies 5 4.8 0.003 0.036 308.105 308.130 scf_env_do_scf 5 5.8 0.000 0.009 291.039 291.041 scf_env_do_scf_inner_loop 105 6.6 0.003 0.016 248.127 248.141 qs_scf_new_mos 105 7.6 0.000 0.001 192.342 192.547 qs_scf_loop_do_ot 105 8.6 0.001 0.002 192.342 192.546 ot_scf_mini 105 9.6 0.003 0.011 181.798 182.057 dbcsr_multiply_generic 1445 12.2 0.127 0.134 144.728 145.444 velocity_verlet 4 3.0 0.005 0.043 130.788 130.791 multiply_cannon 1445 13.2 0.278 0.289 119.492 122.287 multiply_cannon_loop 1445 14.2 2.837 2.998 116.101 117.800 qs_ot_get_p 112 10.4 0.001 0.003 83.306 83.681 qs_ot_p2m_diag 40 11.0 0.020 0.030 69.800 69.909 ot_mini 105 10.6 0.001 0.005 66.598 66.735 cp_dbcsr_syevd 40 12.0 0.002 0.002 65.892 65.896 cp_fm_syevd 40 13.0 0.000 0.000 59.139 59.285 cp_fm_redistribute_end 40 14.0 26.701 53.237 26.711 53.245 cp_fm_syevd_base 40 14.0 26.510 53.054 26.510 53.054 mp_waitall_1 488190 16.1 43.029 50.430 43.029 50.430 qs_ot_get_derivative 55 11.6 0.001 0.001 44.467 44.630 multiply_cannon_multrec 69360 15.2 29.255 34.473 39.297 44.442 init_scf_loop 7 6.6 0.000 0.005 42.856 42.861 prepare_preconditioner 7 7.6 0.000 0.000 36.939 36.972 make_preconditioner 7 8.6 0.004 0.132 36.939 36.972 multiply_cannon_metrocomm3 69360 15.2 0.205 0.215 28.209 36.940 rebuild_ks_matrix 110 8.4 0.000 0.000 35.774 35.952 qs_ks_build_kohn_sham_matrix 110 9.4 0.012 0.019 35.774 35.951 qs_ks_update_qs_env 112 7.6 0.001 0.001 32.843 33.002 multiply_cannon_sync_h2d 69360 15.2 28.580 32.518 28.580 32.518 qs_rho_update_rho_low 110 7.6 0.001 0.002 27.335 27.738 calculate_rho_elec 110 8.6 0.030 0.033 27.334 27.738 make_full_inverse_cholesky 7 9.6 0.000 0.000 27.327 27.419 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 23.356 23.611 apply_single 62 13.6 0.000 0.000 23.356 23.610 ot_new_cg_direction 55 11.6 0.001 0.028 21.365 21.368 density_rs2pw 110 9.6 0.005 0.007 20.399 20.828 cp_fm_cholesky_invert 7 10.6 18.499 18.526 18.499 18.526 transfer_rs2pw 445 10.6 0.007 0.008 16.073 16.534 qs_ot_get_orbitals 105 10.6 0.001 0.001 16.190 16.454 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 15.839 15.931 pw_transfer 1645 12.4 0.079 0.101 14.856 15.124 fft_wrap_pw1pw2 1425 13.5 0.012 0.015 14.718 14.987 mp_sum_l 4764 12.2 13.865 14.820 13.865 14.820 init_scf_run 5 5.8 0.000 0.001 14.064 14.065 scf_env_initial_rho_setup 5 6.8 0.002 0.003 14.064 14.065 sum_up_and_integrate 60 10.3 0.001 0.003 13.163 13.178 integrate_v_rspace 60 11.3 0.002 0.003 13.146 13.161 fft_wrap_pw1pw2_240 915 15.0 1.130 1.221 12.618 12.887 qs_ot_get_derivative_diag 18 12.0 0.000 0.000 12.545 12.645 calculate_dm_sparse 110 9.5 0.000 0.001 12.417 12.639 qs_vxc_create 110 10.4 0.002 0.007 12.146 12.211 check_diag 80 13.5 8.688 8.951 11.749 11.921 make_m2s 2890 13.2 0.079 0.085 10.930 11.563 make_images 2890 14.2 0.238 0.258 10.823 11.455 fft3d_pb 915 16.0 2.359 2.644 10.845 11.206 dbcsr_mm_accdrv_process 154766 15.8 6.125 6.330 9.913 10.714 multiply_cannon_metrocomm1 69360 15.2 0.095 0.102 5.869 10.067 mp_sendrecv_dv 168740 12.6 9.817 10.004 9.817 10.004 transfer_rs2pw_30 110 11.6 1.263 1.320 9.121 9.868 make_full_single_inverse 7 9.6 0.001 0.004 9.126 9.159 acc_transpose_blocks 69360 15.2 0.351 0.368 8.167 8.883 potential_pw2rs 60 12.3 0.003 0.003 8.673 8.742 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.001 8.538 8.555 xc_rho_set_and_dset_create 110 12.4 0.077 0.097 8.251 8.507 mp_alltoall_z22v 2340 17.7 8.039 8.402 8.039 8.402 xc_vxc_pw_create 60 11.3 0.038 0.049 8.320 8.381 calculate_first_density_matrix 1 7.0 0.000 0.004 8.252 8.267 cp_fm_cholesky_decompose 14 10.2 7.943 7.953 7.943 7.953 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 7.861 7.916 xc_pw_derive 510 13.4 0.005 0.006 7.672 7.741 transfer_pw2rs 245 13.2 0.003 0.004 7.266 7.325 mp_alltoall_d11v 1300 13.8 6.231 7.091 6.231 7.091 make_images_sizes 2890 15.2 0.004 0.004 5.457 6.438 mp_alltoall_i44 2890 16.2 5.453 6.434 5.453 6.434 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=313.371000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=560.600000, yerr=1.959592 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: e4289968d5d1c81be94a335915a0cf308a6045a0 Summary: empty Status: OK