=== 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: b3f583d97791bedccf073d04f637ff05a010c441 ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, Cray-FFTW 3.3.8.10, # COSMA 2.6.2, ELPA 2022.11.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.0.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.1, # SIRIUS 7.3.2, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Author: Matthias Krack (26.12.2022) # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.0.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.1 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.3.2 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta_prefixed_scalapack.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b3f583d97791bedccf073d04f637ff05a010c441_performance_tests/01 job id: 44145710 --- 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/b3f583d97791bedccf073d04f637ff05a010c441_performance_tests/02 job id: 44145711 --- 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/b3f583d97791bedccf073d04f637ff05a010c441_performance_tests/03 job id: 44145712 --- 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/b3f583d97791bedccf073d04f637ff05a010c441_performance_tests/04 job id: 44145714 --- 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/b3f583d97791bedccf073d04f637ff05a010c441_performance_tests/05 job id: 44145717 --- 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/b3f583d97791bedccf073d04f637ff05a010c441_performance_tests/06 job id: 44145719 --- 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/b3f583d97791bedccf073d04f637ff05a010c441_performance_tests/07 job id: 44145720 --- 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/b3f583d97791bedccf073d04f637ff05a010c441_performance_tests/08 job id: 44145721 --- 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/b3f583d97791bedccf073d04f637ff05a010c441_performance_tests/09 job id: 44145722 --- 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/b3f583d97791bedccf073d04f637ff05a010c441_performance_tests/10 job id: 44145723 --- 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/b3f583d97791bedccf073d04f637ff05a010c441_performance_tests/11 job id: 44145724 --- 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/b3f583d97791bedccf073d04f637ff05a010c441_performance_tests/12 job id: 44145725 --- 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/b3f583d97791bedccf073d04f637ff05a010c441_performance_tests/13 job id: 44145726 --- 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/b3f583d97791bedccf073d04f637ff05a010c441_performance_tests/14 job id: 44145727 --- 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/b3f583d97791bedccf073d04f637ff05a010c441_performance_tests/15 job id: 44145728 --- 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/b3f583d97791bedccf073d04f637ff05a010c441_performance_tests/16 job id: 44145729 --- 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/b3f583d97791bedccf073d04f637ff05a010c441_performance_tests/17 job id: 44145730 --- 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/b3f583d97791bedccf073d04f637ff05a010c441_performance_tests/18 job id: 44145731 --- 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/b3f583d97791bedccf073d04f637ff05a010c441_performance_tests/19 job id: 44145732 --- 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/b3f583d97791bedccf073d04f637ff05a010c441_performance_tests/20 job id: 44145733 --- 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/b3f583d97791bedccf073d04f637ff05a010c441_performance_tests/21 job id: 44145734 --- 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/b3f583d97791bedccf073d04f637ff05a010c441_performance_tests/22 job id: 44145735 --- 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/b3f583d97791bedccf073d04f637ff05a010c441_performance_tests/23 job id: 44145736 --- 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/b3f583d97791bedccf073d04f637ff05a010c441_performance_tests/24 job id: 44145737 --- 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/b3f583d97791bedccf073d04f637ff05a010c441_performance_tests/25 job id: 44145738 --- 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/b3f583d97791bedccf073d04f637ff05a010c441_performance_tests/26 job id: 44145740 --- Point --- name: 510 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/1r/12t) --- Point --- name: 511 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ === END TESTS (description) === ===== PLOTS (description) ===== ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2_mem", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md_mem", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md_mem", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md_mem", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls_mem", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" === END PLOTS (description) === ============ RESULTS ============ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b3f583d97791bedccf073d04f637ff05a010c441_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 344 9. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.034 134.226 134.226 farming_run 1 2.0 133.400 133.406 134.199 134.202 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.457975E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 228 1113141. MP_Allreduce 485 2282278. MP_Sync 27 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 8 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.041 115.130 115.131 qs_energies 1 2.0 0.000 0.000 114.913 114.915 mp2_main 1 3.0 0.000 0.000 112.999 113.000 mp2_gpw_main 1 4.0 0.019 0.026 112.168 112.170 mp2_ri_gpw_compute_in 1 5.0 0.171 0.174 93.175 93.672 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 55.362 55.860 mp2_eri_3c_integrate_gpw 272 7.0 0.154 0.169 41.680 47.128 get_2c_integrals 1 6.0 0.000 0.001 37.113 37.640 integrate_v_rspace 273 8.0 0.437 0.454 25.072 30.203 pw_transfer 6555 10.6 0.370 0.383 27.335 27.978 fft_wrap_pw1pw2 5465 11.4 0.043 0.046 26.014 26.638 grid_integrate_task_list 273 9.0 20.900 26.495 20.900 26.495 fft_wrap_pw1pw2_100 2178 12.4 1.172 1.240 23.537 24.152 compute_2c_integrals 1 7.0 0.006 0.008 19.351 19.351 rpa_ri_compute_en 1 5.0 0.001 0.001 18.887 19.068 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.842 18.950 mp2_eri_2c_integrate_gpw 1 9.0 2.389 2.422 18.839 18.949 cp_fm_cholesky_decompose 12 8.2 17.718 18.247 17.718 18.247 cholesky_decomp 1 7.0 0.000 0.000 16.614 17.141 fft3d_s 5443 13.4 16.122 16.537 16.143 16.557 ao_to_mo_and_store_B_mult_1 272 7.0 10.844 15.583 10.844 15.583 calculate_wavefunction 272 8.0 5.433 5.586 12.535 13.149 rpa_num_int 1 6.0 0.000 0.001 10.594 10.594 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.550 10.591 calc_mat_Q 8 8.0 0.000 0.000 9.394 9.503 contract_S_to_Q 8 9.0 0.000 0.000 8.815 8.929 calc_potential_gpw 544 9.5 0.005 0.006 8.172 8.592 parallel_gemm_fm 14 9.1 0.000 0.000 8.401 8.474 parallel_gemm_fm_cosma 14 10.1 8.401 8.474 8.401 8.474 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.185 8.421 potential_pw2rs 545 10.0 0.107 0.109 7.718 8.375 collocate_single_gaussian 272 10.0 0.041 0.042 7.452 7.712 create_integ_mat 1 6.0 0.022 0.027 7.699 7.699 array2fm 1 7.0 0.000 0.000 6.732 7.148 pw_scatter_s 2720 13.7 4.415 4.632 4.415 4.632 pw_gather_s 2722 13.2 3.893 4.267 3.893 4.267 array2fm_buffer_send 1 8.0 3.018 3.133 3.018 3.133 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.167961, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2729.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b3f583d97791bedccf073d04f637ff05a010c441_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 205321. MP_Allreduce 344 10. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.026 0.038 394.354 394.355 farming_run 1 2.0 393.541 393.545 394.320 394.322 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.227436E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 703 408373. MP_Allreduce 1821 23730. MP_Sync 38 MP_Alltoall 77 2368424. MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.042 208.641 208.642 qs_energies 1 2.0 0.000 0.000 208.354 208.362 scf_env_do_scf 1 3.0 0.000 0.000 105.880 105.880 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.019 105.026 rebuild_ks_matrix 4 6.0 0.000 0.000 105.018 105.025 qs_ks_build_kohn_sham_matrix 4 7.0 0.054 0.061 105.017 105.025 hfx_ks_matrix 4 8.0 0.001 0.001 104.641 104.645 integrate_four_center 4 9.0 0.143 0.454 104.641 104.644 mp2_main 1 3.0 0.000 0.000 102.182 102.189 mp2_gpw_main 1 4.0 0.031 0.048 101.360 101.369 integrate_four_center_main 4 10.0 0.122 0.575 96.845 98.938 integrate_four_center_bin 264 11.0 96.723 98.758 96.723 98.758 init_scf_loop 1 4.0 0.000 0.000 91.736 91.736 mp2_ri_gpw_compute_in 1 5.0 0.063 0.064 74.918 75.950 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.521 55.553 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.160 42.280 47.002 integrate_v_rspace 95 8.0 0.397 0.565 28.642 33.229 pw_transfer 2240 10.6 0.144 0.160 29.900 30.261 fft_wrap_pw1pw2 1868 11.4 0.017 0.019 28.916 29.356 ao_to_mo_and_store_B_mult_1 91 7.0 10.553 29.352 10.553 29.352 grid_integrate_task_list 95 9.0 23.937 28.764 23.937 28.764 mp2_ri_gpw_compute_en 1 5.0 0.054 0.063 26.294 27.931 fft_wrap_pw1pw2_100 730 12.4 1.272 1.460 26.631 27.157 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.853 1.901 24.611 24.622 get_2c_integrals 1 6.0 0.000 0.000 20.303 20.332 compute_2c_integrals 1 7.0 0.002 0.003 19.285 19.291 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 18.818 19.162 mp2_eri_2c_integrate_gpw 1 9.0 1.736 1.835 18.817 19.160 fft3d_s 1823 13.4 18.426 18.870 18.439 18.884 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.143 14.143 calculate_wavefunction 91 8.0 2.022 2.050 9.751 10.029 potential_pw2rs 186 10.0 0.034 0.035 8.640 9.160 mp2_ri_gpw_compute_en_expansio 172 7.0 0.556 0.578 8.730 9.051 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.216 8.606 local_gemm 172 8.0 8.175 8.480 8.175 8.480 collocate_single_gaussian 91 10.0 0.017 0.020 7.869 8.167 mp2_ri_gpw_compute_en_comm 22 7.0 0.510 0.529 7.639 8.135 calc_potential_gpw 182 9.5 0.002 0.002 7.878 8.060 mp2_ri_gpw_compute_en_ener 172 7.0 6.343 6.441 6.343 6.441 mp_sendrecv_dm3 2068 8.0 5.647 6.147 5.647 6.147 pw_gather_s 912 13.2 4.927 5.286 4.927 5.286 mp_sync 38 10.4 2.506 5.128 2.506 5.128 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.360287, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1514.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b3f583d97791bedccf073d04f637ff05a010c441_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 452.444160E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62385. MP_Allreduce 10249 271. MP_Sync 580 MP_Alltoall 2083 579374. 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 0.011 0.031 51.860 51.861 qs_mol_dyn_low 1 2.0 0.003 0.004 51.638 51.667 qs_forces 11 3.9 0.002 0.002 51.576 51.577 qs_energies 11 4.9 0.001 0.002 50.126 50.137 scf_env_do_scf 11 5.9 0.000 0.001 43.332 43.332 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 41.353 41.354 dbcsr_multiply_generic 2286 12.5 0.094 0.097 33.725 34.076 qs_scf_new_mos 108 7.5 0.000 0.001 31.333 31.634 qs_scf_loop_do_ot 108 8.5 0.000 0.001 31.332 31.634 ot_scf_mini 108 9.5 0.002 0.002 29.714 29.923 multiply_cannon 2286 13.5 0.180 0.185 26.630 28.028 multiply_cannon_loop 2286 14.5 1.486 1.560 25.977 27.408 velocity_verlet 10 3.0 0.001 0.001 24.843 24.844 ot_mini 108 10.5 0.001 0.001 18.875 19.132 qs_ot_get_derivative 108 11.5 0.001 0.005 15.950 16.145 mp_waitall_1 245248 16.5 8.266 15.184 8.266 15.184 multiply_cannon_metrocomm3 54864 15.5 0.068 0.073 5.985 13.249 multiply_cannon_multrec 54864 15.5 4.252 6.573 8.320 11.978 rebuild_ks_matrix 119 8.3 0.000 0.000 7.828 7.954 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 7.827 7.953 multiply_cannon_sync_h2d 54864 15.5 5.949 7.361 5.949 7.361 mp_sum_l 7207 12.9 5.224 7.056 5.224 7.056 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.901 7.016 qs_ot_get_p 119 10.4 0.001 0.001 6.289 6.620 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.516 5.984 dbcsr_mm_accdrv_process 76910 16.1 1.175 1.802 3.991 5.763 init_scf_run 11 5.9 0.000 0.001 5.516 5.516 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.516 5.516 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.193 5.307 sum_up_and_integrate 119 10.3 0.012 0.014 4.455 4.461 integrate_v_rspace 119 11.3 0.002 0.002 4.443 4.450 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.132 4.234 calculate_rho_elec 119 8.7 0.011 0.017 4.131 4.233 jit_kernel_multiply 13 15.8 2.753 3.933 2.753 3.933 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.696 3.754 calculate_first_density_matrix 1 7.0 0.000 0.000 3.401 3.408 calculate_dm_sparse 119 9.5 0.000 0.001 3.226 3.407 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.913 3.099 apply_single 119 13.6 0.000 0.000 2.912 3.099 cp_dbcsr_syevd 50 12.0 0.002 0.003 2.832 2.833 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.787 2.789 rs_pw_transfer 974 11.9 0.012 0.013 2.660 2.773 multiply_cannon_metrocomm1 54864 15.5 0.053 0.059 1.663 2.758 ot_diis_step 108 11.5 0.006 0.006 2.657 2.658 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.643 2.643 cp_fm_redistribute_end 50 14.0 2.408 2.624 2.413 2.625 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.552 2.599 cp_fm_diag_elpa_base 50 14.0 0.210 2.537 0.211 2.552 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.355 2.437 acc_transpose_blocks 54864 15.5 0.233 0.249 1.868 2.348 density_rs2pw 119 9.7 0.004 0.004 2.184 2.332 grid_integrate_task_list 119 12.3 2.018 2.113 2.018 2.113 wfi_extrapolate 11 7.9 0.001 0.001 2.051 2.051 init_scf_loop 11 6.9 0.000 0.000 1.963 1.963 mp_sum_d 4125 12.0 1.265 1.865 1.265 1.865 potential_pw2rs 119 12.3 0.004 0.004 1.818 1.829 pw_transfer 1439 11.6 0.053 0.059 1.751 1.821 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.001 1.763 1.780 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.674 1.749 make_m2s 4572 13.5 0.054 0.057 1.554 1.606 make_images 4572 14.5 0.133 0.139 1.471 1.521 fft3d_ps 1201 14.6 0.370 0.473 1.445 1.518 fft_wrap_pw1pw2_140 487 13.2 0.082 0.094 1.292 1.368 grid_collocate_task_list 119 9.7 1.289 1.352 1.289 1.352 mp_alltoall_d11v 2130 13.8 1.203 1.343 1.203 1.343 mp_waitany 12084 13.8 1.236 1.309 1.236 1.309 acc_transpose_blocks_kernels 54864 16.5 0.251 0.383 0.873 1.209 dbcsr_dot_sd 1205 11.9 0.051 0.063 0.698 1.094 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=51.861000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=431.272727, yerr=0.962091 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b3f583d97791bedccf073d04f637ff05a010c441_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 488.382464E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2194560 MPI messages size (bytes): total size 310.646604E+09 min size 0.000000E+00 max size 1.145520E+06 average size 141.553031E+03 MPI breakdown and total messages size (bytes): size <= 128 724648 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 281952 4619501568 32768 < size <= 131072 494448 39143342080 131072 < size <= 4194304 440000 264807943488 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62664. MP_Allreduce 10226 305. MP_Sync 104 MP_Alltoall 2060 1490547. 