=== 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: 48320402b71dba81277f3b122ba0d11842227d22 ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, Cray-FFTW 3.3.8.10, # COSMA 2.6.2, ELPA 2022.11.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.1.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.1, # SIRIUS 7.3.2, 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 (12.01.2023) # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.1.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.1 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.3.2 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(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/48320402b71dba81277f3b122ba0d11842227d22_performance_tests/01 job id: 44391514 --- 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/48320402b71dba81277f3b122ba0d11842227d22_performance_tests/02 job id: 44391517 --- 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/48320402b71dba81277f3b122ba0d11842227d22_performance_tests/03 job id: 44391519 --- 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/48320402b71dba81277f3b122ba0d11842227d22_performance_tests/04 job id: 44391520 --- 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/48320402b71dba81277f3b122ba0d11842227d22_performance_tests/05 job id: 44391522 --- 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/48320402b71dba81277f3b122ba0d11842227d22_performance_tests/06 job id: 44391523 --- 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/48320402b71dba81277f3b122ba0d11842227d22_performance_tests/07 job id: 44391524 --- 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/48320402b71dba81277f3b122ba0d11842227d22_performance_tests/08 job id: 44391526 --- 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/48320402b71dba81277f3b122ba0d11842227d22_performance_tests/09 job id: 44391527 --- 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/48320402b71dba81277f3b122ba0d11842227d22_performance_tests/10 job id: 44391529 --- 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/48320402b71dba81277f3b122ba0d11842227d22_performance_tests/11 job id: 44391530 --- 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/48320402b71dba81277f3b122ba0d11842227d22_performance_tests/12 job id: 44391531 --- 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/48320402b71dba81277f3b122ba0d11842227d22_performance_tests/13 job id: 44391533 --- 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/48320402b71dba81277f3b122ba0d11842227d22_performance_tests/14 job id: 44391534 --- 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/48320402b71dba81277f3b122ba0d11842227d22_performance_tests/15 job id: 44391535 --- 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/48320402b71dba81277f3b122ba0d11842227d22_performance_tests/16 job id: 44391538 --- 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/48320402b71dba81277f3b122ba0d11842227d22_performance_tests/17 job id: 44391539 --- 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/48320402b71dba81277f3b122ba0d11842227d22_performance_tests/18 job id: 44391541 --- 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/48320402b71dba81277f3b122ba0d11842227d22_performance_tests/19 job id: 44391542 --- 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/48320402b71dba81277f3b122ba0d11842227d22_performance_tests/20 job id: 44391543 --- 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/48320402b71dba81277f3b122ba0d11842227d22_performance_tests/21 job id: 44391544 --- 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/48320402b71dba81277f3b122ba0d11842227d22_performance_tests/22 job id: 44391546 --- 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/48320402b71dba81277f3b122ba0d11842227d22_performance_tests/23 job id: 44391547 --- 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/48320402b71dba81277f3b122ba0d11842227d22_performance_tests/24 job id: 44391549 --- 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/48320402b71dba81277f3b122ba0d11842227d22_performance_tests/25 job id: 44391551 --- 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/48320402b71dba81277f3b122ba0d11842227d22_performance_tests/26 job id: 44391552 --- 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/48320402b71dba81277f3b122ba0d11842227d22_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 0.029 0.045 132.712 132.713 farming_run 1 2.0 132.138 132.141 132.676 132.685 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.474384E+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 0.010 0.026 114.503 114.503 qs_energies 1 2.0 0.000 0.000 114.222 114.224 mp2_main 1 3.0 0.000 0.000 112.353 112.355 mp2_gpw_main 1 4.0 0.020 0.026 111.512 111.514 mp2_ri_gpw_compute_in 1 5.0 0.178 0.181 93.053 93.147 mp2_ri_gpw_compute_in_loop 1 6.0 0.005 0.005 55.318 55.410 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.169 41.625 46.576 get_2c_integrals 1 6.0 0.000 0.000 36.943 37.556 integrate_v_rspace 273 8.0 0.439 0.450 25.012 29.598 pw_transfer 6555 10.6 0.374 0.386 27.319 27.604 fft_wrap_pw1pw2 5465 11.4 0.045 0.047 25.974 26.187 grid_integrate_task_list 273 9.0 20.834 25.863 20.834 25.863 fft_wrap_pw1pw2_100 2178 12.4 1.184 1.277 23.491 23.741 compute_2c_integrals 1 7.0 0.035 0.036 19.316 19.316 compute_2c_integrals_loop_lm 1 8.0 0.008 0.009 18.880 19.015 mp2_eri_2c_integrate_gpw 1 9.0 2.401 2.443 18.871 19.007 rpa_ri_compute_en 1 5.0 0.001 0.013 18.351 18.435 cp_fm_cholesky_decompose 12 8.2 17.579 18.208 17.579 18.208 cholesky_decomp 1 7.0 0.000 0.000 16.477 17.117 fft3d_s 5443 13.4 16.082 16.340 16.105 16.363 ao_to_mo_and_store_B_mult_1 272 7.0 10.853 15.565 10.853 15.565 calculate_wavefunction 272 8.0 5.433 5.517 12.530 13.224 rpa_num_int 1 6.0 0.000 0.006 10.550 10.560 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.502 10.537 calc_mat_Q 8 8.0 0.000 0.000 9.349 9.453 contract_S_to_Q 8 9.0 0.000 0.000 8.773 8.872 calc_potential_gpw 544 9.5 0.005 0.006 8.199 8.593 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.190 8.438 parallel_gemm_fm 14 9.1 0.000 0.000 8.361 8.431 parallel_gemm_fm_cosma 14 10.1 8.361 8.431 8.361 8.431 potential_pw2rs 545 10.0 0.107 0.109 7.710 8.377 collocate_single_gaussian 272 10.0 0.040 0.043 7.457 7.717 create_integ_mat 1 6.0 0.019 0.032 7.603 7.613 array2fm 1 7.0 0.000 0.000 6.645 7.081 pw_scatter_s 2720 13.7 4.427 4.650 4.427 4.650 pw_gather_s 2722 13.2 3.869 4.194 3.869 4.194 array2fm_buffer_send 1 8.0 2.952 3.148 2.952 3.148 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.511561, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2745.