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 0.017 0.030 39.751 39.753 qs_mol_dyn_low 1 2.0 0.003 0.004 39.500 39.507 qs_forces 11 3.9 0.002 0.002 39.435 39.436 qs_energies 11 4.9 0.002 0.005 37.750 37.754 scf_env_do_scf 11 5.9 0.001 0.002 31.806 31.807 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 29.257 29.258 dbcsr_multiply_generic 2286 12.5 0.101 0.105 22.374 22.746 qs_scf_new_mos 108 7.5 0.001 0.001 20.091 20.338 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.090 20.337 ot_scf_mini 108 9.5 0.002 0.003 19.207 19.378 multiply_cannon 2286 13.5 0.207 0.219 17.345 19.000 velocity_verlet 10 3.0 0.001 0.001 18.125 18.126 multiply_cannon_loop 2286 14.5 0.896 0.977 16.281 17.611 ot_mini 108 10.5 0.001 0.001 12.054 12.288 mp_waitall_1 200699 16.5 6.130 11.844 6.130 11.844 multiply_cannon_metrocomm3 27432 15.5 0.069 0.071 4.626 10.242 qs_ot_get_derivative 108 11.5 0.001 0.001 9.599 9.773 multiply_cannon_multrec 27432 15.5 1.967 4.536 6.363 9.702 rebuild_ks_matrix 119 8.3 0.000 0.000 7.250 7.390 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.015 7.250 7.389 dbcsr_mm_accdrv_process 47894 16.0 3.335 5.876 4.328 6.593 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.408 6.535 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.870 4.748 init_scf_run 11 5.9 0.000 0.001 4.711 4.711 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.711 4.711 qs_ot_get_p 119 10.4 0.001 0.002 4.304 4.539 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.138 4.197 apply_single 119 13.6 0.000 0.000 3.138 4.197 sum_up_and_integrate 119 10.3 0.025 0.028 4.184 4.191 integrate_v_rspace 119 11.3 0.002 0.002 4.160 4.169 mp_sum_l 7207 12.9 2.129 4.049 2.129 4.049 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.790 3.832 calculate_rho_elec 119 8.7 0.021 0.024 3.790 3.831 calculate_first_density_matrix 1 7.0 0.000 0.002 3.223 3.225 rs_pw_transfer 974 11.9 0.010 0.011 2.566 2.958 qs_ot_p2m_diag 50 11.0 0.008 0.012 2.896 2.915 multiply_cannon_sync_h2d 27432 15.5 2.172 2.826 2.172 2.826 make_m2s 4572 13.5 0.053 0.054 2.446 2.665 make_images 4572 14.5 0.203 0.236 2.356 2.574 init_scf_loop 11 6.9 0.001 0.004 2.530 2.532 density_rs2pw 119 9.7 0.004 0.004 2.120 2.525 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.478 2.481 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.471 2.471 jit_kernel_multiply 9 16.1 0.941 2.452 0.941 2.452 ot_diis_step 108 11.5 0.011 0.011 2.405 2.406 calculate_dm_sparse 119 9.5 0.000 0.001 2.316 2.392 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.283 2.323 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.095 2.187 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.155 2.156 cp_fm_redistribute_end 50 14.0 1.786 2.134 1.789 2.135 cp_fm_diag_elpa_base 50 14.0 0.330 2.015 0.344 2.087 pw_transfer 1439 11.6 0.065 0.071 1.978 2.015 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.887 1.927 potential_pw2rs 119 12.3 0.006 0.006 1.911 1.924 grid_integrate_task_list 119 12.3 1.832 1.919 1.832 1.919 acc_transpose_blocks 27432 15.5 0.108 0.113 1.299 1.780 fft3d_ps 1201 14.6 0.515 0.567 1.589 1.626 prepare_preconditioner 11 7.9 0.000 0.000 1.556 1.583 make_preconditioner 11 8.9 0.000 0.002 1.556 1.583 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.544 1.552 fft_wrap_pw1pw2_140 487 13.2 0.079 0.086 1.494 1.534 make_images_data 4572 15.5 0.046 0.053 1.120 1.514 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.449 1.507 hybrid_alltoall_any 4725 16.4 0.052 0.113 0.981 1.449 wfi_extrapolate 11 7.9 0.001 0.001 1.431 1.431 mp_alltoall_d11v 2130 13.8 1.174 1.380 1.174 1.380 grid_collocate_task_list 119 9.7 1.228 1.368 1.228 1.368 mp_allgather_i34 2286 14.5 0.507 1.260 0.507 1.260 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.149 1.195 acc_transpose_blocks_kernels 27432 16.5 0.181 0.270 0.766 1.178 mp_sum_d 4125 12.0 0.637 1.121 0.637 1.121 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 0.963 0.964 rs_pw_transfer_RS2PW_140 130 11.5 0.140 0.148 0.556 0.951 mp_waitany 5720 13.7 0.527 0.919 0.527 0.919 jit_kernel_transpose 5 15.5 0.585 0.918 0.585 0.918 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.901 0.915 multiply_cannon_metrocomm1 27432 15.5 0.032 0.035 0.347 0.876 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=39.753000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=465.272727, yerr=1.600620 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b3f583d97791bedccf073d04f637ff05a010c441_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 525.484032E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62660. MP_Allreduce 10225 303. MP_Sync 104 MP_Alltoall 1821 2482779. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_comm_split 50 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.030 32.902 32.903 qs_mol_dyn_low 1 2.0 0.003 0.004 32.676 32.685 qs_forces 11 3.9 0.001 0.002 32.619 32.619 qs_energies 11 4.9 0.002 0.004 31.041 31.043 scf_env_do_scf 11 5.9 0.001 0.002 25.455 25.455 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 22.869 22.870 dbcsr_multiply_generic 2286 12.5 0.095 0.098 17.098 17.195 velocity_verlet 10 3.0 0.001 0.002 15.217 15.218 multiply_cannon 2286 13.5 0.195 0.205 13.883 14.735 qs_scf_new_mos 108 7.5 0.001 0.001 14.634 14.658 qs_scf_loop_do_ot 108 8.5 0.001 0.001 14.633 14.657 multiply_cannon_loop 2286 14.5 0.633 0.663 13.095 13.976 ot_scf_mini 108 9.5 0.002 0.003 13.913 13.929 ot_mini 108 10.5 0.001 0.001 8.613 8.629 multiply_cannon_multrec 18288 15.5 1.952 2.842 6.972 7.940 qs_ot_get_derivative 108 11.5 0.001 0.001 7.114 7.129 rebuild_ks_matrix 119 8.3 0.000 0.001 6.426 6.442 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.426 6.441 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.656 5.670 dbcsr_mm_accdrv_process 38222 16.0 4.158 5.434 4.936 5.577 mp_waitall_1 158411 16.6 3.252 4.570 3.252 4.570 init_scf_run 11 5.9 0.000 0.001 4.420 4.420 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.420 4.420 sum_up_and_integrate 119 10.3 0.030 0.031 3.980 3.988 integrate_v_rspace 119 11.3 0.002 0.004 3.949 3.960 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.472 3.478 calculate_rho_elec 119 8.7 0.030 0.031 3.471 3.477 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.660 3.278 multiply_cannon_metrocomm3 18288 15.5 0.045 0.046 1.770 3.243 calculate_first_density_matrix 1 7.0 0.000 0.002 3.215 3.217 qs_ot_get_p 119 10.4 0.001 0.001 3.158 3.179 calculate_dm_sparse 119 9.5 0.000 0.000 2.709 2.721 init_scf_loop 11 6.9 0.001 0.003 2.569 2.574 rs_pw_transfer 974 11.9 0.009 0.010 2.260 2.518 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.037 2.352 apply_single 119 13.6 0.000 0.000 2.037 2.352 density_rs2pw 119 9.7 0.004 0.004 1.994 2.252 jit_kernel_multiply 11 16.1 0.727 2.160 0.727 2.160 qs_ot_p2m_diag 50 11.0 0.012 0.012 2.112 2.117 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 2.083 2.088 pw_transfer 1439 11.6 0.066 0.072 1.968 1.982 make_m2s 4572 13.5 0.045 0.045 1.754 1.908 grid_integrate_task_list 119 12.3 1.801 1.902 1.801 1.902 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.874 1.890 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.851 1.851 make_images 4572 14.5 0.190 0.201 1.668 1.820 prepare_preconditioner 11 7.9 0.000 0.000 1.763 1.767 make_preconditioner 11 8.9 0.002 0.008 1.763 1.767 potential_pw2rs 119 12.3 0.007 0.008 1.749 1.756 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.710 1.711 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.614 1.700 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.621 1.622 cp_fm_redistribute_end 50 14.0 1.210 1.602 1.211 1.602 fft3d_ps 1201 14.6 0.524 0.541 1.553 1.568 cp_fm_diag_elpa_base 50 14.0 0.374 1.523 0.389 1.564 multiply_cannon_sync_h2d 18288 15.5 1.398 1.562 1.398 1.562 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.549 1.555 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.539 1.547 fft_wrap_pw1pw2_140 487 13.2 0.090 0.094 1.501 1.517 ot_diis_step 108 11.5 0.011 0.011 1.479 1.479 mp_sum_l 7207 12.9 1.069 1.380 1.069 1.380 multiply_cannon_metrocomm1 18288 15.5 0.029 0.030 0.704 1.367 grid_collocate_task_list 119 9.7 1.212 1.360 1.212 1.360 acc_transpose_blocks 18288 15.5 0.075 0.077 1.216 1.236 wfi_extrapolate 11 7.9 0.001 0.001 1.157 1.157 make_images_data 4572 15.5 0.046 0.050 0.793 0.971 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 0.942 0.943 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.871 0.894 hybrid_alltoall_any 4725 16.4 0.056 0.113 0.682 0.868 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.841 0.842 mp_waitany 9880 13.7 0.507 0.788 0.507 0.788 acc_transpose_blocks_kernels 18288 16.5 0.208 0.218 0.776 0.787 rs_pw_transfer_RS2PW_140 130 11.5 0.120 0.123 0.501 0.755 mp_alltoall_d11v 2130 13.8 0.659 0.746 0.659 0.746 mp_alltoall_z22v 1201 16.6 0.635 0.733 0.635 0.733 cp_fm_cholesky_invert 11 10.9 0.720 0.724 0.720 0.724 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.649 0.708 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=32.903000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=498.454545, yerr=3.367320 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b3f583d97791bedccf073d04f637ff05a010c441_performance_tests/06/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 114.044384E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3805952 0.0% 0.0% 100.0% average stack size 0.0 0.0 38.6 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 555.827200E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62659. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_comm_split 50 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.033 0.073 35.342 35.342 qs_mol_dyn_low 1 2.0 0.003 0.005 35.001 35.008 qs_forces 11 3.9 0.009 0.031 34.939 34.944 qs_energies 11 4.9 0.002 0.004 33.185 33.196 scf_env_do_scf 11 5.9 0.001 0.002 28.157 28.159 scf_env_do_scf_inner_loop 108 6.5 0.003 0.008 24.722 24.725 dbcsr_multiply_generic 2286 12.5 0.102 0.106 17.952 18.078 velocity_verlet 10 3.0 0.001 0.001 17.930 17.932 qs_scf_new_mos 108 7.5 0.001 0.001 16.076 16.128 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.076 16.128 ot_scf_mini 108 9.5 0.003 0.003 15.136 15.184 multiply_cannon 2286 13.5 0.231 0.258 14.355 14.780 multiply_cannon_loop 2286 14.5 0.934 0.968 13.420 13.750 ot_mini 108 10.5 0.001 0.001 9.170 9.235 multiply_cannon_multrec 27432 15.5 2.312 2.958 8.554 8.870 qs_ot_get_derivative 108 11.5 0.001 0.001 7.340 7.390 dbcsr_mm_accdrv_process 47916 15.9 5.170 6.849 6.150 7.322 rebuild_ks_matrix 119 8.3 0.000 0.000 6.664 6.718 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 6.663 6.718 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.916 5.963 sum_up_and_integrate 119 10.3 0.035 0.038 3.866 3.876 integrate_v_rspace 119 11.3 0.002 0.003 3.831 3.841 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.665 3.695 calculate_rho_elec 119 8.7 0.040 0.046 3.664 3.694 init_scf_run 11 5.9 0.000 0.001 3.628 3.629 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.628 3.629 init_scf_loop 11 6.9 0.001 0.002 3.415 3.420 qs_ot_get_p 119 10.4 0.001 0.001 3.349 3.419 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.767 3.171 prepare_preconditioner 11 7.9 0.000 0.000 2.554 2.562 make_preconditioner 11 8.9 0.000 0.001 2.554 2.562 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.156 2.480 mp_waitall_1 137007 16.6 1.813 2.401 1.813 2.401 rs_pw_transfer 974 11.9 0.009 0.009 2.101 2.308 make_m2s 4572 13.5 0.055 0.056 2.190 2.302 calculate_first_density_matrix 1 7.0 0.001 0.002 2.261 2.263 density_rs2pw 119 9.7 0.004 0.004 2.074 2.255 pw_transfer 1439 11.6 0.066 0.071 2.181 2.217 make_images 4572 14.5 0.270 0.331 2.082 2.191 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.051 2.178 apply_single 119 13.6 0.000 0.000 2.050 2.178 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.088 2.129 calculate_dm_sparse 119 9.5 0.000 0.000 2.031 2.097 qs_ot_p2m_diag 50 11.0 0.015 0.022 2.054 2.063 grid_integrate_task_list 119 12.3 1.823 1.916 1.823 1.916 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.890 1.915 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.801 1.802 fft3d_ps 1201 14.6 0.560 0.614 1.762 1.797 ot_diis_step 108 11.5 0.012 0.012 1.790 1.791 fft_wrap_pw1pw2_140 487 13.2 0.088 0.097 1.721 1.764 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.744 1.745 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.630 1.646 jit_kernel_multiply 9 15.7 0.919 1.630 0.919 1.630 potential_pw2rs 119 12.3 0.008 0.009 1.621 1.626 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.499 1.500 cp_fm_redistribute_end 50 14.0 0.993 1.477 0.995 1.477 acc_transpose_blocks 27432 15.5 0.112 0.116 1.439 1.460 cp_fm_diag_elpa_base 50 14.0 0.460 1.409 0.481 1.453 multiply_cannon_metrocomm3 27432 15.5 0.039 0.040 0.778 1.404 grid_collocate_task_list 119 9.7 1.220 1.334 1.220 1.334 wfi_extrapolate 11 7.9 0.001 0.001 1.316 1.316 mp_sum_l 7207 12.9 0.979 1.254 0.979 1.254 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.209 1.221 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.142 1.147 cp_fm_upper_to_full 72 13.5 0.832 1.139 0.832 1.139 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.100 1.120 multiply_cannon_sync_h2d 27432 15.5 0.987 1.044 0.987 1.044 dbcsr_complete_redistribute 329 12.2 0.119 0.143 0.766 1.036 make_images_data 4572 15.5 0.047 0.051 0.855 0.986 hybrid_alltoall_any 4725 16.4 0.063 0.152 0.728 0.908 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.790 0.867 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.839 0.844 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.567 0.833 acc_transpose_blocks_kernels 27432 16.5 0.266 0.275 0.815 0.824 mp_alltoall_z22v 1201 16.6 0.786 0.823 0.786 0.823 mp_alltoall_d11v 2130 13.8 0.729 0.808 0.729 0.808 cp_fm_cholesky_invert 11 10.9 0.742 0.745 0.742 0.745 mp_waitany 7280 13.7 0.518 0.728 0.518 0.728 rs_pw_transfer_RS2PW_140 130 11.5 0.108 0.113 0.529 0.724 mp_alltoall_i22 627 13.8 0.433 0.712 0.433 0.712 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=35.342000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=528.363636, yerr=2.568081 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b3f583d97791bedccf073d04f637ff05a010c441_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 612.954112E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62658. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_comm_split 50 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.031 28.695 28.696 qs_mol_dyn_low 1 2.0 0.003 0.004 28.468 28.477 qs_forces 11 3.9 0.002 0.004 28.397 28.398 qs_energies 11 4.9 0.001 0.001 26.685 26.689 scf_env_do_scf 11 5.9 0.000 0.001 21.821 21.821 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 19.319 19.319 velocity_verlet 10 3.0 0.001 0.001 14.463 14.466 dbcsr_multiply_generic 2286 12.5 0.114 0.118 12.545 12.617 qs_scf_new_mos 108 7.5 0.001 0.001 11.197 11.219 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.197 11.219 ot_scf_mini 108 9.5 0.002 0.002 10.521 10.545 multiply_cannon 2286 13.5 0.231 0.238 10.035 10.510 multiply_cannon_loop 2286 14.5 0.328 0.337 9.145 9.325 multiply_cannon_multrec 9144 15.5 1.787 2.059 6.120 6.422 rebuild_ks_matrix 119 8.3 0.000 0.000 6.025 6.044 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.025 6.044 ot_mini 108 10.5 0.001 0.001 5.978 6.009 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.373 5.390 qs_ot_get_derivative 108 11.5 0.001 0.001 4.681 4.707 dbcsr_mm_accdrv_process 12550 15.8 3.128 4.226 4.232 4.325 sum_up_and_integrate 119 10.3 0.038 0.041 3.700 3.704 integrate_v_rspace 119 11.3 0.002 0.003 3.661 3.667 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.618 3.624 calculate_rho_elec 119 8.7 0.060 0.061 3.617 3.623 init_scf_run 11 5.9 0.000 0.001 3.428 3.429 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.428 3.428 qs_ot_get_p 119 10.4 0.001 0.001 2.768 2.808 init_scf_loop 11 6.9 0.000 0.000 2.482 2.483 calculate_first_density_matrix 1 7.0 0.000 0.000 2.301 2.302 pw_transfer 1439 11.6 0.066 0.069 2.185 2.195 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.091 2.102 density_rs2pw 119 9.7 0.004 0.004 1.940 2.080 mp_waitall_1 115863 16.7 1.493 1.975 1.493 1.975 grid_integrate_task_list 119 12.3 1.840 1.949 1.840 1.949 jit_kernel_multiply 10 15.6 1.066 1.932 1.066 1.932 make_m2s 4572 13.5 0.035 0.036 1.725 1.877 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.857 1.859 rs_pw_transfer 974 11.9 0.008 0.008 1.701 1.850 make_images 4572 14.5 0.268 0.304 1.635 1.786 calculate_dm_sparse 119 9.5 0.000 0.000 1.763 1.783 fft3d_ps 1201 14.6 0.564 0.575 1.751 1.759 prepare_preconditioner 11 7.9 0.000 0.000 1.742 1.747 make_preconditioner 11 8.9 0.000 0.000 1.742 1.747 fft_wrap_pw1pw2_140 487 13.2 0.087 0.091 1.731 1.744 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.691 1.693 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.664 1.664 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.633 1.657 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.594 1.606 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.528 1.536 potential_pw2rs 119 12.3 0.010 0.010 1.433 1.436 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.390 1.391 grid_collocate_task_list 119 9.7 1.269 1.387 1.269 1.387 cp_fm_redistribute_end 50 14.0 0.694 1.370 0.694 1.371 cp_fm_diag_elpa_base 50 14.0 0.629 1.294 0.675 1.356 ot_diis_step 108 11.5 0.012 0.013 1.285 1.285 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.216 1.218 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.181 1.187 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.158 1.184 apply_single 119 13.6 0.000 0.000 1.158 1.184 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.116 1.130 wfi_extrapolate 11 7.9 0.001 0.001 1.078 1.078 hybrid_alltoall_any 4725 16.4 0.064 0.176 0.795 1.020 acc_transpose_blocks 9144 15.5 0.038 0.040 1.010 1.017 make_images_data 4572 15.5 0.041 0.044 0.811 0.997 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.865 0.921 mp_alltoall_d11v 2130 13.8 0.777 0.918 0.777 0.918 cp_fm_cholesky_invert 11 10.9 0.855 0.858 0.855 0.858 multiply_cannon_metrocomm3 9144 15.5 0.019 0.020 0.395 0.852 multiply_cannon_sync_h2d 9144 15.5 0.715 0.787 0.715 0.787 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.775 0.777 acc_transpose_blocks_kernels 9144 16.5 0.117 0.119 0.769 0.771 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.719 0.729 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.672 0.722 mp_alltoall_z22v 1201 16.6 0.633 0.685 0.633 0.685 jit_kernel_transpose 5 15.6 0.653 0.654 0.653 0.654 mp_allgather_i34 2286 14.5 0.224 0.641 0.224 0.641 yz_to_x 606 15.1 0.265 0.276 0.597 0.612 x_to_yz 595 16.2 0.276 0.289 0.578 0.583 mp_waitany 5200 13.7 0.451 0.582 0.451 0.582 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=28.696000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=575.818182, yerr=7.649340 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b3f583d97791bedccf073d04f637ff05a010c441_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 744.787968E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63729. MP_Allreduce 10074 433. MP_Sync 54 MP_Alltoall 1582 7383731. MP_SendRecv 2499 189067. MP_ISendRecv 2499 189067. MP_Wait 6399 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.038 42.163 42.164 qs_mol_dyn_low 1 2.0 0.003 0.004 41.958 41.965 qs_forces 11 3.9 0.001 0.002 41.872 41.881 qs_energies 11 4.9 0.001 0.001 39.871 39.877 scf_env_do_scf 11 5.9 0.001 0.001 33.831 33.831 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 25.975 25.976 velocity_verlet 10 3.0 0.001 0.001 23.464 23.471 dbcsr_multiply_generic 2286 12.5 0.102 0.103 17.791 17.977 qs_scf_new_mos 108 7.5 0.001 0.001 15.849 15.948 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.848 15.947 ot_scf_mini 108 9.5 0.002 0.002 14.767 14.868 multiply_cannon 2286 13.5 0.304 0.314 13.921 14.813 multiply_cannon_loop 2286 14.5 0.342 0.349 12.679 13.564 multiply_cannon_multrec 9144 15.5 3.431 4.955 8.880 9.034 ot_mini 108 10.5 0.001 0.001 8.843 8.959 init_scf_loop 11 6.9 0.000 0.000 7.830 7.834 rebuild_ks_matrix 119 8.3 0.000 0.000 7.395 7.537 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.395 7.537 qs_ot_get_derivative 108 11.5 0.001 0.001 6.816 6.917 prepare_preconditioner 11 7.9 0.000 0.000 6.863 6.876 make_preconditioner 11 8.9 0.000 0.000 6.863 6.876 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.679 6.807 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.415 6.754 dbcsr_mm_accdrv_process 12550 15.8 4.142 5.934 5.327 6.705 cp_fm_upper_to_full 72 14.2 3.244 4.633 3.244 4.633 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.289 4.293 calculate_rho_elec 119 8.7 0.118 0.121 4.288 4.293 sum_up_and_integrate 119 10.3 0.065 0.066 4.074 4.079 integrate_v_rspace 119 11.3 0.003 0.003 4.008 4.015 init_scf_run 11 5.9 0.000 0.001 3.998 3.998 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.998 3.998 qs_ot_get_p 119 10.4 0.001 0.001 3.204 3.342 mp_waitall_1 94719 16.7 2.446 3.317 2.446 3.317 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.455 2.856 pw_transfer 1439 11.6 0.069 0.070 2.834 2.840 dbcsr_complete_redistribute 329 12.2 0.287 0.293 1.985 2.825 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.736 2.740 calculate_first_density_matrix 1 7.0 0.000 0.000 2.521 2.522 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.674 2.492 calculate_dm_sparse 119 9.5 0.000 0.000 2.462 2.481 make_m2s 4572 13.5 0.038 0.039 2.287 2.453 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.168 2.437 apply_single 119 13.6 0.000 0.000 2.168 2.437 fft3d_ps 1201 14.6 0.596 0.605 2.360 2.364 make_images 4572 14.5 0.350 0.382 2.166 2.331 fft_wrap_pw1pw2_140 487 13.2 0.095 0.097 2.299 2.306 mp_alltoall_i22 627 13.8 1.465 2.295 1.465 2.295 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.443 2.255 density_rs2pw 119 9.7 0.004 0.004 2.220 2.235 multiply_cannon_metrocomm3 9144 15.5 0.020 0.021 1.348 2.187 grid_integrate_task_list 119 12.3 2.054 2.063 2.054 2.063 ot_diis_step 108 11.5 0.014 0.014 2.001 2.001 jit_kernel_multiply 10 15.7 1.157 1.965 1.157 1.965 qs_ot_p2m_diag 50 11.0 0.043 0.043 1.962 1.964 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.950 1.952 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.781 1.790 mp_sum_l 7207 12.9 1.040 1.722 1.040 1.722 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.671 1.722 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.687 1.687 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.644 1.653 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.527 1.534 rs_pw_transfer 974 11.9 0.009 0.010 1.481 1.502 grid_collocate_task_list 119 9.7 1.487 1.499 1.487 1.499 potential_pw2rs 119 12.3 0.014 0.014 1.471 1.473 cp_fm_cholesky_invert 11 10.9 1.431 1.434 1.431 1.434 wfi_extrapolate 11 7.9 0.001 0.001 1.413 1.413 hybrid_alltoall_any 4725 16.4 0.088 0.150 1.165 1.413 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.386 1.387 cp_fm_diag_elpa_base 50 14.0 1.238 1.289 1.384 1.384 make_images_data 4572 15.5 0.045 0.048 1.113 1.317 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.127 1.145 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.120 1.138 mp_alltoall_d11v 2130 13.8 1.072 1.093 1.072 1.093 multiply_cannon_sync_h2d 9144 15.5 1.044 1.049 1.044 1.049 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.981 1.018 yz_to_x 606 15.1 0.460 0.470 0.945 0.957 qs_create_task_list 11 7.9 0.000 0.000 0.942 0.956 generate_qs_task_list 11 8.9 0.371 0.390 0.942 0.956 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.929 0.941 acc_transpose_blocks 9144 15.5 0.038 0.039 0.902 0.910 mp_alltoall_z22v 1201 16.6 0.811 0.856 0.811 0.856 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=42.164000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=696.000000, yerr=13.436720 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b3f583d97791bedccf073d04f637ff05a010c441_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 503.709696E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65372. MP_Allreduce 9840 486. MP_Sync 100 MP_Alltoall 1938 1379060. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_comm_split 48 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.029 84.940 84.940 qs_mol_dyn_low 1 2.0 0.003 0.004 84.686 84.695 qs_forces 11 3.9 0.008 0.028 84.611 84.618 qs_energies 11 4.9 0.001 0.002 81.753 81.775 scf_env_do_scf 11 5.9 0.000 0.001 70.982 70.985 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 65.375 65.375 dbcsr_multiply_generic 2055 12.4 0.107 0.111 53.533 53.862 qs_scf_new_mos 99 7.5 0.000 0.001 47.693 47.831 qs_scf_loop_do_ot 99 8.5 0.001 0.001 47.693 47.831 ot_scf_mini 99 9.5 0.002 0.002 45.255 45.411 multiply_cannon 2055 13.4 0.174 0.181 44.301 45.268 multiply_cannon_loop 2055 14.4 1.538 1.589 43.348 44.326 velocity_verlet 10 3.0 0.001 0.001 41.799 41.799 ot_mini 99 10.5 0.001 0.001 27.113 27.264 qs_ot_get_derivative 99 11.5 0.004 0.031 20.230 20.352 multiply_cannon_multrec 49320 15.4 12.217 12.764 18.289 19.838 rebuild_ks_matrix 110 8.3 0.000 0.000 14.808 14.984 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.808 14.983 mp_waitall_1 220248 16.4 12.129 13.444 12.129 13.444 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.012 13.162 multiply_cannon_sync_h2d 49320 15.4 10.031 10.697 10.031 10.697 qs_ot_get_p 110 10.4 0.001 0.001 9.570 9.726 init_scf_run 11 5.9 0.000 0.001 8.700 8.700 scf_env_initial_rho_setup 11 6.9 0.000 0.001 8.700 8.700 multiply_cannon_metrocomm3 49320 15.4 0.080 0.085 6.725 8.298 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.208 7.774 apply_single 110 13.6 0.000 0.001 7.207 7.774 sum_up_and_integrate 110 10.3 0.036 0.042 7.248 7.262 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 6.682 7.237 integrate_v_rspace 110 11.3 0.003 0.003 7.211 7.235 dbcsr_mm_accdrv_process 87628 16.1 2.052 2.162 5.951 7.105 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.718 6.864 calculate_rho_elec 110 8.6 0.021 0.026 6.717 6.863 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.574 6.658 ot_diis_step 99 11.5 0.006 0.006 6.602 6.602 qs_ot_p2m_diag 48 11.0 0.012 0.018 6.438 6.480 calculate_dm_sparse 110 9.5 0.000 0.001 5.774 5.892 mp_sum_l 6514 12.8 4.890 5.875 4.890 5.875 init_scf_loop 11 6.9 0.000 0.000 5.580 5.581 cp_dbcsr_syevd 48 12.0 0.002 0.002 5.414 5.414 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.935 4.936 cp_fm_redistribute_end 48 14.0 4.310 4.912 4.314 4.913 cp_fm_diag_elpa_base 48 14.0 0.590 4.735 0.596 4.781 multiply_cannon_metrocomm1 49320 15.4 0.062 0.065 3.507 4.662 calculate_first_density_matrix 1 7.0 0.000 0.000 4.560 4.575 jit_kernel_multiply 13 15.9 3.617 4.543 3.617 4.543 rs_pw_transfer 902 11.9 0.011 0.013 3.773 4.260 wfi_extrapolate 11 7.9 0.001 0.001 4.042 4.042 make_m2s 4110 13.4 0.061 0.065 3.878 3.999 density_rs2pw 110 9.6 0.004 0.005 3.479 3.972 make_images 4110 14.4 0.178 0.193 3.782 3.906 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 3.610 3.640 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.528 3.531 pw_transfer 1331 11.6 0.055 0.066 3.370 3.437 prepare_preconditioner 11 7.9 0.000 0.000 3.406 3.426 make_preconditioner 11 8.9 0.000 0.000 3.406 3.426 grid_integrate_task_list 110 12.3 3.245 3.404 3.245 3.404 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.289 3.354 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 3.282 3.349 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.203 3.271 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.183 3.229 fft3d_ps 1111 14.6 0.783 0.889 2.811 2.863 fft_wrap_pw1pw2_140 451 13.1 0.171 0.192 2.779 2.849 potential_pw2rs 110 12.3 0.006 0.007 2.752 2.772 mp_alltoall_d11v 2046 13.8 2.141 2.573 2.141 2.573 mp_waitany 14300 13.8 1.852 2.518 1.852 2.518 grid_collocate_task_list 110 9.6 2.086 2.292 2.086 2.292 acc_transpose_blocks 49320 15.4 0.223 0.235 2.185 2.289 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.939 1.963 mp_sum_d 3879 11.9 1.446 1.927 1.446 1.927 make_images_data 4110 15.4 0.044 0.047 1.761 1.897 cp_fm_cholesky_invert 11 10.9 1.854 1.858 1.854 1.858 hybrid_alltoall_any 4261 16.3 0.082 0.483 1.529 1.824 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=84.940000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=478.181818, yerr=3.009626 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b3f583d97791bedccf073d04f637ff05a010c441_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 590.471168E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1972800 MPI messages size (bytes): total size 1.077520E+12 min size 0.000000E+00 max size 4.537280E+06 average size 546.188250E+03 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 222984 1826684928 8192 < size <= 32768 520356 13399818240 32768 < size <= 131072 372336 35386294272 131072 < size <= 4194304 787758 788321309808 4194304 < size <= 16777216 54450 238588003280 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65587. MP_Allreduce 9839 562. MP_Sync 100 MP_Alltoall 1717 2235246. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 MP_comm_split 48 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.032 69.542 69.544 qs_mol_dyn_low 1 2.0 0.003 0.004 69.331 69.341 qs_forces 11 3.9 0.004 0.005 69.259 69.260 qs_energies 11 4.9 0.001 0.001 65.927 65.932 scf_env_do_scf 11 5.9 0.000 0.001 56.411 56.414 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 48.811 48.812 dbcsr_multiply_generic 2055 12.4 0.115 0.120 38.478 38.647 velocity_verlet 10 3.0 0.001 0.001 36.196 36.197 multiply_cannon 2055 13.4 0.222 0.245 31.951 33.221 qs_scf_new_mos 99 7.5 0.001 0.001 32.347 32.463 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.347 32.463 multiply_cannon_loop 2055 14.4 0.922 0.945 30.657 31.580 ot_scf_mini 99 9.5 0.003 0.003 30.690 30.794 ot_mini 99 10.5 0.001 0.001 18.031 18.144 multiply_cannon_multrec 24660 15.4 7.613 8.853 14.504 16.178 rebuild_ks_matrix 110 8.3 0.000 0.000 13.910 13.992 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 13.909 13.991 qs_ot_get_derivative 99 11.5 0.001 0.003 12.252 12.355 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.253 12.326 mp_waitall_1 176588 16.5 7.897 10.816 7.897 10.816 multiply_cannon_sync_h2d 24660 15.4 7.052 8.027 7.052 8.027 multiply_cannon_metrocomm3 24660 15.4 0.071 0.073 5.288 8.002 init_scf_loop 11 6.9 0.000 0.000 7.564 7.564 dbcsr_mm_accdrv_process 52282 16.1 5.276 6.469 6.730 7.432 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.473 7.253 apply_single 110 13.6 0.000 0.001 6.473 7.252 init_scf_run 11 5.9 0.000 0.001 7.154 7.154 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.154 7.154 sum_up_and_integrate 110 10.3 0.053 0.059 6.698 6.712 integrate_v_rspace 110 11.3 0.002 0.003 6.645 6.659 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.209 6.217 calculate_rho_elec 110 8.6 0.040 0.047 6.208 6.217 qs_ot_get_p 110 10.4 0.001 0.001 6.042 6.200 ot_diis_step 99 11.5 0.010 0.010 5.732 5.732 prepare_preconditioner 11 7.9 0.000 0.000 5.529 5.549 make_preconditioner 11 8.9 0.000 0.000 5.529 5.549 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.616 5.343 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.128 5.290 make_m2s 4110 13.4 0.057 0.060 4.151 4.653 make_images 4110 14.4 0.397 0.444 4.042 4.540 qs_ot_p2m_diag 48 11.0 0.028 0.043 4.196 4.216 calculate_dm_sparse 110 9.5 0.001 0.001 4.002 4.033 pw_transfer 1331 11.6 0.066 0.074 3.781 3.918 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.674 3.814 density_rs2pw 110 9.6 0.004 0.004 3.320 3.795 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.774 3.775 wfi_extrapolate 11 7.9 0.001 0.001 3.543 3.543 calculate_first_density_matrix 1 7.0 0.000 0.000 3.521 3.524 rs_pw_transfer 902 11.9 0.012 0.013 2.952 3.464 grid_integrate_task_list 110 12.3 3.155 3.353 3.155 3.353 jit_kernel_multiply 12 16.3 1.101 3.298 1.101 3.298 fft_wrap_pw1pw2_140 451 13.1 0.203 0.219 3.143 3.287 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.262 3.276 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.256 3.258 cp_fm_redistribute_end 48 14.0 2.435 3.232 2.437 3.232 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.148 3.210 cp_fm_diag_elpa_base 48 14.0 0.757 3.075 0.792 3.175 fft3d_ps 1111 14.6 1.103 1.319 3.022 3.156 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.919 2.949 make_images_data 4110 15.4 0.048 0.052 2.331 2.836 hybrid_alltoall_any 4261 16.3 0.103 0.442 2.058 2.802 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 2.579 2.588 cp_fm_cholesky_invert 11 10.9 2.537 2.544 2.537 2.544 potential_pw2rs 110 12.3 0.008 0.009 2.475 2.484 grid_collocate_task_list 110 9.6 2.053 2.483 2.053 2.483 mp_sum_l 6514 12.8 1.744 2.373 1.744 2.373 mp_alltoall_d11v 2046 13.8 1.751 2.095 1.751 2.095 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.937 1.953 qs_energies_init_hamiltonians 11 5.9 0.001 0.014 1.858 1.861 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.821 1.837 mp_waitany 10164 13.8 1.207 1.747 1.207 1.747 multiply_cannon_metrocomm4 22605 15.4 0.074 0.078 0.786 1.634 mp_allgather_i34 2055 14.4 0.623 1.633 0.623 1.633 acc_transpose_blocks 24660 15.4 0.109 0.112 1.508 1.527 mp_irecv_dv 57340 16.2 0.660 1.513 0.660 1.513 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.363 1.469 rs_pw_transfer_RS2PW_140 121 11.5 0.208 0.220 0.949 1.455 dbcsr_complete_redistribute 325 12.2 0.242 0.308 1.125 1.406 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=69.544000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=556.181818, yerr=7.649340 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b3f583d97791bedccf073d04f637ff05a010c441_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 662.417408E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65578. MP_Allreduce 9838 559. MP_Sync 100 MP_Alltoall 1496 4511006. MP_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 MP_comm_split 48 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.024 0.035 61.132 61.134 qs_mol_dyn_low 1 2.0 0.003 0.004 60.774 60.784 qs_forces 11 3.9 0.002 0.006 60.705 60.707 qs_energies 11 4.9 0.005 0.017 57.481 57.485 scf_env_do_scf 11 5.9 0.001 0.001 48.403 48.403 scf_env_do_scf_inner_loop 99 6.5 0.007 0.014 39.805 39.807 velocity_verlet 10 3.0 0.001 0.001 32.563 32.565 dbcsr_multiply_generic 2055 12.4 0.117 0.133 29.391 29.618 qs_scf_new_mos 99 7.5 0.001 0.001 24.610 24.706 qs_scf_loop_do_ot 99 8.5 0.001 0.001 24.610 24.705 multiply_cannon 2055 13.4 0.212 0.225 23.149 24.322 ot_scf_mini 99 9.5 0.003 0.003 23.397 23.510 multiply_cannon_loop 2055 14.4 0.613 0.629 21.957 22.963 ot_mini 99 10.5 0.001 0.001 13.352 13.462 rebuild_ks_matrix 110 8.3 0.000 0.000 12.506 12.634 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.506 12.634 multiply_cannon_multrec 16440 15.4 4.016 5.620 10.317 11.853 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.010 11.130 mp_waitall_1 139946 16.5 7.357 10.271 7.357 10.271 qs_ot_get_derivative 99 11.5 0.001 0.001 8.965 9.078 init_scf_loop 11 6.9 0.001 0.003 8.564 8.566 multiply_cannon_metrocomm3 16440 15.4 0.044 0.045 4.439 7.619 dbcsr_mm_accdrv_process 34862 16.1 5.069 6.074 6.155 6.842 prepare_preconditioner 11 7.9 0.000 0.000 6.760 6.777 make_preconditioner 11 8.9 0.000 0.001 6.760 6.777 init_scf_run 11 5.9 0.000 0.001 6.679 6.683 scf_env_initial_rho_setup 11 6.9 0.002 0.006 6.679 6.683 sum_up_and_integrate 110 10.3 0.060 0.061 6.607 6.622 integrate_v_rspace 110 11.3 0.003 0.004 6.546 6.562 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.088 6.436 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.060 6.070 calculate_rho_elec 110 8.6 0.058 0.059 6.060 6.070 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.982 5.394 apply_single 110 13.6 0.000 0.000 4.982 5.393 qs_ot_get_p 110 10.4 0.001 0.001 5.233 5.368 multiply_cannon_sync_h2d 16440 15.4 3.724 4.590 3.724 4.590 make_m2s 4110 13.4 0.050 0.052 4.140 4.526 make_images 4110 14.4 0.391 0.509 4.024 4.411 ot_diis_step 99 11.5 0.011 0.011 4.360 4.361 density_rs2pw 110 9.6 0.004 0.005 3.144 4.312 rs_pw_transfer 902 11.9 0.010 0.011 2.661 3.818 pw_transfer 1331 11.6 0.065 0.074 3.763 3.773 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.047 3.715 calculate_dm_sparse 110 9.5 0.001 0.001 3.680 3.711 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.667 3.672 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.656 3.668 calculate_first_density_matrix 1 7.0 0.001 0.003 3.622 3.628 grid_integrate_task_list 110 12.3 3.156 3.384 3.156 3.384 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.333 3.333 fft_wrap_pw1pw2_140 451 13.1 0.214 0.217 3.164 3.178 fft3d_ps 1111 14.6 1.088 1.098 2.955 2.965 wfi_extrapolate 11 7.9 0.001 0.001 2.953 2.953 make_images_data 4110 15.4 0.045 0.050 2.400 2.912 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.832 2.833 hybrid_alltoall_any 4261 16.3 0.106 0.379 2.105 2.822 cp_fm_redistribute_end 48 14.0 1.766 2.805 1.768 2.806 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.786 2.787 cp_fm_diag_elpa_base 48 14.0 0.970 2.645 1.033 2.761 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 2.646 2.659 cp_fm_cholesky_invert 11 10.9 2.615 2.621 2.615 2.621 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.462 2.509 grid_collocate_task_list 110 9.6 2.087 2.470 2.087 2.470 jit_kernel_multiply 10 16.4 0.695 2.457 0.695 2.457 mp_waitany 17072 13.8 1.176 2.390 1.176 2.390 multiply_cannon_metrocomm4 14385 15.4 0.046 0.050 0.861 2.365 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.285 2.350 potential_pw2rs 110 12.3 0.011 0.011 2.327 2.333 mp_irecv_dv 48980 15.7 0.790 2.244 0.790 2.244 mp_alltoall_d11v 2046 13.8 1.759 2.201 1.759 2.201 rs_pw_transfer_RS2PW_140 121 11.5 0.177 0.182 0.899 2.053 mp_sum_l 6514 12.8 1.475 2.032 1.475 2.032 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 1.970 1.972 cp_fm_upper_to_full 70 13.6 1.495 1.962 1.495 1.962 dbcsr_complete_redistribute 325 12.2 0.339 0.409 1.447 1.923 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.695 1.709 multiply_cannon_metrocomm1 16440 15.4 0.029 0.030 0.617 1.701 cp_fm_cholesky_decompose 22 10.9 1.556 1.574 1.556 1.574 mp_allgather_i34 2055 14.4 0.516 1.559 0.516 1.559 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.349 1.456 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.973 1.437 rs_gather_matrices 110 12.3 0.231 0.259 0.984 1.388 acc_transpose_blocks 16440 15.4 0.071 0.073 1.208 1.257 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.240 1.248 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=61.134000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=627.272727, yerr=8.873770 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b3f583d97791bedccf073d04f637ff05a010c441_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 727.658496E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65576. MP_Allreduce 9838 600. MP_Sync 100 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 MP_comm_split 48 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.038 65.158 65.159 qs_mol_dyn_low 1 2.0 0.003 0.004 64.879 64.888 qs_forces 11 3.9 0.002 0.002 64.812 64.813 qs_energies 11 4.9 0.002 0.003 61.384 61.388 scf_env_do_scf 11 5.9 0.001 0.001 53.031 53.034 scf_env_do_scf_inner_loop 99 6.5 0.010 0.066 41.423 41.425 velocity_verlet 10 3.0 0.001 0.001 36.866 36.868 dbcsr_multiply_generic 2055 12.4 0.116 0.119 29.289 29.535 qs_scf_new_mos 99 7.5 0.001 0.001 26.169 26.255 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.168 26.254 ot_scf_mini 99 9.5 0.003 0.004 24.560 24.657 multiply_cannon 2055 13.4 0.246 0.266 22.545 23.710 multiply_cannon_loop 2055 14.4 0.883 0.901 21.041 21.633 ot_mini 99 10.5 0.001 0.001 13.832 13.947 multiply_cannon_multrec 24660 15.4 4.204 6.770 12.731 13.877 rebuild_ks_matrix 110 8.3 0.000 0.000 12.330 12.459 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.330 12.458 init_scf_loop 11 6.9 0.000 0.002 11.569 11.573 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.884 10.995 qs_ot_get_derivative 99 11.5 0.001 0.001 9.723 9.823 prepare_preconditioner 11 7.9 0.000 0.000 9.799 9.815 make_preconditioner 11 8.9 0.000 0.001 9.799 9.815 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.037 9.465 dbcsr_mm_accdrv_process 52304 16.0 6.974 8.339 8.381 9.307 sum_up_and_integrate 110 10.3 0.067 0.071 6.563 6.577 integrate_v_rspace 110 11.3 0.003 0.003 6.495 6.510 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.218 6.233 calculate_rho_elec 110 8.6 0.077 0.081 6.218 6.233 mp_waitall_1 121746 16.5 4.112 6.218 4.112 6.218 qs_ot_get_p 110 10.4 0.001 0.001 5.758 5.913 make_m2s 4110 13.4 0.060 0.061 5.201 5.652 init_scf_run 11 5.9 0.000 0.001 5.594 5.595 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.594 5.595 make_images 4110 14.4 0.572 0.695 5.059 5.505 cp_fm_upper_to_full 70 13.8 3.388 4.736 3.388 4.736 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.995 4.124 apply_single 110 13.6 0.000 0.000 3.995 4.124 ot_diis_step 99 11.5 0.011 0.011 4.074 4.074 pw_transfer 1331 11.6 0.065 0.075 3.914 3.945 qs_ot_p2m_diag 48 11.0 0.054 0.063 3.914 3.928 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.808 3.841 dbcsr_complete_redistribute 325 12.2 0.433 0.473 2.612 3.735 density_rs2pw 110 9.6 0.004 0.004 3.084 3.597 grid_integrate_task_list 110 12.3 3.268 3.449 3.268 3.449 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.442 3.443 multiply_cannon_sync_h2d 24660 15.4 3.182 3.350 3.182 3.350 fft_wrap_pw1pw2_140 451 13.1 0.202 0.212 3.289 3.324 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.245 3.296 multiply_cannon_metrocomm3 24660 15.4 0.036 0.038 1.419 3.274 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.108 3.216 fft3d_ps 1111 14.6 1.091 1.131 3.101 3.126 hybrid_alltoall_any 4261 16.3 0.121 0.458 2.252 3.076 make_images_data 4110 15.4 0.048 0.052 2.600 3.031 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.944 2.997 wfi_extrapolate 11 7.9 0.001 0.001 2.957 2.957 calculate_dm_sparse 110 9.5 0.001 0.001 2.906 2.934 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.925 2.927 cp_fm_redistribute_end 48 14.0 1.464 2.902 1.465 2.903 rs_pw_transfer 902 11.9 0.010 0.011 2.377 2.900 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.881 2.883 cp_fm_diag_elpa_base 48 14.0 1.350 2.753 1.435 2.883 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.753 2.846 mp_alltoall_i22 605 13.7 1.604 2.746 1.604 2.746 cp_fm_cholesky_invert 11 10.9 2.602 2.610 2.602 2.610 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.528 2.561 calculate_first_density_matrix 1 7.0 0.000 0.000 2.547 