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/48320402b71dba81277f3b122ba0d11842227d22_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 0.051 0.060 397.451 397.452 farming_run 1 2.0 396.618 396.666 397.390 397.394 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827141120 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788822 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.222746E+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 2369671. 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 0.179 0.588 210.934 210.970 qs_energies 1 2.0 0.001 0.007 209.943 210.082 scf_env_do_scf 1 3.0 0.000 0.000 107.069 107.069 qs_ks_update_qs_env 5 5.0 0.000 0.000 106.198 106.207 rebuild_ks_matrix 4 6.0 0.000 0.000 106.095 106.102 qs_ks_build_kohn_sham_matrix 4 7.0 0.056 0.063 106.095 106.102 hfx_ks_matrix 4 8.0 0.001 0.001 105.712 105.715 integrate_four_center 4 9.0 0.144 0.461 105.711 105.715 mp2_main 1 3.0 0.005 0.029 102.580 102.676 mp2_gpw_main 1 4.0 0.037 0.090 101.733 101.800 integrate_four_center_main 4 10.0 0.096 0.520 97.081 99.553 integrate_four_center_bin 263 11.0 96.985 99.548 96.985 99.548 init_scf_loop 1 4.0 0.000 0.000 92.337 92.337 mp2_ri_gpw_compute_in 1 5.0 0.065 0.074 74.971 76.057 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.523 55.605 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.157 42.191 47.046 integrate_v_rspace 95 8.0 0.398 0.566 28.559 33.347 pw_transfer 2240 10.6 0.145 0.165 29.918 30.366 fft_wrap_pw1pw2 1868 11.4 0.018 0.021 28.929 29.370 ao_to_mo_and_store_B_mult_1 91 7.0 10.639 29.113 10.639 29.113 grid_integrate_task_list 95 9.0 23.846 28.859 23.846 28.859 mp2_ri_gpw_compute_en 1 5.0 0.060 0.106 26.573 28.220 fft_wrap_pw1pw2_100 730 12.4 1.269 1.464 26.630 27.083 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.848 1.894 24.798 24.808 get_2c_integrals 1 6.0 0.005 0.046 20.319 20.388 compute_2c_integrals 1 7.0 0.004 0.011 19.284 19.297 compute_2c_integrals_loop_lm 1 8.0 0.002 0.009 18.833 19.157 mp2_eri_2c_integrate_gpw 1 9.0 1.732 1.876 18.831 19.155 fft3d_s 1823 13.4 18.433 18.876 18.447 18.890 scf_env_do_scf_inner_loop 4 4.0 0.002 0.104 14.730 14.731 calculate_wavefunction 91 8.0 2.020 2.052 9.750 9.971 potential_pw2rs 186 10.0 0.033 0.035 8.630 9.204 mp2_ri_gpw_compute_en_expansio 172 7.0 0.557 0.578 8.723 9.179 local_gemm 172 8.0 8.166 8.633 8.166 8.633 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.232 8.611 mp2_ri_gpw_compute_en_comm 22 7.0 0.515 0.626 7.830 8.384 calc_potential_gpw 182 9.5 0.002 0.002 7.910 8.167 collocate_single_gaussian 91 10.0 0.017 0.028 7.867 8.096 mp2_ri_gpw_compute_en_ener 172 7.0 6.352 6.439 6.352 6.439 mp_sendrecv_dm3 2068 8.0 5.823 6.370 5.823 6.370 mp_sync 38 10.4 3.336 6.093 3.336 6.093 pw_gather_s 912 13.2 4.915 5.346 4.915 5.346 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.799593, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1507.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/48320402b71dba81277f3b122ba0d11842227d22_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 451.444736E+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 589622. 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 0.028 0.054 54.194 54.209 qs_mol_dyn_low 1 2.0 0.003 0.004 53.817 53.824 qs_forces 11 3.9 0.012 0.079 53.712 53.713 qs_energies 11 4.9 0.002 0.005 52.127 52.140 scf_env_do_scf 11 5.9 0.001 0.028 45.953 45.953 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 43.836 43.837 dbcsr_multiply_generic 2286 12.5 0.095 0.097 34.959 35.449 qs_scf_new_mos 108 7.5 0.000 0.001 33.534 33.849 qs_scf_loop_do_ot 108 8.5 0.000 0.001 33.534 33.848 ot_scf_mini 108 9.5 0.002 0.002 31.879 32.091 multiply_cannon 2286 13.5 0.178 0.185 26.180 28.552 multiply_cannon_loop 2286 14.5 1.494 1.558 25.480 27.874 velocity_verlet 10 3.0 0.006 0.014 25.951 25.951 ot_mini 108 10.5 0.001 0.001 20.823 21.061 qs_ot_get_derivative 108 11.5 0.001 0.001 17.868 18.069 mp_waitall_1 245248 16.5 8.544 15.127 8.544 15.127 multiply_cannon_metrocomm3 54864 15.5 0.067 0.074 5.964 13.450 multiply_cannon_multrec 54864 15.5 4.209 6.519 7.685 11.153 mp_sum_l 7207 12.9 6.770 8.675 6.770 8.675 rebuild_ks_matrix 119 8.3 0.000 0.000 8.034 8.167 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.013 8.034 8.167 multiply_cannon_sync_h2d 54864 15.5 5.890 7.947 5.890 7.947 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.909 7.274 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.101 7.225 qs_ot_get_p 119 10.4 0.001 0.001 6.491 6.796 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.658 5.801 init_scf_run 11 5.9 0.000 0.001 4.920 4.920 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.920 4.920 dbcsr_mm_accdrv_process 76910 16.1 1.173 1.816 3.399 4.754 sum_up_and_integrate 119 10.3 0.012 0.014 4.589 4.596 integrate_v_rspace 119 11.3 0.002 0.002 4.576 4.585 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.233 4.349 calculate_rho_elec 119 8.7 0.011 0.017 4.232 4.348 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.851 4.188 apply_single 119 13.6 0.000 0.000 3.851 4.188 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.645 3.728 multiply_cannon_metrocomm1 54864 15.5 0.052 0.057 1.873 3.680 calculate_dm_sparse 119 9.5 0.000 0.001 3.127 