2.548 grid_collocate_task_list 110 9.6 2.180 2.448 2.180 2.448 jit_kernel_multiply 12 15.9 1.076 2.293 1.076 2.293 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 2.278 2.279 potential_pw2rs 110 12.3 0.012 0.013 2.208 2.214 mp_alltoall_d11v 2046 13.8 1.806 1.971 1.806 1.971 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.676 1.711 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.690 1.705 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.594 1.700 cp_fm_cholesky_decompose 22 10.9 1.594 1.638 1.594 1.638 mp_allgather_i34 2055 14.4 0.603 1.569 0.603 1.569 mp_waitany 13376 13.8 1.063 1.565 1.063 1.565 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.531 1.547 acc_transpose_blocks 24660 15.4 0.106 0.109 1.498 1.527 multiply_cannon_metrocomm4 20550 15.4 0.058 0.061 0.853 1.460 mp_irecv_dv 62702 16.1 0.752 1.378 0.752 1.378 mp_sum_l 6514 12.8 0.855 1.374 0.855 1.374 rs_pw_transfer_RS2PW_140 121 11.5 0.169 0.176 0.832 1.362 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.230 1.339 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=65.159000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=688.181818, yerr=10.025587 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b3f583d97791bedccf073d04f637ff05a010c441_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 842.665984E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65574. MP_Allreduce 9838 640. MP_Sync 100 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 MP_comm_split 48 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.031 55.654 55.656 qs_mol_dyn_low 1 2.0 0.003 0.004 55.204 55.351 qs_forces 11 3.9 0.002 0.002 54.977 54.978 qs_energies 11 4.9 0.001 0.002 51.293 51.297 scf_env_do_scf 11 5.9 0.000 0.001 42.999 42.999 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 35.195 35.196 velocity_verlet 10 3.0 0.001 0.001 31.281 31.294 dbcsr_multiply_generic 2055 12.4 0.107 0.108 22.659 22.781 qs_scf_new_mos 99 7.5 0.001 0.001 20.031 20.078 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.030 20.077 ot_scf_mini 99 9.5 0.002 0.002 18.794 18.817 multiply_cannon 2055 13.4 0.247 0.256 17.296 18.638 multiply_cannon_loop 2055 14.4 0.320 0.333 15.971 16.254 rebuild_ks_matrix 110 8.3 0.000 0.000 11.984 12.017 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 11.984 12.017 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.647 10.676 ot_mini 99 10.5 0.001 0.001 10.207 10.223 multiply_cannon_multrec 8220 15.4 3.230 4.472 7.534 8.523 init_scf_loop 11 6.9 0.000 0.000 7.758 7.759 mp_waitall_1 103326 16.6 5.920 7.537 5.920 7.537 sum_up_and_integrate 110 10.3 0.080 0.081 6.533 6.547 qs_ot_get_derivative 99 11.5 0.001 0.001 6.447 6.469 integrate_v_rspace 110 11.3 0.003 0.003 6.453 6.468 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.348 6.364 calculate_rho_elec 110 8.6 0.116 0.116 6.347 6.363 prepare_preconditioner 11 7.9 0.000 0.000 6.094 6.100 make_preconditioner 11 8.9 0.000 0.000 6.094 6.100 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.665 5.742 init_scf_run 11 5.9 0.000 0.001 5.148 5.148 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.147 5.148 dbcsr_mm_accdrv_process 17442 15.9 2.866 3.918 4.174 5.113 qs_ot_get_p 110 10.4 0.001 0.001 4.678 4.711 make_m2s 4110 13.4 0.039 0.040 4.119 4.392 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 2.945 4.328 make_images 4110 14.4 0.634 0.689 3.989 4.259 pw_transfer 1331 11.6 0.066 0.074 4.139 4.162 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.031 4.058 ot_diis_step 99 11.5 0.012 0.012 3.738 3.739 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.658 3.686 apply_single 110 13.6 0.000 0.000 3.657 3.686 grid_integrate_task_list 110 12.3 3.351 3.575 3.351 3.575 fft_wrap_pw1pw2_140 451 13.1 0.214 0.217 3.504 3.541 density_rs2pw 110 9.6 0.004 0.004 3.100 3.474 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.409 3.412 fft3d_ps 1111 14.6 1.144 1.194 3.278 3.303 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.109 3.109 multiply_cannon_sync_h2d 8220 15.4 2.913 3.029 2.913 3.029 cp_fm_cholesky_invert 11 10.9 2.829 2.832 2.829 2.832 make_images_data 4110 15.4 0.040 0.046 2.359 2.746 hybrid_alltoall_any 4261 16.3 0.201 0.867 2.291 2.739 wfi_extrapolate 11 7.9 0.001 0.001 2.724 2.724 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.692 2.694 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.616 2.616 cp_fm_redistribute_end 48 14.0 0.663 2.590 0.667 2.590 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.570 2.572 cp_fm_diag_elpa_base 48 14.0 1.729 2.387 1.915 2.564 grid_collocate_task_list 110 9.6 2.277 2.536 2.277 2.536 rs_pw_transfer 902 11.9 0.010 0.010 2.122 2.524 calculate_dm_sparse 110 9.5 0.001 0.001 2.460 2.494 calculate_first_density_matrix 1 7.0 0.000 0.000 2.327 2.329 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.231 2.237 potential_pw2rs 110 12.3 0.015 0.015 2.088 2.091 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.767 1.985 mp_alltoall_d11v 2046 13.8 1.583 1.861 1.583 1.861 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.838 1.854 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.734 1.744 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.672 1.676 cp_fm_cholesky_decompose 22 10.9 1.649 1.661 1.649 1.661 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.533 1.660 mp_allgather_i34 2055 14.4 0.512 1.594 0.512 1.594 dbcsr_complete_redistribute 325 12.2 0.566 0.583 1.427 1.521 mp_waitany 9240 13.8 1.046 1.465 1.046 1.465 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.424 1.434 qs_create_task_list 11 7.9 0.000 0.001 1.241 1.344 generate_qs_task_list 11 8.9 0.380 0.449 1.240 1.343 multiply_cannon_metrocomm1 8220 15.4 0.021 0.022 0.805 1.291 jit_kernel_multiply 7 15.7 1.000 1.272 1.000 1.272 rs_pw_transfer_RS2PW_140 121 11.5 0.161 0.165 0.794 1.198 rs_gather_matrices 110 12.3 0.322 0.365 0.926 1.190 multiply_cannon_metrocomm4 6165 15.4 0.018 0.019 0.485 1.134 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=55.656000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=790.454545, yerr=12.470792 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b3f583d97791bedccf073d04f637ff05a010c441_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 1.354400E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67104. MP_Allreduce 9672 819. MP_Sync 52 MP_Alltoall 1474 16505187. MP_SendRecv 2310 360267. MP_ISendRecv 2310 360267. MP_Wait 5214 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.036 87.140 87.141 qs_mol_dyn_low 1 2.0 0.003 0.004 86.822 86.831 qs_forces 11 3.9 0.002 0.003 86.757 86.758 qs_energies 11 4.9 0.001 0.001 82.592 82.598 scf_env_do_scf 11 5.9 0.001 0.001 72.586 72.586 velocity_verlet 10 3.0 0.001 0.001 55.828 55.833 scf_env_do_scf_inner_loop 99 6.5 0.042 0.046 43.855 43.856 init_scf_loop 11 6.9 0.000 0.000 28.659 28.661 dbcsr_multiply_generic 2055 12.4 0.120 0.121 27.969 28.010 prepare_preconditioner 11 7.9 0.000 0.000 26.481 26.486 make_preconditioner 11 8.9 0.000 0.000 26.481 26.486 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.770 25.964 qs_scf_new_mos 99 7.5 0.001 0.001 25.487 25.509 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.487 25.509 ot_scf_mini 99 9.5 0.002 0.002 23.758 23.770 multiply_cannon 2055 13.4 0.347 0.367 21.132 22.068 multiply_cannon_loop 2055 14.4 0.339 0.342 19.357 19.775 cp_fm_upper_to_full 70 14.2 12.933 18.365 12.933 18.365 rebuild_ks_matrix 110 8.3 0.000 0.001 14.154 14.160 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 14.153 14.159 ot_mini 99 10.5 0.001 0.001 13.227 13.238 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.821 12.826 dbcsr_complete_redistribute 325 12.2 1.023 1.070 7.266 10.400 multiply_cannon_multrec 8220 15.4 4.352 4.518 9.549 9.638 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.254 9.389 qs_ot_get_derivative 99 11.5 0.001 0.001 8.804 8.814 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.696 8.811 mp_alltoall_i22 605 13.7 5.316 8.457 5.316 8.457 mp_waitall_1 84994 16.7 7.123 7.840 7.123 7.840 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.764 7.799 calculate_rho_elec 110 8.6 0.227 0.227 7.763 7.798 sum_up_and_integrate 110 10.3 0.150 0.152 7.391 7.402 integrate_v_rspace 110 11.3 0.004 0.004 7.240 7.251 init_scf_run 11 5.9 0.000 0.001 5.737 5.737 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.737 5.737 make_m2s 4110 13.4 0.043 0.044 5.127 5.618 qs_ot_get_p 110 10.4 0.001 0.001 5.580 5.587 make_images 4110 14.4 0.877 0.932 4.939 5.428 cp_fm_cholesky_invert 11 10.9 5.345 5.349 5.345 5.349 pw_transfer 1331 11.6 0.075 0.076 5.308 5.316 dbcsr_mm_accdrv_process 11614 15.7 3.299 3.648 5.056 5.284 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 5.190 5.197 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.543 4.994 apply_single 110 13.6 0.000 0.000 4.543 4.994 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 4.462 4.871 fft_wrap_pw1pw2_140 451 13.1 0.220 0.222 4.515 4.526 ot_diis_step 99 11.5 0.015 0.015 4.405 4.406 fft3d_ps 1111 14.6 1.295 1.304 4.389 4.402 qs_ot_p2m_diag 48 11.0 0.151 0.155 3.951 3.958 multiply_cannon_sync_h2d 8220 15.4 3.951 3.954 3.951 3.954 density_rs2pw 110 9.6 0.004 0.004 3.801 3.828 grid_integrate_task_list 110 12.3 3.658 3.717 3.658 3.717 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.675 3.676 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.513 3.514 hybrid_alltoall_any 4261 16.3 0.258 0.560 2.768 3.506 make_images_data 4110 15.4 0.044 0.047 2.748 3.390 wfi_extrapolate 11 7.9 0.001 0.001 3.294 3.294 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.802 3.250 calculate_dm_sparse 110 9.5 0.001 0.001 3.116 3.137 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.926 2.926 cp_fm_diag_elpa_base 48 14.0 2.385 2.580 2.925 2.925 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.859 2.861 grid_collocate_task_list 110 9.6 2.630 2.653 2.630 2.653 potential_pw2rs 110 12.3 0.021 0.021 2.527 2.532 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.368 2.380 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.327 2.331 calculate_first_density_matrix 1 7.0 0.000 0.000 2.322 2.323 rs_pw_transfer 902 11.9 0.011 0.011 2.268 2.294 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.209 2.269 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.080 2.170 cp_fm_cholesky_decompose 22 10.9 1.993 2.013 1.993 2.013 mp_alltoall_d11v 2046 13.8 1.901 1.967 1.901 1.967 qs_create_task_list 11 7.9 0.000 0.001 1.901 1.949 generate_qs_task_list 11 8.9 0.738 0.792 1.900 1.948 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.941 1.942 jit_kernel_multiply 10 15.3 1.557 1.796 1.557 1.796 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.742 1.761 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=87.141000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1217.363636, yerr=56.568981 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b3f583d97791bedccf073d04f637ff05a010c441_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 628.244480E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175954870160 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 56898. MP_Allreduce 11228 786. MP_Sync 170 MP_Alltoall 2226 2523700. MP_SendRecv 24320 18752. MP_ISendRecv 24320 18752. MP_Wait 42476 MP_comm_split 83 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.053 0.094 204.922 204.924 qs_mol_dyn_low 1 2.0 0.003 0.006 204.038 204.052 qs_forces 11 3.9 0.004 0.008 203.910 203.953 qs_energies 11 4.9 0.001 0.002 198.334 198.385 scf_env_do_scf 11 5.9 0.001 0.001 181.725 181.729 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 161.131 161.133 dbcsr_multiply_generic 2507 12.6 0.177 0.183 124.012 124.675 velocity_verlet 10 3.0 0.001 0.001 122.838 122.839 qs_scf_new_mos 117 7.6 0.001 0.001 121.061 121.203 qs_scf_loop_do_ot 117 8.6 0.001 0.001 121.060 121.202 ot_scf_mini 117 9.6 0.003 0.003 114.492 114.638 multiply_cannon 2507 13.6 0.236 0.243 100.830 102.258 multiply_cannon_loop 2507 14.6 2.099 2.185 98.656 99.921 ot_mini 117 10.6 0.001 0.001 65.798 65.946 multiply_cannon_multrec 60168 15.6 33.299 35.518 41.617 43.339 qs_ot_get_derivative 117 11.6 0.001 0.002 40.912 41.051 rebuild_ks_matrix 128 8.3 0.001 0.001 33.733 34.004 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 33.732 34.004 mp_waitall_1 267128 16.5 27.852 31.136 27.852 31.136 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.315 30.556 multiply_cannon_sync_h2d 60168 15.6 27.536 29.736 27.536 29.736 qs_ot_get_p 128 10.4 0.001 0.001 26.763 27.082 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.292 25.189 apply_single 128 13.6 0.001 0.001 24.292 25.189 ot_diis_step 117 11.6 0.007 0.008 24.488 24.490 init_scf_loop 11 6.9 0.000 0.000 20.518 20.519 qs_ot_p2m_diag 83 11.4 0.077 0.091 20.144 20.258 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.082 19.190 multiply_cannon_metrocomm3 60168 15.6 0.117 0.122 15.768 18.223 cp_dbcsr_syevd 83 12.4 0.004 0.005 17.591 17.592 prepare_preconditioner 11 7.9 0.000 0.000 15.926 15.964 make_preconditioner 11 8.9 0.000 0.000 15.926 15.964 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.189 15.361 cp_fm_diag_elpa 83 13.4 0.000 0.000 14.797 14.802 cp_fm_redistribute_end 83 14.4 11.698 14.727 11.711 14.731 cp_fm_diag_elpa_base 83 14.4 2.960 14.300 3.007 14.466 sum_up_and_integrate 128 10.3 0.090 0.107 14.230 14.250 integrate_v_rspace 128 11.3 0.003 0.004 14.140 14.162 make_m2s 5014 13.6 0.105 0.114 13.609 13.975 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.725 13.841 calculate_rho_elec 128 8.7 0.045 0.064 13.725 13.840 make_images 5014 14.6 0.402 0.422 13.426 13.801 init_scf_run 11 5.9 0.000 0.001 12.410 12.411 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.410 12.411 density_rs2pw 128 9.7 0.006 0.007 7.104 10.292 mp_sum_l 7870 13.0 8.361 9.543 8.361 9.543 wfi_extrapolate 11 7.9 0.001 0.001 9.185 9.185 cp_fm_cholesky_invert 11 10.9 9.070 9.078 9.070 9.078 rs_pw_transfer 1046 11.9 0.016 0.018 5.670 8.852 dbcsr_mm_accdrv_process 124484 16.2 3.204 3.419 7.879 8.505 calculate_dm_sparse 128 9.5 0.001 0.001 8.369 8.440 pw_transfer 1547 11.6 0.076 0.097 7.982 8.222 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.776 8.021 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.882 8.018 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.907 8.016 grid_integrate_task_list 128 12.3 7.087 7.526 7.087 7.526 make_images_data 5014 15.6 0.069 0.076 6.636 7.446 multiply_cannon_metrocomm1 60168 15.6 0.091 