3.294 acc_transpose_blocks 54864 15.5 0.230 0.249 1.875 3.242 rs_pw_transfer 974 11.9 0.011 0.012 2.759 2.883 jit_kernel_multiply 13 15.8 2.164 2.878 2.164 2.878 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.858 2.859 calculate_first_density_matrix 1 7.0 0.001 0.004 2.780 2.784 ot_diis_step 108 11.5 0.006 0.006 2.677 2.677 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.657 2.658 cp_fm_redistribute_end 50 14.0 2.419 2.636 2.425 2.638 cp_fm_diag_elpa_base 50 14.0 0.211 2.544 0.212 2.557 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.378 2.442 density_rs2pw 119 9.7 0.004 0.004 2.256 2.400 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.249 2.250 acc_transpose_blocks_kernels 54864 16.5 0.249 0.378 0.894 2.144 grid_integrate_task_list 119 12.3 2.023 2.129 2.023 2.129 wfi_extrapolate 11 7.9 0.001 0.001 2.078 2.078 init_scf_loop 11 6.9 0.001 0.003 2.068 2.068 mp_sum_d 4125 12.0 1.449 2.060 1.449 2.060 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.976 2.030 pw_transfer 1439 11.6 0.053 0.059 1.858 1.934 potential_pw2rs 119 12.3 0.004 0.004 1.919 1.931 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.781 1.862 jit_kernel_transpose 5 15.5 0.645 1.783 0.645 1.783 make_m2s 4572 13.5 0.054 0.056 1.717 1.758 make_images 4572 14.5 0.133 0.138 1.634 1.676 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.634 1.655 fft3d_ps 1201 14.6 0.372 0.476 1.552 1.631 fft_wrap_pw1pw2_140 487 13.2 0.082 0.095 1.372 1.455 mp_alltoall_d11v 2130 13.8 1.282 1.433 1.282 1.433 mp_waitany 12084 13.8 1.249 1.400 1.249 1.400 grid_collocate_task_list 119 9.7 1.296 1.368 1.296 1.368 dbcsr_dot_sd 1205 11.9 0.051 0.062 0.778 1.212 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=54.209000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=430.727273, yerr=0.445362 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/48320402b71dba81277f3b122ba0d11842227d22_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 489.242624E+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 1943370. 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 0.030 0.061 40.755 40.765 qs_mol_dyn_low 1 2.0 0.003 0.003 40.384 40.391 qs_forces 11 3.9 0.002 0.002 40.312 40.313 qs_energies 11 4.9 0.002 0.008 38.603 38.606 scf_env_do_scf 11 5.9 0.001 0.002 32.791 32.792 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 30.177 30.177 dbcsr_multiply_generic 2286 12.5 0.126 0.129 22.279 22.647 qs_scf_new_mos 108 7.5 0.001 0.001 20.584 20.830 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.583 20.830 ot_scf_mini 108 9.5 0.002 0.003 19.683 19.861 velocity_verlet 10 3.0 0.005 0.013 18.869 18.870 multiply_cannon 2286 13.5 0.209 0.228 17.046 18.700 multiply_cannon_loop 2286 14.5 0.901 0.981 15.882 17.435 ot_mini 108 10.5 0.001 0.001 12.174 12.412 mp_waitall_1 200699 16.5 6.067 11.563 6.067 11.563 multiply_cannon_metrocomm3 27432 15.5 0.068 0.071 4.417 9.928 qs_ot_get_derivative 108 11.5 0.001 0.001 9.670 9.854 multiply_cannon_multrec 27432 15.5 1.967 4.550 6.174 9.030 rebuild_ks_matrix 119 8.3 0.000 0.000 7.419 7.556 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.015 7.418 7.555 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.563 6.689 dbcsr_mm_accdrv_process 47894 16.0 3.139 5.586 4.139 6.183 qs_ot_get_p 119 10.4 0.001 0.001 4.592 4.822 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.763 4.626 init_scf_run 11 5.9 0.000 0.001 4.554 4.554 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.553 4.554 sum_up_and_integrate 119 10.3 0.024 0.026 4.380 4.385 integrate_v_rspace 119 11.3 0.002 0.002 4.355 4.363 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.089 4.152 apply_single 119 13.6 0.000 0.000 3.089 4.152 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.034 4.113 calculate_rho_elec 119 8.7 0.021 0.024 4.033 4.113 mp_sum_l 7207 12.9 2.150 4.086 2.150 4.086 rs_pw_transfer 974 11.9 0.010 0.011 2.832 3.253 qs_ot_p2m_diag 50 11.0 0.008 0.012 3.064 3.083 calculate_first_density_matrix 1 7.0 0.000 0.003 2.948 2.953 make_m2s 4572 13.5 0.052 0.054 2.615 2.859 density_rs2pw 119 9.7 0.004 0.004 2.349 2.789 multiply_cannon_sync_h2d 27432 15.5 2.192 2.769 2.192 2.769 make_images 4572 14.5 0.199 0.235 2.526 2.767 cp_dbcsr_syevd 50 12.0 0.003 0.004 2.622 2.623 init_scf_loop 11 6.9 0.001 0.006 2.585 2.585 calculate_dm_sparse 119 9.5 0.000 0.000 2.415 2.494 ot_diis_step 108 11.5 0.011 0.011 2.441 2.442 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.230 2.321 jit_kernel_multiply 10 16.1 0.949 2.308 0.949 2.308 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.236 2.237 cp_fm_redistribute_end 50 14.0 1.848 2.208 1.852 2.209 pw_transfer 1439 11.6 0.065 0.073 2.120 2.155 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.150 2.152 cp_fm_diag_elpa_base 50 14.0 0.342 2.087 0.355 2.150 potential_pw2rs 119 12.3 0.006 0.006 2.098 2.110 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.028 2.067 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.891 1.929 grid_integrate_task_list 119 12.3 1.830 1.917 1.830 1.917 fft3d_ps 1201 14.6 0.512 0.563 1.728 1.763 prepare_preconditioner 11 7.9 0.000 0.000 1.656 1.685 make_preconditioner 11 8.9 0.000 0.002 1.656 1.685 acc_transpose_blocks 27432 15.5 0.108 0.113 1.283 1.638 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.628 1.637 make_images_data 4572 15.5 0.045 0.051 1.156 1.598 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.538 1.597 fft_wrap_pw1pw2_140 487 13.2 0.079 0.086 1.513 1.557 wfi_extrapolate 11 7.9 0.001 0.001 1.545 1.545 hybrid_alltoall_any 4725 16.4 0.051 0.111 1.016 1.530 mp_alltoall_d11v 2130 13.8 1.311 1.434 1.311 1.434 grid_collocate_task_list 119 9.7 1.228 1.357 1.228 1.357 mp_allgather_i34 2286 14.5 0.599 1.332 0.599 1.332 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.183 1.228 mp_sum_d 4125 12.0 0.637 1.073 0.637 1.073 acc_transpose_blocks_kernels 27432 16.5 0.181 0.272 0.756 1.026 mp_waitany 5720 13.7 0.526 0.989 0.526 0.989 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 0.972 0.973 rs_pw_transfer_RS2PW_140 130 11.5 0.139 0.148 0.542 0.963 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.917 0.930 make_images_sizes 4572 15.5 0.005 0.005 0.649 0.909 mp_alltoall_i44 4572 16.5 0.644 0.904 0.644 0.904 mp_alltoall_z22v 1201 16.6 0.769 0.874 0.769 0.874 rs_pw_transfer_PW2RS_50 119 14.3 0.589 0.606 0.789 0.830 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=40.765000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=464.727273, yerr=1.420045 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/48320402b71dba81277f3b122ba0d11842227d22_performance_tests/05/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 59.051995E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3143552 0.0% 0.0% 100.0% average stack size 0.0 0.0 46.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 523.694080E+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 1607811. 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 0.025 0.052 32.923 32.924 qs_mol_dyn_low 1 2.0 0.003 0.003 32.624 32.631 qs_forces 11 3.9 0.009 0.059 32.524 32.524 qs_energies 11 4.9 0.002 0.005 30.879 30.881 scf_env_do_scf 11 5.9 0.001 0.001 25.866 25.866 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 23.254 23.255 dbcsr_multiply_generic 2286 12.5 0.096 0.099 16.366 16.478 velocity_verlet 10 3.0 0.005 0.007 15.522 15.525 qs_scf_new_mos 108 7.5 0.001 0.001 14.879 14.891 qs_scf_loop_do_ot 108 8.5 0.001 0.001 14.878 14.891 ot_scf_mini 108 9.5 0.002 0.004 14.158 14.166 multiply_cannon 2286 13.5 0.195 0.202 13.021 13.843 multiply_cannon_loop 2286 14.5 0.637 0.666 12.217 13.052 ot_mini 108 10.5 0.001 0.001 8.768 8.778 qs_ot_get_derivative 108 11.5 0.001 0.001 7.260 7.270 multiply_cannon_multrec 18288 15.5 1.924 2.927 6.746 7.050 rebuild_ks_matrix 119 8.3 0.000 0.000 6.498 6.512 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 6.497 6.512 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.749 5.762 dbcsr_mm_accdrv_process 38222 16.0 3.946 5.348 4.739 5.539 sum_up_and_integrate 119 10.3 0.029 0.030 4.056 4.069 integrate_v_rspace 119 11.3 0.002 0.003 4.027 4.039 mp_waitall_1 158411 16.6 2.635 3.922 2.635 3.922 init_scf_run 11 5.9 0.000 0.001 3.565 3.565 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.565 3.565 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.539 3.550 calculate_rho_elec 119 8.7 0.030 0.031 3.539 3.550 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.715 3.345 qs_ot_get_p 119 10.4 0.001 0.001 3.212 3.238 rs_pw_transfer 974 11.9 0.009 0.010 2.423 2.667 init_scf_loop 11 6.9 0.001 0.003 2.592 2.593 multiply_cannon_metrocomm3 18288 15.5 0.045 0.047 1.436 2.513 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.058 2.371 apply_single 119 13.6 0.000 0.000 2.057 2.371 density_rs2pw 119 9.7 0.004 0.004 2.062 2.311 calculate_first_density_matrix 1 7.0 0.000 0.001 2.309 2.310 qs_ot_p2m_diag 50 11.0 0.012 0.012 2.144 2.151 make_m2s 4572 13.5 0.044 0.045 1.814 1.986 pw_transfer 1439 11.6 0.066 0.071 1.961 1.971 make_images 4572 14.5 0.190 0.201 1.729 1.899 cp_dbcsr_syevd 50 12.0 0.004 0.004 1.879 1.880 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.867 1.880 grid_integrate_task_list 119 12.3 1.800 1.877 1.800 1.877 jit_kernel_multiply 10 16.1 0.742 1.872 0.742 1.872 potential_pw2rs 119 12.3 0.007 0.008 1.832 1.839 calculate_dm_sparse 119 9.5 0.000 0.000 1.814 1.823 prepare_preconditioner 11 7.9 0.000 0.001 1.786 1.787 make_preconditioner 11 8.9 0.000 0.001 1.785 1.787 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.720 1.722 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.636 1.721 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.639 1.641 cp_fm_redistribute_end 50 14.0 1.220 1.615 1.221 1.615 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.593 1.603 cp_fm_diag_elpa_base 50 14.0 0.378 1.530 0.392 1.576 multiply_cannon_sync_h2d 18288 15.5 1.405 1.568 1.405 1.568 fft3d_ps 1201 14.6 0.523 0.542 1.546 1.560 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.541 1.547 ot_diis_step 108 11.5 0.011 0.011 1.488 1.488 mp_sum_l 7207 12.9 1.153 1.475 1.153 1.475 fft_wrap_pw1pw2_140 487 13.2 0.090 0.096 1.458 1.469 grid_collocate_task_list 119 9.7 1.216 1.358 1.216 1.358 acc_transpose_blocks 18288 15.5 0.075 0.077 1.244 1.277 wfi_extrapolate 11 7.9 0.001 0.001 1.208 1.208 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.186 1.191 make_images_data 4572 15.5 0.045 0.049 0.799 0.988 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 0.978 0.979 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.884 0.907 hybrid_alltoall_any 4725 16.4 0.055 0.112 0.676 0.887 acc_transpose_blocks_kernels 18288 16.5 0.209 0.219 0.808 0.836 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.821 0.824 mp_waitany 9880 13.7 0.551 0.810 0.551 0.810 rs_pw_transfer_RS2PW_140 130 11.5 0.120 0.123 0.526 0.769 mp_alltoall_d11v 2130 13.8 0.665 0.753 0.665 0.753 cp_fm_cholesky_invert 11 10.9 0.746 0.750 0.746 0.750 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.656 0.744 mp_alltoall_z22v 1201 16.6 0.620 0.716 0.620 0.716 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=32.924000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.909091, yerr=2.678349 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/48320402b71dba81277f3b122ba0d11842227d22_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 546.598912E+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 0.030 0.073 36.420 36.421 qs_mol_dyn_low 1 2.0 0.003 0.004 36.089 36.097 qs_forces 11 3.9 0.002 0.004 36.024 36.027 qs_energies 11 4.9 0.001 0.004 34.272 34.280 scf_env_do_scf 11 5.9 0.001 0.001 28.970 28.971 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 25.398 25.399 dbcsr_multiply_generic 2286 12.5 0.102 0.103 18.344 18.474 velocity_verlet 10 3.0 0.007 0.012 18.451 18.453 qs_scf_new_mos 108 7.5 0.001 0.001 16.588 