0.094 5.754 7.154 fft_wrap_pw1pw2_140 523 13.2 0.446 0.511 6.786 7.087 hybrid_alltoall_any 5200 16.5 0.292 2.253 5.810 6.973 fft3d_ps 1291 14.7 2.111 2.885 6.559 6.733 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.631 6.639 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.767 5.876 mp_waitany 16020 13.9 2.656 5.771 2.656 5.771 grid_collocate_task_list 128 9.7 4.553 5.758 4.553 5.758 rs_pw_transfer_RS2PW_140 139 11.5 0.284 0.312 2.129 5.314 mp_alltoall_d11v 2415 14.1 4.053 5.104 4.053 5.104 potential_pw2rs 128 12.3 0.009 0.011 4.677 4.705 cp_fm_cholesky_decompose 22 10.9 4.604 4.618 4.604 4.618 mp_sum_d 4455 12.2 3.664 4.487 3.664 4.487 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=204.924000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=595.454545, yerr=7.414527 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b3f583d97791bedccf073d04f637ff05a010c441_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 2.199914E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 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 832.978944E+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 1473826772352 16777216 < size 68112 1190343475200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4107 56904. MP_Allreduce 11307 944. MP_Sync 170 MP_Alltoall 1983 5621117. MP_SendRecv 12126 47072. MP_ISendRecv 12126 47072. MP_Wait 26114 MP_comm_split 83 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 0.067 0.108 191.655 191.658 qs_mol_dyn_low 1 2.0 0.003 0.004 191.069 191.082 qs_forces 11 3.9 0.009 0.026 190.986 190.988 qs_energies 11 4.9 0.002 0.007 184.215 184.228 scf_env_do_scf 11 5.9 0.001 0.001 167.594 167.606 scf_env_do_scf_inner_loop 118 6.6 0.003 0.010 134.358 134.362 velocity_verlet 10 3.0 0.001 0.001 120.824 120.827 dbcsr_multiply_generic 2527 12.6 0.188 0.194 97.684 98.778 qs_scf_new_mos 118 7.6 0.001 0.001 94.162 94.552 qs_scf_loop_do_ot 118 8.6 0.001 0.001 94.161 94.551 ot_scf_mini 118 9.6 0.004 0.005 89.323 89.736 multiply_cannon 2527 13.6 0.487 0.540 77.959 82.169 multiply_cannon_loop 2527 14.6 1.260 1.300 74.609 76.811 ot_mini 118 10.6 0.001 0.001 49.814 50.198 mp_waitall_1 216438 16.6 23.975 38.281 23.975 38.281 multiply_cannon_multrec 30324 15.6 22.380 26.278 32.062 36.230 rebuild_ks_matrix 129 8.3 0.001 0.001 33.430 33.808 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.020 33.429 33.808 init_scf_loop 11 6.9 0.001 0.005 33.151 33.151 qs_ks_update_qs_env 129 7.6 0.001 0.001 30.128 30.474 prepare_preconditioner 11 7.9 0.000 0.000 28.747 28.823 make_preconditioner 11 8.9 0.000 0.001 28.747 28.823 multiply_cannon_metrocomm3 30324 15.6 0.096 0.101 15.491 28.253 qs_ot_get_derivative 118 11.6 0.001 0.001 27.763 28.190 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.463 27.999 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 22.097 23.185 apply_single 129 13.6 0.001 0.001 22.097 23.185 ot_diis_step 118 11.6 0.014 0.015 21.876 21.878 multiply_cannon_sync_h2d 30324 15.6 19.616 21.872 19.616 21.872 qs_ot_get_p 129 10.4 0.001 0.001 21.118 21.727 cp_fm_cholesky_invert 11 10.9 16.653 16.665 16.653 16.665 qs_ot_p2m_diag 83 11.4 0.187 0.215 16.393 16.427 make_m2s 5054 13.6 0.092 0.096 14.089 15.749 make_images 5054 14.6 1.157 1.344 13.875 15.535 cp_dbcsr_syevd 83 12.4 0.005 0.006 15.251 15.252 sum_up_and_integrate 129 10.3 0.117 0.134 14.653 14.682 integrate_v_rspace 129 11.3 0.003 0.003 14.535 14.568 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.028 14.067 calculate_rho_elec 129 8.7 0.089 0.106 14.028 14.066 cp_fm_diag_elpa 83 13.4 0.000 0.001 12.187 12.197 cp_fm_redistribute_end 83 14.4 7.118 12.128 7.129 12.131 cp_fm_diag_elpa_base 83 14.4 4.745 11.578 4.984 11.990 init_scf_run 11 5.9 0.000 0.001 11.611 11.613 scf_env_initial_rho_setup 11 6.9 0.001 0.002 11.611 11.613 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 10.814 11.116 make_images_data 5054 15.6 0.067 0.074 8.517 10.342 multiply_cannon_metrocomm4 27797 15.6 0.098 0.111 3.757 10.210 density_rs2pw 129 9.7 0.006 0.007 7.434 9.843 mp_irecv_dv 70031 16.3 3.559 9.821 3.559 9.821 dbcsr_mm_accdrv_process 62734 16.2 4.475 5.298 9.129 9.732 hybrid_alltoall_any 5240 16.5 0.347 1.523 7.277 9.719 pw_transfer 1559 11.6 0.087 0.106 8.821 8.897 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 8.593 8.669 wfi_extrapolate 11 7.9 0.001 0.001 8.368 8.369 rs_pw_transfer 1054 12.0 0.014 0.016 5.719 8.175 fft_wrap_pw1pw2_140 527 13.2 0.476 0.533 7.565 7.666 grid_integrate_task_list 129 12.3 7.256 7.609 7.256 7.609 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 6.322 7.055 cp_fm_cholesky_decompose 22 10.9 6.921 7.008 6.921 7.008 fft3d_ps 1301 14.7 2.778 2.946 6.947 6.989 calculate_dm_sparse 129 9.5 0.001 0.001 6.482 6.622 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.262 6.270 grid_collocate_task_list 129 9.7 4.738 5.931 4.738 5.931 mp_sum_l 7930 13.1 3.854 5.646 3.854 5.646 qs_ot_get_orbitals 118 10.6 0.001 0.001 5.393 5.462 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.295 5.420 mp_waitany 11836 13.9 2.504 4.999 2.504 4.999 potential_pw2rs 129 12.3 0.015 0.018 4.940 4.957 mp_allgather_i34 2527 14.6 1.906 4.870 1.906 4.870 mp_alltoall_d11v 2423 14.1 3.831 4.649 3.831 4.649 rs_pw_transfer_RS2PW_140 140 11.5 0.383 0.426 2.181 4.615 dbcsr_complete_redistribute 395 12.7 0.867 1.011 3.270 4.115 cp_fm_upper_to_full 105 14.2 3.072 3.994 3.072 3.994 mp_sum_d 4489 12.2 2.663 3.955 2.663 3.955 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=191.658000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=794.000000, yerr=3.162278 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b3f583d97791bedccf073d04f637ff05a010c441_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420243808256 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 11528908111872 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 93.514772E+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 6755942624 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 954.294272E+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 931531401248 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4103 56951. MP_Allreduce 11296 984. MP_Sync 170 MP_Alltoall 1712 9388896. MP_SendRecv 7936 75008. MP_ISendRecv 7936 75008. MP_Wait 21820 MP_comm_split 83 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 0.045 0.093 176.456 176.457 qs_mol_dyn_low 1 2.0 0.003 0.004 175.949 175.962 qs_forces 11 3.9 0.004 0.006 175.813 175.825 qs_energies 11 4.9 0.012 0.043 169.212 169.228 scf_env_do_scf 11 5.9 0.001 0.002 153.737 153.738 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 118.393 118.395 velocity_verlet 10 3.0 0.001 0.002 113.091 113.094 dbcsr_multiply_generic 2507 12.6 0.182 0.187 80.694 81.909 qs_scf_new_mos 117 7.6 0.001 0.001 80.376 80.656 qs_scf_loop_do_ot 117 8.6 0.001 0.001 80.375 80.655 ot_scf_mini 117 9.6 0.004 0.004 76.258 76.588 multiply_cannon 2507 13.6 0.499 0.520 61.132 65.322 multiply_cannon_loop 2507 14.6 0.854 0.884 58.137 61.225 ot_mini 117 10.6 0.001 0.001 42.227 42.563 init_scf_loop 11 6.9 0.001 0.003 35.245 35.254 mp_waitall_1 170520 16.6 24.298 33.467 24.298 33.467 prepare_preconditioner 11 7.9 0.000 0.000 31.235 31.278 make_preconditioner 11 8.9 0.000 0.001 31.235 31.278 rebuild_ks_matrix 128 8.3 0.001 0.001 30.524 30.937 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.019 30.523 30.936 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.853 30.272 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.523 27.899 multiply_cannon_multrec 20056 15.6 13.628 16.416 22.270 24.990 multiply_cannon_metrocomm3 20056 15.6 0.061 0.065 14.739 24.194 qs_ot_get_derivative 117 11.6 0.001 0.002 22.670 23.003 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.652 20.607 apply_single 128 13.6 0.001 0.001 19.651 20.606 qs_ot_get_p 128 10.4 0.001 0.002 19.571 19.996 ot_diis_step 117 11.6 0.018 0.018 19.452 19.452 multiply_cannon_sync_h2d 20056 15.6 14.305 16.109 14.305 16.109 make_m2s 5014 13.6 0.082 0.088 14.542 15.445 qs_ot_p2m_diag 83 11.4 0.265 0.272 15.309 15.316 make_images 5014 14.6 1.161 1.242 14.306 15.209 sum_up_and_integrate 128 10.3 0.133 0.145 14.597 14.623 integrate_v_rspace 128 11.3 0.004 0.005 14.464 14.493 cp_fm_cholesky_invert 11 10.9 14.363 14.371 14.363 14.371 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.271 14.313 calculate_rho_elec 128 8.7 0.132 0.147 14.270 14.312 cp_dbcsr_syevd 83 12.4 0.005 0.005 14.298 14.299 cp_fm_diag_elpa 83 13.4 0.000 0.001 11.154 11.156 cp_fm_redistribute_end 83 14.4 4.216 11.090 4.229 11.094 cp_fm_diag_elpa_base 83 14.4 6.402 10.439 6.844 10.971 init_scf_run 11 5.9 0.000 0.001 10.444 10.445 scf_env_initial_rho_setup 11 6.9 0.013 0.046 10.444 10.445 make_images_data 5014 15.6 0.063 0.071 8.978 10.186 hybrid_alltoall_any 5200 16.5 0.437 1.998 7.720 9.282 density_rs2pw 128 9.7 0.006 0.006 7.190 9.211 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 8.870 9.103 pw_transfer 1547 11.6 0.086 0.107 8.856 8.971 multiply_cannon_metrocomm4 17549 15.6 0.063 0.072 3.445 8.800 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.631 8.753 mp_irecv_dv 50230 16.2 3.320 8.556 3.320 8.556 dbcsr_mm_accdrv_process 41502 16.2 4.380 5.326 8.093 8.261 cp_fm_upper_to_full 105 14.5 6.186 7.942 6.186 7.942 fft_wrap_pw1pw2_140 523 13.2 0.476 0.521 7.641 7.764 grid_integrate_task_list 128 12.3 7.307 7.697 7.307 7.697 wfi_extrapolate 11 7.9 0.001 0.001 7.383 7.383 cp_fm_cholesky_decompose 22 10.9 7.314 7.340 7.314 7.340 rs_pw_transfer 1046 11.9 0.013 0.015 5.184 7.191 fft3d_ps 1291 14.7 2.664 2.895 6.897 6.982 dbcsr_complete_redistribute 395 12.7 1.185 1.254 4.620 6.379 calculate_dm_sparse 128 9.5 0.001 0.001 5.703 5.808 grid_collocate_task_list 128 9.7 4.903 5.769 4.903 5.769 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.420 5.428 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.575 5.273 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.453 5.217 potential_pw2rs 128 12.3 0.020 0.023 4.772 4.786 mp_alltoall_d11v 2415 14.1 4.143 4.731 4.143 4.731 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.486 4.662 mp_sum_l 7870 13.0 3.228 4.579 3.228 4.579 mp_allgather_i34 2507 14.6 1.525 4.461 1.525 4.461 mp_waitany 11748 13.9 2.304 4.296 2.304 4.296 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.363 4.084 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.018 4.042 mp_alltoall_i22 716 14.1 1.970 3.926 1.970 3.926 rs_pw_transfer_RS2PW_140 139 11.5 0.333 0.351 1.910 3.893 qs_energies_init_hamiltonians 11 5.9 0.001 0.005 3.758 3.767 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.515 3.554 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=176.457000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=902.454545, yerr=18.207709 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b3f583d97791bedccf073d04f637ff05a010c441_performance_tests/18/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 4.353791E+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 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 1.154511E+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 1.770320E+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 665379475120 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4103 56948. MP_Allreduce 11296 1066. MP_Sync 170 MP_Alltoall 1712 12503107. MP_SendRecv 5888 75008. MP_ISendRecv 5888 75008. MP_Wait 22442 MP_comm_split 83 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 0.022 0.058 187.971 187.973 qs_mol_dyn_low 1 2.0 0.003 0.004 187.439 187.452 qs_forces 11 3.9 0.003 0.006 187.328 187.344 qs_energies 11 4.9 0.002 0.011 180.169 180.183 scf_env_do_scf 11 5.9 0.001 0.003 163.458 163.465 velocity_verlet 10 3.0 0.001 0.002 124.551 124.554 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 116.865 116.866 qs_scf_new_mos 117 7.6 0.001 0.001 79.435 79.683 qs_scf_loop_do_ot 117 8.6 0.001 0.001 79.434 79.683 dbcsr_multiply_generic 2507 12.6 0.191 0.195 78.720 79.351 ot_scf_mini 117 9.6 0.003 0.004 74.996 75.295 multiply_cannon 2507 13.6 0.558 0.592 54.424 57.423 multiply_cannon_loop 2507 14.6 1.184 1.211 50.717 52.263 init_scf_loop 11 6.9 0.001 0.007 46.475 46.476 prepare_preconditioner 11 7.9 0.000 0.000 42.382 42.416 make_preconditioner 11 8.9 0.000 0.002 42.382 42.416 ot_mini 117 10.6 0.001 0.001 41.944 42.246 make_full_inverse_cholesky 11 9.9 0.000 0.000 36.105 40.999 multiply_cannon_multrec 30084 15.6 14.221 19.365 26.167 30.999 rebuild_ks_matrix 128 8.3 0.001 0.001 29.859 30.131 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.020 29.858 30.130 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.912 27.161 mp_waitall_1 147882 16.7 16.625 26.863 16.625 26.863 qs_ot_get_derivative 117 11.6 0.001 0.002 22.363 22.662 make_m2s 5014 13.6 0.096 0.100 19.972 21.019 make_images 5014 14.6 1.932 2.275 19.667 20.714 qs_ot_get_p 128 10.4 0.001 0.001 19.323 19.677 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 18.948 19.476 apply_single 128 13.6 0.001 0.001 18.948 19.476 ot_diis_step 117 11.6 0.017 0.018 19.454 19.456 cp_fm_upper_to_full 105 14.7 11.401 16.634 11.401 16.634 cp_fm_cholesky_invert 11 10.9 16.151 16.160 16.151 16.160 qs_ot_p2m_diag 83 11.4 0.342 0.389 15.250 15.301 multiply_cannon_metrocomm3 30084 15.6 0.048 0.050 6.089 14.870 sum_up_and_integrate 128 10.3 0.140 0.152 14.697 14.725 integrate_v_rspace 128 11.3 0.003 0.005 14.556 14.587 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.526 14.562 calculate_rho_elec 128 8.7 0.175 0.190 14.526 14.561 cp_dbcsr_syevd 83 12.4 0.005 0.005 13.951 13.953 multiply_cannon_sync_h2d 30084 15.6 11.815 12.855 11.815 12.855 dbcsr_complete_redistribute 395 12.7 1.511 1.625 8.815 12.447 make_images_data 5014 15.6 0.066 0.071 10.657 12.293 dbcsr_mm_accdrv_process 62264 16.2 7.303 8.227 11.523 12.042 hybrid_alltoall_any 5200 16.5 0.526 2.202 9.445 11.325 copy_fm_to_dbcsr 209 11.7 0.002 0.002 7.475 11.091 cp_fm_diag_elpa 83 13.4 0.000 0.001 10.808 10.811 init_scf_run 11 5.9 0.000 0.001 10.774 10.775 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.774 10.775 cp_fm_redistribute_end 83 14.4 1.872 10.723 1.889 10.729 cp_fm_diag_elpa_base 83 14.4 8.188 10.060 8.807 10.586 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.253 9.801 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.254 9.488 pw_transfer 1547 11.6 0.088 0.106 9.219 9.295 mp_alltoall_i22 716 14.1 5.484 9.128 5.484 9.128 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.991 9.069 density_rs2pw 