16.638 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.588 16.637 ot_scf_mini 108 9.5 0.002 0.003 15.642 15.693 multiply_cannon 2286 13.5 0.241 0.296 14.472 14.896 multiply_cannon_loop 2286 14.5 0.937 0.973 13.441 13.795 ot_mini 108 10.5 0.001 0.001 9.502 9.568 multiply_cannon_multrec 27432 15.5 2.323 2.977 8.569 8.890 qs_ot_get_derivative 108 11.5 0.001 0.001 7.665 7.716 dbcsr_mm_accdrv_process 47916 15.9 5.287 6.952 6.155 7.349 rebuild_ks_matrix 119 8.3 0.000 0.000 6.795 6.853 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.015 6.794 6.853 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.044 6.096 sum_up_and_integrate 119 10.3 0.035 0.038 3.972 3.981 integrate_v_rspace 119 11.3 0.002 0.002 3.937 3.946 init_scf_run 11 5.9 0.000 0.001 3.895 3.895 scf_env_initial_rho_setup 11 6.9 0.014 0.041 3.894 3.895 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.769 3.799 calculate_rho_elec 119 8.7 0.040 0.046 3.768 3.798 qs_ot_get_p 119 10.4 0.001 0.001 3.506 3.581 init_scf_loop 11 6.9 0.001 0.010 3.548 3.550 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.763 3.235 prepare_preconditioner 11 7.9 0.000 0.001 2.629 2.638 make_preconditioner 11 8.9 0.000 0.001 2.629 2.638 mp_waitall_1 137007 16.6 1.957 2.608 1.957 2.608 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.127 2.607 apply_single 119 13.6 0.000 0.000 2.127 2.606 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.226 2.553 make_m2s 4572 13.5 0.054 0.056 2.322 2.470 rs_pw_transfer 974 11.9 0.009 0.009 2.241 2.441 calculate_first_density_matrix 1 7.0 0.000 0.001 2.417 2.431 density_rs2pw 119 9.7 0.004 0.004 2.168 2.363 make_images 4572 14.5 0.269 0.331 2.214 2.361 pw_transfer 1439 11.6 0.066 0.072 2.205 2.245 qs_ot_p2m_diag 50 11.0 0.015 0.022 2.169 2.177 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.112 2.156 calculate_dm_sparse 119 9.5 0.000 0.000 2.070 2.140 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.975 2.002 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.919 1.920 grid_integrate_task_list 119 12.3 1.826 1.883 1.826 1.883 cp_dbcsr_syevd 50 12.0 0.003 0.004 1.845 1.845 fft3d_ps 1201 14.6 0.560 0.613 1.787 1.823 ot_diis_step 108 11.5 0.012 0.012 1.795 1.795 fft_wrap_pw1pw2_140 487 13.2 0.088 0.095 1.704 1.753 potential_pw2rs 119 12.3 0.008 0.009 1.725 1.732 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.643 1.659 mp_sum_l 7207 12.9 1.135 1.619 1.135 1.619 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.572 1.573 cp_fm_redistribute_end 50 14.0 1.040 1.543 1.041 1.544 cp_fm_diag_elpa_base 50 14.0 0.481 1.476 0.501 1.514 jit_kernel_multiply 8 15.7 0.809 1.499 0.809 1.499 acc_transpose_blocks 27432 15.5 0.112 0.114 1.440 1.461 wfi_extrapolate 11 7.9 0.001 0.001 1.413 1.413 grid_collocate_task_list 119 9.7 1.222 1.340 1.222 1.340 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.761 1.305 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.245 1.258 dbcsr_complete_redistribute 329 12.2 0.162 0.209 0.895 1.171 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.127 1.144 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.114 1.116 cp_fm_upper_to_full 72 13.5 0.799 1.110 0.799 1.110 multiply_cannon_sync_h2d 27432 15.5 0.991 1.104 0.991 1.104 make_images_data 4572 15.5 0.045 0.049 0.905 1.071 hybrid_alltoall_any 4725 16.4 0.062 0.151 0.787 0.994 mp_alltoall_d11v 2130 13.8 0.810 0.938 0.810 0.938 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.664 0.931 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.795 0.869 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.843 0.849 mp_alltoall_z22v 1201 16.6 0.810 0.846 0.810 0.846 acc_transpose_blocks_kernels 27432 16.5 0.267 0.276 0.821 0.831 cp_fm_cholesky_invert 11 10.9 0.794 0.797 0.794 0.797 mp_alltoall_i22 627 13.8 0.442 0.744 0.442 0.744 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=36.421000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=518.090909, yerr=3.342687 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/48320402b71dba81277f3b122ba0d11842227d22_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 607.137792E+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 0.029 0.062 28.493 28.503 qs_mol_dyn_low 1 2.0 0.003 0.004 28.158 28.166 qs_forces 11 3.9 0.002 0.002 28.079 28.080 qs_energies 11 4.9 0.001 0.001 26.370 26.372 scf_env_do_scf 11 5.9 0.001 0.007 21.444 21.444 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 18.960 18.960 velocity_verlet 10 3.0 0.011 0.012 14.360 14.363 dbcsr_multiply_generic 2286 12.5 0.093 0.096 12.338 12.436 qs_scf_new_mos 108 7.5 0.001 0.001 10.912 10.936 qs_scf_loop_do_ot 108 8.5 0.001 0.001 10.912 10.936 ot_scf_mini 108 9.5 0.002 0.002 10.239 10.266 multiply_cannon 2286 13.5 0.232 0.241 9.883 10.144 multiply_cannon_loop 2286 14.5 0.330 0.342 8.957 9.162 multiply_cannon_multrec 9144 15.5 1.636 2.062 5.946 6.336 rebuild_ks_matrix 119 8.3 0.000 0.000 6.009 6.030 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 6.008 6.030 ot_mini 108 10.5 0.001 0.001 5.818 5.848 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.355 5.374 qs_ot_get_derivative 108 11.5 0.001 0.001 4.521 4.549 dbcsr_mm_accdrv_process 12550 15.8 3.154 4.100 4.211 4.310 sum_up_and_integrate 119 10.3 0.038 0.041 3.674 3.681 integrate_v_rspace 119 11.3 0.002 0.003 3.636 3.643 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.568 3.575 calculate_rho_elec 119 8.7 0.060 0.061 3.568 3.575 init_scf_run 11 5.9 0.000 0.001 3.493 3.493 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.493 3.493 qs_ot_get_p 119 10.4 0.001 0.001 2.666 2.705 init_scf_loop 11 6.9 0.000 0.000 2.458 2.458 calculate_first_density_matrix 1 7.0 0.000 0.000 2.363 2.364 pw_transfer 1439 11.6 0.066 0.069 2.167 2.177 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.073 2.085 density_rs2pw 119 9.7 0.004 0.004 1.896 2.035 mp_waitall_1 115863 16.7 1.455 1.999 1.455 1.999 grid_integrate_task_list 119 12.3 1.850 