128 9.7 0.006 0.006 7.036 8.582 fft_wrap_pw1pw2_140 523 13.2 0.481 0.490 7.953 8.040 grid_integrate_task_list 128 12.3 7.529 7.953 7.529 7.953 wfi_extrapolate 11 7.9 0.001 0.001 7.654 7.654 cp_fm_cholesky_decompose 22 10.9 7.520 7.611 7.520 7.611 fft3d_ps 1291 14.7 2.777 2.886 7.219 7.311 multiply_cannon_metrocomm4 25070 15.6 0.077 0.086 2.752 7.179 mp_irecv_dv 76098 16.2 2.606 6.927 2.606 6.927 rs_pw_transfer 1046 11.9 0.013 0.014 4.747 6.300 calculate_dm_sparse 128 9.5 0.001 0.001 6.090 6.181 grid_collocate_task_list 128 9.7 5.068 5.802 5.068 5.802 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.460 5.533 mp_alltoall_d11v 2415 14.1 4.604 5.456 4.604 5.456 potential_pw2rs 128 12.3 0.022 0.023 4.605 4.617 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.426 4.506 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 4.477 4.483 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.310 4.393 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.185 4.242 mp_waitany 14952 13.9 2.269 3.818 2.269 3.818 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=187.973000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1089.636364, yerr=18.217238 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b3f583d97791bedccf073d04f637ff05a010c441_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430457200640 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986252263424 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992001093632 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613077360640 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239162695680 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239162695680 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.232994E+12 0.0% 0.0% 100.0% flops max/rank 5.910120E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806382528 0.0% 0.0% 100.0% number of processed stacks 1976928 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.9 marketing flops 145.650931E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 1.502900E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 242784 MPI messages size (bytes): total size 1.341806E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.526746E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 115008 60297314304 4194304 < size <= 16777216 105840 554906419200 16777216 < size 20352 726592214928 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 9010 51. MP_Alltoall 9724 794507. MP_ISend 40420 2100460. MP_IRecv 40420 2099564. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 57630. MP_Allreduce 11104 1171. MP_Sync 88 MP_Alltoall 1724 18848081. 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 0.041 0.076 177.600 177.616 qs_mol_dyn_low 1 2.0 0.003 0.009 177.064 177.079 qs_forces 11 3.9 0.006 0.019 176.927 176.954 qs_energies 11 4.9 0.001 0.002 169.351 169.390 scf_env_do_scf 11 5.9 0.001 0.001 151.871 151.881 velocity_verlet 10 3.0 0.001 0.001 116.136 116.141 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 115.601 115.602 qs_scf_new_mos 118 7.6 0.001 0.001 76.246 76.323 qs_scf_loop_do_ot 118 8.6 0.001 0.001 76.245 76.322 dbcsr_multiply_generic 2529 12.6 0.185 0.192 75.155 75.776 ot_scf_mini 118 9.6 0.004 0.004 71.766 71.852 multiply_cannon 2529 13.6 0.590 0.621 55.277 59.617 multiply_cannon_loop 2529 14.6 0.448 0.461 50.520 51.696 ot_mini 118 10.6 0.001 0.001 40.271 40.358 init_scf_loop 11 6.9 0.000 0.000 36.118 36.120 mp_waitall_1 126876 16.7 26.379 34.031 26.379 34.031 prepare_preconditioner 11 7.9 0.000 0.000 32.160 32.182 make_preconditioner 11 8.9 0.000 0.000 32.160 32.182 rebuild_ks_matrix 129 8.3 0.001 0.001 30.383 30.508 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.018 30.383 30.507 make_full_inverse_cholesky 11 9.9 0.000 0.000 30.021 30.275 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.632 27.748 multiply_cannon_multrec 10116 15.6 10.492 15.593 18.006 21.887 multiply_cannon_metrocomm3 10116 15.6 0.024 0.025 12.818 20.388 ot_diis_step 118 11.6 0.021 0.037 20.270 20.270 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.756 20.175 apply_single 129 13.6 0.001 0.001 19.755 20.175 qs_ot_get_derivative 118 11.6 0.001 0.002 19.933 20.017 make_m2s 5058 13.6 0.067 0.072 16.260 18.700 qs_ot_get_p 129 10.4 0.001 0.001 18.415 18.543 make_images 5058 14.6 2.297 2.640 15.949 18.382 cp_fm_cholesky_invert 11 10.9 17.981 17.986 17.981 17.986 qs_rho_update_rho_low 129 7.7 0.001 0.001 15.631 15.697 calculate_rho_elec 129 8.7 0.261 0.271 15.631 15.696 sum_up_and_integrate 129 10.3 0.182 0.192 15.300 15.356 integrate_v_rspace 129 11.3 0.004 0.004 15.117 15.182 qs_ot_p2m_diag 84 11.4 0.501 0.507 14.629 14.644 cp_dbcsr_syevd 84 12.4 0.005 0.005 13.480 13.482 make_images_data 5058 15.6 0.055 0.063 9.880 12.434 hybrid_alltoall_any 5245 16.5 0.821 3.832 9.726 12.132 multiply_cannon_sync_h2d 10116 15.6 11.623 12.028 11.623 12.028 init_scf_run 11 5.9 0.000 0.001 10.585 10.585 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.585 10.585 cp_fm_diag_elpa 84 13.4 0.000 0.000 10.395 10.408 cp_fm_diag_elpa_base 84 14.4 10.096 10.168 10.388 10.401 pw_transfer 1559 11.6 0.086 0.093 10.015 10.058 fft_wrap_pw1pw2 1301 12.7 0.010 0.010 9.790 9.837 fft_wrap_pw1pw2_140 527 13.2 0.499 0.520 8.639 8.688 grid_integrate_task_list 129 12.3 7.811 8.355 7.811 8.355 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 8.062 8.128 density_rs2pw 129 9.7 0.005 0.006 7.145 8.022 fft3d_ps 1301 14.7 2.725 2.804 7.943 7.961 cp_fm_cholesky_decompose 22 10.9 7.819 7.933 7.819 7.933 dbcsr_mm_accdrv_process 20934 16.1 2.744 3.494 7.139 7.820 multiply_cannon_metrocomm1 10116 15.6 0.029 0.030 4.416 7.730 mp_allgather_i34 2529 14.6 2.951 7.586 2.951 7.586 wfi_extrapolate 11 7.9 0.001 0.001 7.392 7.392 calculate_dm_sparse 129 9.5 0.001 0.001 6.157 6.244 grid_collocate_task_list 129 9.7 5.436 6.138 5.436 6.138 mp_alltoall_d11v 2429 14.1 5.066 6.017 5.066 6.017 dbcsr_complete_redistribute 397 12.7 2.181 2.289 5.312 5.769 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.370 5.379 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 5.339 5.367 multiply_cannon_metrocomm4 7587 15.6 0.024 0.027 1.899 5.185 mp_irecv_dv 29102 15.9 1.862 5.101 1.862 5.101 rs_pw_transfer 1054 12.0 0.012 0.013 4.204 5.090 potential_pw2rs 129 12.3 0.027 0.028 4.645 4.661 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.180 4.213 copy_fm_to_dbcsr 210 11.7 0.002 0.002 3.528 3.877 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.552 3.852 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.775 3.813 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.668 3.679 copy_dbcsr_to_fm 187 11.8 0.004 0.004 3.523 3.625 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=177.616000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1397.363636, yerr=44.719881 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b3f583d97791bedccf073d04f637ff05a010c441_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 11.785520E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 1979552 0.0% 0.0% 100.0% average stack size 0.0 0.0 3438.3 marketing flops 145.646636E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 2.964697E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101080 MPI messages size (bytes): total size 1.144901E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.326685E+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 45568 35366371328 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726591885696 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4024 58831. MP_Allreduce 11057 1512. MP_Sync 87 MP_Alltoall 1724 36993716. 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 0.041 0.088 301.178 301.184 qs_mol_dyn_low 1 2.0 0.003 0.004 300.582 300.594 qs_forces 11 3.9 0.003 0.003 300.481 300.489 qs_energies 11 4.9 0.001 0.001 291.056 291.066 scf_env_do_scf 11 5.9 0.001 0.001 268.773 268.790 velocity_verlet 10 3.0 0.001 0.001 216.657 216.665 scf_env_do_scf_inner_loop 118 6.6 0.004 0.009 143.341 143.343 init_scf_loop 11 6.9 0.000 0.000 125.174 125.179 prepare_preconditioner 11 7.9 0.000 0.000 120.019 120.048 make_preconditioner 11 8.9 0.000 0.000 120.019 120.048 make_full_inverse_cholesky 11 9.9 0.000 0.000 96.306 117.177 qs_scf_new_mos 118 7.6 0.001 0.001 91.445 91.543 qs_scf_loop_do_ot 118 8.6 0.001 0.001 91.444 91.542 ot_scf_mini 118 9.6 0.004 0.004 86.616 86.654 dbcsr_multiply_generic 2527 12.6 0.229 0.249 84.452 85.110 cp_fm_upper_to_full 105 14.8 53.986 76.620 53.986 76.620 multiply_cannon 2527 13.6 0.704 0.748 59.201 60.020 multiply_cannon_loop 2527 14.6 0.477 0.489 55.356 56.863 ot_mini 118 10.6 0.001 0.001 45.113 45.144 dbcsr_complete_redistribute 395 12.7 3.984 4.023 29.356 42.253 rebuild_ks_matrix 129 8.3 0.001 0.001 39.134 39.210 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.018 39.133 39.210 copy_fm_to_dbcsr 209 11.7 0.001 0.002 25.988 38.852 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 23.668 36.435 qs_ks_update_qs_env 129 7.6 0.001 0.001 36.068 36.143 mp_alltoall_i22 718 14.1 21.467 34.213 21.467 34.213 cp_fm_cholesky_invert 11 10.9 33.270 33.276 33.270 33.276 mp_waitall_1 104500 16.8 28.842 32.428 28.842 32.428 qs_ot_get_p 129 10.4 0.001 0.001 25.793 25.861 qs_ot_get_derivative 118 11.6 0.002 0.002 24.428 24.469 qs_ot_p2m_diag 83 11.4 0.878 0.884 21.675 21.706 make_m2s 5054 13.6 0.077 0.080 20.564 21.697 make_images 5054 14.6 3.808 3.939 20.080 21.216 qs_rho_update_rho_low 129 7.7 0.001 0.001 21.035 21.058 calculate_rho_elec 129 8.7 0.486 0.487 21.034 21.058 ot_diis_step 118 11.6 0.022 0.022 20.648 20.648 multiply_cannon_metrocomm3 10108 15.6 0.025 0.025 18.913 20.271 sum_up_and_integrate 129 10.3 0.324 0.326 20.093 20.171 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.878 20.100 apply_single 129 13.6 0.001 0.001 19.878 20.100 cp_dbcsr_syevd 83 12.4 0.006 0.006 19.952 19.954 integrate_v_rspace 129 11.3 0.004 0.004 19.769 19.847 multiply_cannon_multrec 10108 15.6 10.591 12.454 18.074 18.130 cp_fm_diag_elpa 83 13.4 0.000 0.000 16.741 16.742 cp_fm_diag_elpa_base 83 14.4 12.346 13.875 16.738 16.738 multiply_cannon_sync_h2d 10108 15.6 15.780 15.806 15.780 15.806 hybrid_alltoall_any 5240 16.5 1.311 3.051 11.299 13.152 make_images_data 5054 15.6 0.063 0.068 11.184 13.027 pw_transfer 1559 11.6 0.093 0.094 12.510 12.522 fft_wrap_pw1pw2 1301 12.7 0.011 0.011 12.275 12.287 init_scf_run 11 5.9 0.000 0.001 12.257 12.258 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.257 12.258 fft_wrap_pw1pw2_140 527 13.2 0.545 0.547 10.910 10.924 fft3d_ps 1301 14.7 2.760 2.769 10.301 10.325 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.455 9.493 mp_alltoall_d11v 2423 14.1 8.277 9.345 8.277 9.345 wfi_extrapolate 11 7.9 0.001 0.001 9.129 9.130 dbcsr_mm_accdrv_process 20926 16.1 3.866 5.714 7.244 9.096 cp_fm_cholesky_decompose 22 10.9 8.812 8.842 8.812 8.842 grid_integrate_task_list 129 12.3 8.641 8.814 8.641 8.814 density_rs2pw 129 9.7 0.005 0.005 8.490 8.638 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 8.045 8.046 calculate_dm_sparse 129 9.5 0.001 0.001 6.694 6.761 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.402 6.456 grid_collocate_task_list 129 9.7 6.411 6.451 6.411 6.451 rs_scatter_matrices 140 9.7 3.733 4.620 6.104 6.339 copy_dbcsr_to_fm 186 11.8 0.004 0.004 6.032 6.153 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=301.184000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2647.818182, yerr=160.464563 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b3f583d97791bedccf073d04f637ff05a010c441_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 1.261777E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255669. MP_Allreduce 3059 6274. MP_Sync 4 MP_Alltoall 54 6805335. MP_SendRecv 285 19200. MP_ISendRecv 285 19200. MP_Wait 1017 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.020 0.058 85.298 85.300 qs_energies 1 2.0 0.000 0.000 84.820 84.834 ls_scf 1 3.0 0.000 0.001 83.902 83.915 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.857 73.064 multiply_cannon 111 7.7 0.017 0.020 56.146 57.036 multiply_cannon_loop 111 8.7 0.203 0.224 52.626 53.722 ls_scf_main 1 4.0 0.000 0.000 52.343 52.343 density_matrix_trs4 2 5.0 0.002 0.003 46.751 46.837 ls_scf_init_scf 1 4.0 0.000 0.001 28.510 28.510 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.381 27.442 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.156 25.179 mp_waitall_1 11031 10.9 22.454 24.507 22.454 24.507 multiply_cannon_multrec 2664 9.7 8.197 8.962 15.507 17.251 multiply_cannon_sync_h2d 2664 9.7 13.812 15.336 13.812 15.336 make_m2s 222 7.7 0.008 0.011 13.123 13.614 make_images 222 8.7 0.099 0.109 13.101 13.595 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 9.602 12.078 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.544 8.691 make_images_data 222 9.7 0.004 0.005 7.634 8.255 hybrid_alltoall_any 227 10.6 0.215 1.853 6.521 8.088 dbcsr_mm_accdrv_process 4760 10.4 0.510 0.635 6.927 8.022 dbcsr_mm_accdrv_process_sort 4760 11.4 6.200 7.074 6.200 7.074 calculate_norms 4752 9.8 5.538 6.166 5.538 6.166 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.168 5.335 mp_sum_l 807 5.4 3.045 4.658 3.045 4.658 make_images_sizes 222 9.7 0.000 0.000 0.781 3.596 mp_alltoall_i44 222 10.7 0.781 3.596 0.781 3.596 multiply_cannon_metrocomm4 2442 9.7 0.012 0.014 2.062 3.367 mp_irecv_dv 6231 10.9 2.043 3.337 2.043 3.337 arnoldi_extremal 4 6.8 0.000 0.000 3.231 3.271 arnoldi_normal_ev 4 7.8 0.001 0.003 3.231 3.271 build_subspace 16 8.4 0.009 0.012 3.124 3.126 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.202 3.123 ls_scf_post 1 4.0 0.000 0.000 3.049 3.062 ls_scf_store_result 1 5.0 0.000 0.000 2.870 2.909 dbcsr_special_finalize 555 9.7 0.005 0.006 2.354 2.777 dbcsr_merge_single_wm 555 10.7 0.486 0.630 2.346 2.769 make_images_pack 222 9.7 2.207 2.620 2.209 2.622 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.314 2.560 dbcsr_sort_data 658 11.4 2.118 2.498 2.118 2.498 dbcsr_matrix_vector_mult_local 304 10.0 2.065 2.463 2.067 2.465 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.342 2.414 buffer_matrices_ensure_size 222 8.7 1.816 2.164 1.816 2.164 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.824 1.825 rebuild_ks_matrix 3 7.3 0.000 0.000 1.814 1.815 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.003 1.814 1.815 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.300000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1142.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b3f583d97791bedccf073d04f637ff05a010c441_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 2.104689E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 10339. MP_Sync 4 MP_Alltoall 47 15335933. MP_SendRecv 141 57600. MP_ISendRecv 141 57600. MP_Wait 687 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.025 0.044 90.846 90.848 qs_energies 1 2.0 0.000 0.001 90.388 90.391 ls_scf 1 3.0 0.000 0.000 89.080 89.083 dbcsr_multiply_generic 111 6.7 0.015 0.016 75.075 75.401 multiply_cannon 111 7.7 0.028 0.043 53.184 56.564 ls_scf_main 1 4.0 0.000 0.001 54.830 54.836 multiply_cannon_loop 111 8.7 0.115 0.121 49.934 52.626 density_matrix_trs4 2 5.0 0.002 0.006 49.140 49.313 ls_scf_init_scf 1 4.0 0.000 0.002 30.697 30.699 mp_waitall_1 9105 10.9 21.153 30.473 21.153 30.473 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.476 29.578 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.040 27.051 multiply_cannon_multrec 1332 9.7 13.097 16.590 22.283 26.916 multiply_cannon_metrocomm3 1332 9.7 0.006 0.008 11.826 21.325 make_m2s 222 7.7 0.006 0.007 15.242 15.935 make_images 222 8.7 1.571 1.967 15.212 15.906 dbcsr_mm_accdrv_process 4041 10.4 0.271 0.458 8.782 10.354 dbcsr_mm_accdrv_process_sort 4041 11.4 8.374 9.901 8.374 9.901 make_images_data 222 9.7 0.004 0.004 8.718 9.635 hybrid_alltoall_any 227 10.6 0.522 2.488 8.148 9.189 mp_sum_l 807 5.4 5.333 8.098 5.333 8.098 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.239 7.657 mp_irecv_dv 3311 11.0 3.219 7.593 3.219 7.593 calculate_norms 2376 9.8 5.998 6.888 5.998 6.888 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.135 6.465 multiply_cannon_sync_h2d 1332 9.7 4.786 6.279 4.786 6.279 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.055 5.234 arnoldi_extremal 4 6.8 0.000 0.002 4.657 4.683 arnoldi_normal_ev 4 7.8 0.001 0.005 4.657 4.681 build_subspace 16 8.4 0.014 0.021 4.393 4.396 ls_scf_post 1 4.0 0.000 0.001 3.552 3.557 dbcsr_matrix_vector_mult 304 9.0 0.009 0.021 3.143 3.394 ls_scf_store_result 1 5.0 0.000 0.000 3.218 3.372 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.241 3.241 dbcsr_matrix_vector_mult_local 304 10.0 2.745 3.220 2.747 3.222 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.602 2.704 make_images_pack 222 9.7 2.020 2.432 2.022 2.434 mp_allgather_i34 111 8.7 0.956 2.382 0.956 2.382 dbcsr_sort_data 436 11.2 1.855 2.073 1.855 2.073 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 1.897 1.899 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.003 1.897 1.899 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=90.848000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1750.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b3f583d97791bedccf073d04f637ff05a010c441_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 2.790658E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265470. MP_Allreduce 3058 11181. MP_Sync 4 MP_Alltoall 47 23526250. MP_SendRecv 93 57600. MP_ISendRecv 93 57600. MP_Wait 639 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.036 0.051 94.755 94.758 qs_energies 1 2.0 0.000 0.000 94.148 94.153 ls_scf 1 3.0 0.000 0.001 92.739 92.742 dbcsr_multiply_generic 111 6.7 0.017 0.025 77.404 77.683 ls_scf_main 1 4.0 0.000 0.000 58.313 58.318 multiply_cannon 111 7.7 0.037 0.069 53.294 57.487 multiply_cannon_loop 111 8.7 0.099 0.105 49.754 54.174 density_matrix_trs4 2 5.0 0.002 0.003 52.307 52.524 mp_waitall_1 7281 11.0 24.458 34.616 24.458 34.616 ls_scf_init_scf 1 4.0 0.000 0.002 30.849 30.852 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.622 29.724 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.207 27.225 multiply_cannon_multrec 888 9.7 12.491 15.091 20.981 24.178 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.437 23.898 make_m2s 222 7.7 0.006 0.007 17.097 18.358 make_images 222 8.7 1.969 2.312 17.059 18.319 hybrid_alltoall_any 227 10.6 0.618 2.853 9.474 10.840 make_images_data 222 9.7 0.004 0.004 9.776 10.839 mp_sum_l 807 5.4 5.457 9.621 5.457 9.621 dbcsr_mm_accdrv_process 3754 10.4 0.247 0.411 8.015 9.211 dbcsr_mm_accdrv_process_sort 3754 11.4 7.644 8.800 7.644 8.800 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.236 7.710 multiply_cannon_sync_h2d 888 9.7 6.113 7.543 6.113 7.543 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.868 7.360 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.494 7.194 mp_irecv_dv 2335 11.1 2.479 7.145 2.479 7.145 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.919 5.134 arnoldi_extremal 4 6.8 0.000 0.000 5.067 5.089 arnoldi_normal_ev 4 7.8 0.001 0.005 5.067 5.089 build_subspace 16 8.4 0.014 0.021 4.748 4.754 calculate_norms 1584 9.8 4.256 4.648 4.256 4.648 mp_allgather_i34 111 8.7 1.380 3.857 1.380 3.857 dbcsr_matrix_vector_mult 304 9.0 0.009 0.020 3.448 3.764 dbcsr_matrix_vector_mult_local 304 10.0 3.037 3.599 3.039 3.601 ls_scf_post 1 4.0 0.000 0.001 3.578 3.584 ls_scf_store_result 1 5.0 0.000 0.000 3.302 3.386 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.872 2.968 make_images_sizes 222 9.7 0.000 0.000 1.018 2.259 mp_alltoall_i44 222 10.7 1.017 2.258 1.017 2.258 dbcsr_sort_data 325 11.1 1.875 2.111 1.875 2.111 make_images_pack 222 9.7 1.805 2.082 1.808 2.085 dbcsr_data_release 9322 10.9 1.312 1.975 1.312 1.975 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.928 1.930 rebuild_ks_matrix 3 7.3 0.000 0.000 1.910 1.912 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.910 1.912 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=94.758000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2216.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b3f583d97791bedccf073d04f637ff05a010c441_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 3.314332E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 13371. MP_Sync 4 MP_Alltoall 47 30278988. MP_SendRecv 69 86400. MP_ISendRecv 69 86400. MP_Wait 531 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.041 0.059 96.866 96.868 qs_energies 1 2.0 0.000 0.000 96.204 96.210 ls_scf 1 3.0 0.000 0.001 94.557 94.561 dbcsr_multiply_generic 111 6.7 0.017 0.017 78.315 78.557 ls_scf_main 1 4.0 0.000 0.000 58.683 58.689 multiply_cannon 111 7.7 0.052 0.146 51.420 56.313 density_matrix_trs4 2 5.0 0.002 0.003 52.556 52.674 multiply_cannon_loop 111 8.7 0.114 0.126 46.358 49.645 ls_scf_init_scf 1 4.0 0.000 0.002 32.629 32.630 ls_scf_init_matrix_S 1 5.0 0.000 0.001 31.394 31.464 mp_waitall_1 6369 11.0 22.543 29.279 22.543 29.279 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.942 28.956 multiply_cannon_multrec 1332 9.7 14.183 17.275 22.029 25.048 make_m2s 222 7.7 0.007 0.008 21.088 22.533 make_images 222 8.7 3.129 3.592 21.038 22.484 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.050 17.369 make_images_data 222 9.7 0.004 0.004 11.797 13.428 hybrid_alltoall_any 227 10.6 0.797 3.778 11.150 13.020 dbcsr_mm_accdrv_process 3641 10.4 0.238 0.483 7.485 8.978 dbcsr_mm_accdrv_process_sort 3641 11.4 7.088 8.562 7.088 8.562 mp_sum_l 807 5.4 4.223 7.681 4.223 7.681 multiply_cannon_sync_h2d 1332 9.7 5.506 6.056 5.506 6.056 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.082 6.008 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.234 5.991 mp_irecv_dv 3229 10.9 2.060 5.963 2.060 5.963 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.526 5.935 arnoldi_extremal 4 6.8 0.000 0.000 5.203 5.219 arnoldi_normal_ev 4 7.8 0.001 0.004 5.203 5.219 build_subspace 16 8.4 0.014 0.021 4.865 4.872 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.533 4.713 mp_allgather_i34 111 8.7 2.196 4.640 2.196 4.640 calculate_norms 2376 9.8 4.184 4.577 4.184 4.577 dbcsr_matrix_vector_mult 304 9.0 0.010 0.020 3.593 3.890 dbcsr_matrix_vector_mult_local 304 10.0 3.206 3.714 3.208 3.716 dbcsr_sort_data 658 11.4 3.091 3.445 3.091 3.445 ls_scf_post 1 4.0 0.000 0.001 3.245 3.250 dbcsr_special_finalize 555 9.7 0.006 0.007 2.828 3.220 dbcsr_merge_single_wm 555 10.7 0.537 0.657 2.820 3.211 ls_scf_dm_to_ks 2 5.0 0.000 0.004 3.044 3.104 ls_scf_store_result 1 5.0 0.000 0.000 2.995 3.047 dbcsr_data_release 10477 10.7 1.580 2.417 1.580 2.417 dbcsr_finalize 304 7.8 0.049 0.061 1.808 1.989 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.983 1.985 rebuild_ks_matrix 3 7.3 0.000 0.000 1.960 1.962 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.960 1.962 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=96.868000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2723.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b3f583d97791bedccf073d04f637ff05a010c441_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 4.630995E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265558. MP_Allreduce 3049 15663. MP_Sync 4 MP_Alltoall 47 46208988. MP_SendRecv 45 115200. MP_ISendRecv 45 115200. MP_Wait 528 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.040 0.057 99.592 99.594 qs_energies 1 2.0 0.000 0.000 98.825 98.829 ls_scf 1 3.0 0.000 0.000 96.860 96.867 dbcsr_multiply_generic 111 6.7 0.017 0.019 78.167 78.449 ls_scf_main 1 4.0 0.000 0.000 62.476 62.477 multiply_cannon 111 7.7 0.106 0.180 55.215 60.109 density_matrix_trs4 2 5.0 0.002 0.003 55.163 55.268 multiply_cannon_loop 111 8.7 0.069 0.077 50.748 53.410 mp_waitall_1 5436 11.0 26.583 32.301 26.583 32.301 ls_scf_init_scf 1 4.0 0.000 0.001 30.790 30.795 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.414 29.466 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.253 27.262 multiply_cannon_multrec 444 9.7 13.962 16.196 20.957 24.232 make_m2s 222 7.7 0.004 0.005 17.584 20.033 make_images 222 8.7 3.710 4.403 17.522 19.973 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 5.961 17.044 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.466 16.274 make_images_data 222 9.7 0.003 0.004 9.958 12.221 hybrid_alltoall_any 227 10.6 0.790 3.764 9.678 11.986 multiply_cannon_sync_h2d 444 9.7 6.473 8.144 6.473 8.144 dbcsr_mm_accdrv_process 3003 10.4 0.187 0.343 6.706 7.851 dbcsr_mm_accdrv_process_sort 3003 11.4 6.392 7.508 6.392 7.508 mp_allgather_i34 111 8.7 2.700 7.012 2.700 7.012 mp_sum_l 807 5.4 3.416 6.963 3.416 6.963 arnoldi_extremal 4 6.8 0.000 0.000 5.791 5.804 arnoldi_normal_ev 4 7.8 0.002 0.005 5.791 5.804 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.470 5.578 build_subspace 16 8.4 0.015 0.020 5.395 5.409 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.585 4.778 dbcsr_matrix_vector_mult 304 9.0 0.010 0.020 4.136 4.352 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.689 4.215 mp_irecv_dv 1241 11.2 1.671 4.192 1.671 4.192 dbcsr_matrix_vector_mult_local 304 10.0 3.672 4.148 3.674 4.150 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.702 3.809 calculate_norms 792 9.8 3.527 3.702 3.527 3.702 ls_scf_post 1 4.0 0.000 0.000 3.594 3.598 ls_scf_store_result 1 5.0 0.000 0.000 3.369 3.414 make_images_sizes 222 9.7 0.000 0.000 1.004 3.312 mp_alltoall_i44 222 10.7 1.003 3.311 1.003 3.311 dbcsr_finalize 304 7.8 0.062 0.078 2.198 2.256 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.197 2.199 rebuild_ks_matrix 3 7.3 0.000 0.000 2.165 2.166 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.165 2.166 dbcsr_merge_all 275 8.9 0.473 0.521 2.047 2.094 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=99.594000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3659.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b3f583d97791bedccf073d04f637ff05a010c441_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 8.845238E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284111. MP_Allreduce 3043 21950. MP_Sync 4 MP_Alltoall 47 88727262. MP_SendRecv 42 732600. MP_ISendRecv 42 732600. MP_Wait 267 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.119 0.175 108.230 108.232 qs_energies 1 2.0 0.000 0.000 106.750 106.762 ls_scf 1 3.0 0.000 0.000 103.748 103.760 dbcsr_multiply_generic 111 6.7 0.023 0.028 77.175 77.309 ls_scf_main 1 4.0 0.000 0.000 65.934 65.935 density_matrix_trs4 2 5.0 0.002 0.003 56.590 56.668 multiply_cannon 111 7.7 0.146 0.181 49.484 50.854 multiply_cannon_loop 111 8.7 0.067 0.070 46.137 47.488 ls_scf_init_scf 1 4.0 0.002 0.003 34.094 34.095 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.445 32.453 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.623 29.632 mp_waitall_1 4527 11.1 21.746 25.913 21.746 25.913 make_m2s 222 7.7 0.005 0.005 23.708 24.578 make_images 222 8.7 4.585 4.991 23.602 24.471 multiply_cannon_multrec 444 9.7 17.926 18.528 22.499 22.945 hybrid_alltoall_any 227 10.6 1.659 3.620 12.810 15.397 make_images_data 222 9.7 0.003 0.003 12.973 15.243 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.220 11.105 multiply_cannon_sync_h2d 444 9.7 8.846 8.884 8.846 8.884 arnoldi_extremal 4 6.8 0.000 0.000 7.332 7.348 arnoldi_normal_ev 4 7.8 0.003 0.009 7.332 7.348 build_subspace 16 8.4 0.026 0.036 6.790 6.801 dbcsr_matrix_vector_mult 304 9.0 0.016 0.032 5.436 5.580 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.309 5.402 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.027 5.303 dbcsr_matrix_vector_mult_local 304 10.0 4.994 5.296 4.997 5.299 dbcsr_mm_accdrv_process 1814 10.4 0.206 0.304 4.409 4.535 dbcsr_mm_accdrv_process_sort 1814 11.4 4.112 4.238 4.112 4.238 ls_scf_post 1 4.0 0.000 0.000 3.721 3.733 make_images_sizes 222 9.7 0.000 0.000 1.408 3.472 mp_alltoall_i44 222 10.7 1.408 3.471 1.408 3.471 ls_scf_store_result 1 5.0 0.000 0.000 3.421 3.445 calculate_norms 792 9.8 3.242 3.272 3.242 3.272 dbcsr_finalize 304 7.8 0.082 0.089 3.086 3.140 mp_allgather_i34 111 8.7 0.995 3.116 0.995 3.116 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.971 2.971 dbcsr_merge_all 275 8.9 0.890 0.915 2.872 2.918 dbcsr_complete_redistribute 5 7.6 1.438 1.492 2.765 2.885 dbcsr_data_release 12724 10.6 2.335 2.875 2.335 2.875 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.619 2.621 rebuild_ks_matrix 3 7.3 0.000 0.000 2.553 2.554 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.553 2.554 mp_sum_l 807 5.4 1.672 2.540 1.672 2.540 matrix_ls_to_qs 2 6.0 0.000 0.000 2.404 2.532 dbcsr_sort_data 325 11.1 2.442 2.492 2.442 2.492 dbcsr_new_transposed 4 7.5 0.243 0.252 2.270 2.288 dbcsr_frobenius_norm 74 6.6 2.056 2.128 2.182 2.225 dbcsr_add_d 103 6.2 0.000 0.000 2.131 2.203 dbcsr_add_anytype 103 7.2 0.859 0.892 2.131 2.203 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=108.232000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6835.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: b3f583d97791bedccf073d04f637ff05a010c441 Summary: empty Status: OK