1.954 1.850 1.954 make_m2s 4572 13.5 0.034 0.035 1.700 1.892 rs_pw_transfer 974 11.9 0.008 0.009 1.651 1.805 calculate_dm_sparse 119 9.5 0.000 0.000 1.785 1.805 make_images 4572 14.5 0.267 0.301 1.611 1.801 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.779 1.781 fft3d_ps 1201 14.6 0.562 0.572 1.733 1.740 fft_wrap_pw1pw2_140 487 13.2 0.087 0.090 1.726 1.737 prepare_preconditioner 11 7.9 0.000 0.000 1.716 1.720 make_preconditioner 11 8.9 0.000 0.000 1.716 1.720 jit_kernel_multiply 9 15.8 1.019 1.715 1.019 1.715 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.639 1.640 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.609 1.636 cp_dbcsr_syevd 50 12.0 0.004 0.004 1.588 1.588 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.551 1.566 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.486 1.493 potential_pw2rs 119 12.3 0.010 0.010 1.407 1.409 grid_collocate_task_list 119 9.7 1.272 1.393 1.272 1.393 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.306 1.306 cp_fm_redistribute_end 50 14.0 0.652 1.287 0.653 1.287 ot_diis_step 108 11.5 0.012 0.013 1.285 1.285 cp_fm_diag_elpa_base 50 14.0 0.592 1.218 0.634 1.274 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.228 1.229 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.204 1.208 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.159 1.181 apply_single 119 13.6 0.000 0.000 1.159 1.181 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.116 1.129 wfi_extrapolate 11 7.9 0.001 0.001 1.081 1.081 acc_transpose_blocks 9144 15.5 0.038 0.039 1.005 1.058 hybrid_alltoall_any 4725 16.4 0.063 0.175 0.777 1.057 make_images_data 4572 15.5 0.039 0.042 0.798 1.020 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.866 0.919 mp_alltoall_d11v 2130 13.8 0.762 0.905 0.762 0.905 cp_fm_cholesky_invert 11 10.9 0.849 0.852 0.849 0.852 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.375 0.839 acc_transpose_blocks_kernels 9144 16.5 0.117 0.120 0.766 0.817 multiply_cannon_sync_h2d 9144 15.5 0.714 0.782 0.714 0.782 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.774 0.777 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.681 0.733 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.708 0.715 jit_kernel_transpose 5 15.6 0.649 0.701 0.649 0.701 mp_allgather_i34 2286 14.5 0.259 0.672 0.259 0.672 mp_alltoall_z22v 1201 16.6 0.618 0.669 0.618 0.669 yz_to_x 606 15.1 0.265 0.276 0.591 0.606 multiply_cannon_metrocomm1 9144 15.5 0.022 0.023 0.306 0.597 qs_create_task_list 11 7.9 0.000 0.001 0.551 0.577 mp_waitany 5200 13.7 0.438 0.576 0.438 0.576 generate_qs_task_list 11 8.9 0.192 0.216 0.551 0.576 x_to_yz 595 16.2 0.275 0.287 0.568 0.576 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=28.503000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=573.545455, yerr=8.574419 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/48320402b71dba81277f3b122ba0d11842227d22_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 747.372544E+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 0.033 0.089 42.881 42.908 qs_mol_dyn_low 1 2.0 0.003 0.004 42.483 42.490 qs_forces 11 3.9 0.002 0.002 42.352 42.352 qs_energies 11 4.9 0.001 0.002 40.349 40.352 scf_env_do_scf 11 5.9 0.003 0.019 33.577 33.577 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 25.767 25.768 velocity_verlet 10 3.0 0.011 0.012 23.545 23.559 dbcsr_multiply_generic 2286 12.5 0.101 0.102 18.581 18.805 qs_scf_new_mos 108 7.5 0.001 0.001 15.828 15.913 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.827 15.912 multiply_cannon 2286 13.5 0.314 0.337 14.769 15.687 ot_scf_mini 108 9.5 0.002 0.002 14.750 14.842 multiply_cannon_loop 2286 14.5 0.345 0.350 13.481 14.454 multiply_cannon_multrec 9144 15.5 3.461 4.764 9.670 9.790 ot_mini 108 10.5 0.001 0.001 8.865 8.976 init_scf_loop 11 6.9 0.000 0.000 7.763 7.765 rebuild_ks_matrix 119 8.3 0.000 0.000 7.400 7.532 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.399 7.531 dbcsr_mm_accdrv_process 12550 15.8 4.953 6.598 6.088 7.452 qs_ot_get_derivative 108 11.5 0.001 0.001 6.852 6.944 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.690 6.807 prepare_preconditioner 11 7.9 0.000 0.000 6.791 6.803 make_preconditioner 11 8.9 0.000 0.000 6.791 6.803 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.364 6.683 init_scf_run 11 5.9 0.000 0.001 4.720 4.720 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.720 4.720 cp_fm_upper_to_full 72 14.2 3.136 4.504 3.136 4.504 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.328 4.332 calculate_rho_elec 119 8.7 0.118 0.121 4.328 4.331 sum_up_and_integrate 119 10.3 0.065 0.067 4.057 4.063 integrate_v_rspace 119 11.3 0.003 0.003 3.992 4.000 mp_waitall_1 94719 16.7 2.302 3.329 2.302 3.329 qs_ot_get_p 119 10.4 0.001 0.001 3.140 3.268 calculate_first_density_matrix 1 7.0 0.000 0.000 3.250 3.252 calculate_dm_sparse 119 9.5 0.000 0.000 3.185 3.204 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.500 2.910 pw_transfer 1439 11.6 0.069 0.070 2.876 2.880 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.777 2.781 dbcsr_complete_redistribute 329 12.2 0.285 0.290 1.968 2.767 jit_kernel_multiply 8 15.5 1.107 2.466 1.107 2.466 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.173 2.464 apply_single 119 13.6 0.000 0.000 2.173 2.464 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.654 2.455 make_m2s 4572 13.5 0.038 0.038 2.273 2.447 fft3d_ps 1201 14.6 0.597 0.603 2.401 2.405 fft_wrap_pw1pw2_140 487 13.2 0.095 0.096 2.344 2.352 make_images 4572 14.5 0.351 0.381 2.152 2.325 density_rs2pw 119 9.7 0.004 0.004 2.250 2.269 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 2.254 2.261 mp_alltoall_i22 627 13.8 1.419 2.248 1.419 2.248 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.422 2.217 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 1.270 2.166 grid_integrate_task_list 119 12.3 2.058 2.081 2.058 2.081 ot_diis_step 108 11.5 0.014 0.014 1.987 1.988 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.929 1.930 qs_ot_p2m_diag 50 11.0 0.042 0.043 1.901 1.902 qs_energies_init_hamiltonians 11 5.9 0.002 0.002 1.785 1.786 mp_sum_l 7207 12.9 1.010 1.741 1.010 1.741 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.689 1.735 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.667 1.695 cp_dbcsr_syevd 50 12.0 0.004 0.004 1.626 1.626 grid_collocate_task_list 119 9.7 1.490 1.507 1.490 1.507 rs_pw_transfer 974 11.9 0.009 0.010 1.459 1.497 cp_fm_cholesky_invert 11 10.9 1.478 1.481 1.478 1.481 potential_pw2rs 119 12.3 0.014 0.015 1.451 1.454 wfi_extrapolate 11 7.9 0.001 0.001 1.396 1.396 hybrid_alltoall_any 4725 16.4 0.087 0.148 1.117 1.373 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.339 1.339 cp_fm_diag_elpa_base 50 14.0 1.197 1.248 1.337 1.337 make_images_data 4572 15.5 0.043 0.046 1.072 1.287 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.142 1.162 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.114 1.134 mp_alltoall_d11v 2130 13.8 1.079 1.094 1.079 1.094 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 0.985 1.038 acc_transpose_blocks 9144 15.5 0.038 0.038 1.000 1.020 yz_to_x 606 15.1 0.459 0.471 0.963 0.973 qs_create_task_list 11 7.9 0.001 0.001 0.946 0.959 generate_qs_task_list 11 8.9 0.376 0.394 0.945 0.958 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.922 0.936 mp_alltoall_z22v 1201 16.6 0.853 0.895 0.853 0.895 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=42.908000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=700.909091, yerr=16.228228 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/48320402b71dba81277f3b122ba0d11842227d22_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 502.177792E+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 0.024 0.072 82.977 82.979 qs_mol_dyn_low 1 2.0 0.003 0.004 82.507 82.602 qs_forces 11 3.9 0.004 0.012 82.180 82.182 qs_energies 11 4.9 0.002 0.004 79.307 79.335 scf_env_do_scf 11 5.9 0.000 0.001 70.494 70.496 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 64.960 64.960 dbcsr_multiply_generic 2055 12.4 0.108 0.113 51.501 51.833 qs_scf_new_mos 99 7.5 0.000 0.001 47.583 47.689 qs_scf_loop_do_ot 99 8.5 0.001 0.001 47.582 47.689 ot_scf_mini 99 9.5 0.002 0.002 45.191 45.280 multiply_cannon 2055 13.4 0.173 0.179 42.160 43.362 multiply_cannon_loop 2055 14.4 1.540 1.572 41.228 42.474 velocity_verlet 10 3.0 0.003 0.010 42.127 42.130 ot_mini 99 10.5 0.001 0.001 27.254 27.345 qs_ot_get_derivative 99 11.5 0.001 0.001 20.487 20.583 multiply_cannon_multrec 49320 15.4 12.125 12.943 17.137 18.152 rebuild_ks_matrix 110 8.3 0.000 0.001 14.549 14.671 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.549 14.671 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.766 12.874 mp_waitall_1 220248 16.4 11.156 12.178 11.156 12.178 multiply_cannon_sync_h2d 49320 15.4 9.979 10.592 9.979 10.592 qs_ot_get_p 110 10.4 0.001 0.001 9.516 9.645 multiply_cannon_metrocomm3 49320 15.4 0.079 0.084 6.807 8.218 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.163 7.640 apply_single 110 13.6 0.000 0.001 7.162 7.639 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.029 7.563 sum_up_and_integrate 110 10.3 0.036 0.043 7.063 7.075 integrate_v_rspace 110 11.3 0.003 0.003 7.026 7.047 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.606 6.757 calculate_rho_elec 110 8.6 0.021 0.026 6.605 6.756 init_scf_run 11 5.9 0.000 0.001 6.744 6.744 scf_env_initial_rho_setup 11 6.9 0.010 0.019 6.744 6.744 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.513 6.577 ot_diis_step 99 11.5 0.005 0.006 6.535 6.535 qs_ot_p2m_diag 48 11.0 0.012 0.018 6.361 6.389 mp_sum_l 6514 12.8 5.151 6.085 5.151 6.085 init_scf_loop 11 6.9 0.000 0.001 5.507 5.507 cp_dbcsr_syevd 48 12.0 0.003 0.003 5.382 5.382 dbcsr_mm_accdrv_process 87628 16.1 2.078 2.179 4.891 5.171 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.925 4.927 cp_fm_redistribute_end 48 14.0 4.300 4.902 4.304 4.903 cp_fm_diag_elpa_base 48 14.0 0.593 4.757 0.597 4.785 rs_pw_transfer 902 11.9 0.011 0.013 3.620 4.184 wfi_extrapolate 11 7.9 0.001 0.001 4.007 4.007 density_rs2pw 110 9.6 0.004 0.005 3.389 3.885 make_m2s 4110 13.4 0.060 0.065 3.758 3.879 calculate_dm_sparse 110 9.5 0.000 0.001 3.752 3.868 make_images 4110 14.4 0.177 0.190 3.663 3.787 multiply_cannon_metrocomm1 49320 15.4 0.060 0.063 2.545 3.709 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.505 3.509 grid_integrate_task_list 110 12.3 3.248 3.402 3.248 3.402 prepare_preconditioner 11 7.9 0.000 0.000 3.340 3.357 make_preconditioner 11 8.9 0.000 0.000 3.340 3.357 pw_transfer 1331 11.6 0.055 0.069 3.214 3.276 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.199 3.239 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.181 3.227 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 3.125 3.190 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.122 3.168 fft_wrap_pw1pw2_140 451 13.1 0.171 0.192 2.671 2.739 fft3d_ps 1111 14.6 0.784 0.883 2.652 2.709 calculate_first_density_matrix 1 7.0 0.000 0.002 2.640 2.654 potential_pw2rs 110 12.3 0.006 0.007 2.599 2.618 mp_alltoall_d11v 2046 13.8 2.082 2.552 2.082 2.552 jit_kernel_multiply 13 15.9 2.530 2.540 2.530 2.540 mp_waitany 14300 13.8 1.811 2.346 1.811 2.346 grid_collocate_task_list 110 9.6 2.085 2.289 2.085 2.289 acc_transpose_blocks 49320 15.4 0.223 0.232 2.155 2.228 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.923 1.943 mp_sum_d 3879 11.9 1.447 1.940 1.447 1.940 make_images_data 4110 15.4 0.043 0.046 1.750 1.877 cp_fm_cholesky_invert 11 10.9 1.789 1.793 1.789 1.793 hybrid_alltoall_any 4261 16.3 0.082 0.482 1.534 1.785 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.636 1.666 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=82.979000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.818182, yerr=2.248967 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/48320402b71dba81277f3b122ba0d11842227d22_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 584.818688E+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 2462973. 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 0.142 0.215 71.242 71.249 qs_mol_dyn_low 1 2.0 0.005 0.020 70.736 70.744 qs_forces 11 3.9 0.014 0.032 69.642 69.643 qs_energies 11 4.9 0.002 0.006 66.274 66.290 scf_env_do_scf 11 5.9 0.001 0.001 56.384 56.388 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 48.838 48.838 dbcsr_multiply_generic 2055 12.4 0.115 0.121 38.676 38.814 velocity_verlet 10 3.0 0.007 0.012 37.064 37.086 multiply_cannon 2055 13.4 0.222 0.248 32.176 33.246 qs_scf_new_mos 99 7.5 0.001 0.001 32.371 32.505 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.371 32.504 multiply_cannon_loop 2055 14.4 0.928 0.959 30.918 31.796 ot_scf_mini 99 9.5 0.003 0.004 30.725 30.851 ot_mini 99 10.5 0.001 0.001 18.060 18.189 multiply_cannon_multrec 24660 15.4 7.690 9.247 13.987 15.914 rebuild_ks_matrix 110 8.3 0.000 0.001 13.869 13.968 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.015 13.869 13.968 qs_ot_get_derivative 99 11.5 0.001 0.001 12.268 12.398 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.210 12.295 mp_waitall_1 176588 16.5 7.890 10.603 7.890 10.603 multiply_cannon_metrocomm3 24660 15.4 0.070 0.072 5.409 8.290 multiply_cannon_sync_h2d 24660 15.4 7.107 8.263 7.107 8.263 init_scf_loop 11 6.9 0.001 0.004 7.512 7.513 init_scf_run 11 5.9 0.000 0.001 7.481 7.481 scf_env_initial_rho_setup 11 6.9 0.001 0.002 7.480 7.481 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.515 7.275 apply_single 110 13.6 0.000 0.001 6.515 7.275 dbcsr_mm_accdrv_process 52282 16.1 4.706 5.810 6.138 6.805 sum_up_and_integrate 110 10.3 0.051 0.057 6.653 6.663 integrate_v_rspace 110 11.3 0.002 0.003 6.601 6.611 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.245 6.257 calculate_rho_elec 110 8.6 0.039 0.047 6.244 6.256 qs_ot_get_p 110 10.4 0.001 0.001 6.035 6.183 ot_diis_step 99 11.5 0.010 0.010 5.746 5.747 prepare_preconditioner 11 7.9 0.000 0.000 5.477 5.504 make_preconditioner 11 8.9 0.000 0.002 5.477 5.504 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.662 5.407 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.056 5.218 make_m2s 4110 13.4 0.057 0.060 4.126 4.576 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 4.469 4.470 make_images 4110 14.4 0.396 0.440 4.017 4.464 qs_ot_p2m_diag 48 11.0 0.028 0.043 4.196 4.216 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.050 4.094 pw_transfer 1331 11.6 0.067 0.074 3.750 3.891 calculate_first_density_matrix 1 7.0 0.000 0.003 3.883 3.885 density_rs2pw 110 9.6 0.004 0.004 3.343 3.852 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.643 3.786 cp_dbcsr_syevd 48 12.0 0.007 0.009 3.777 3.778 rs_pw_transfer 902 11.9 0.012 0.014 3.003 3.579 wfi_extrapolate 11 7.9 0.001 0.001 3.510 3.510 grid_integrate_task_list 110 12.3 3.153 3.325 3.153 3.325 fft_wrap_pw1pw2_140 451 13.1 0.203 0.221 3.131 3.280 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.255 3.260 cp_fm_redistribute_end 48 14.0 2.434 3.228 2.436 3.230 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.137 3.202 cp_fm_diag_elpa_base 48 14.0 0.760 3.089 0.791 3.176 fft3d_ps 1111 14.6 1.103 1.325 2.991 3.127 calculate_dm_sparse 110 9.5 0.001 0.001 3.003 3.031 make_images_data 4110 15.4 0.047 0.051 2.319 2.785 hybrid_alltoall_any 4261 16.3 0.102 0.442 2.015 2.765 acc_transpose_blocks 24660 15.4 0.109 0.112 2.258 2.650 cp_fm_cholesky_invert 11 10.9 2.516 2.523 2.516 2.523 grid_collocate_task_list 110 9.6 2.048 2.483 2.048 2.483 potential_pw2rs 110 12.3 0.008 0.009 2.452 2.466 mp_sum_l 6514 12.8 1.780 2.437 1.780 2.437 jit_kernel_multiply 11 16.4 1.084 2.175 1.084 2.175 acc_transpose_blocks_kernels 24660 16.4 0.308 0.318 1.647 2.036 mp_alltoall_d11v 2046 13.8 1.736 2.004 1.736 2.004 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.926 1.947 qs_energies_init_hamiltonians 11 5.9 0.003 0.013 1.889 1.902 mp_waitany 10164 13.8 1.267 1.892 1.267 1.892 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.823 1.833 jit_kernel_transpose 5 15.6 1.339 1.721 1.339 1.721 multiply_cannon_metrocomm4 22605 15.4 0.072 0.077 0.745 1.633 mp_allgather_i34 2055 14.4 0.585 1.597 0.585 1.597 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.587 1.597 rs_pw_transfer_RS2PW_140 121 11.5 0.207 0.219 0.996 1.557 mp_irecv_dv 57340 16.2 0.622 1.515 0.622 1.515 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.371 1.476 dbcsr_complete_redistribute 325 12.2 0.294 0.386 1.201 1.470 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=71.249000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=553.545455, yerr=5.852173 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/48320402b71dba81277f3b122ba0d11842227d22_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 661.159936E+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 0.111 0.179 61.569 61.584 qs_mol_dyn_low 1 2.0 0.003 0.004 61.033 61.043 qs_forces 11 3.9 0.005 0.016 60.965 60.966 qs_energies 11 4.9 0.008 0.031 57.701 57.705 scf_env_do_scf 11 5.9 0.001 0.001 48.331 48.331 scf_env_do_scf_inner_loop 99 6.5 0.002 0.008 39.800 39.801 velocity_verlet 10 3.0 0.004 0.011 32.531 32.533 dbcsr_multiply_generic 2055 12.4 0.109 0.112 29.281 29.533 qs_scf_new_mos 99 7.5 0.001 0.001 24.622 24.716 qs_scf_loop_do_ot 99 8.5 0.001 0.001 24.622 24.715 multiply_cannon 2055 13.4 0.210 0.222 23.040 24.382 ot_scf_mini 99 9.5 0.003 0.003 23.411 23.518 multiply_cannon_loop 2055 14.4 0.617 0.635 21.902 23.031 ot_mini 99 10.5 0.001 0.001 13.370 13.474 rebuild_ks_matrix 110 8.3 0.000 0.000 12.475 12.599 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.475 12.598 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.993 11.107 multiply_cannon_multrec 16440 15.4 3.903 4.741 9.801 11.043 mp_waitall_1 139946 16.5 7.038 9.970 7.038 9.970 qs_ot_get_derivative 99 11.5 0.001 0.001 8.963 9.071 init_scf_loop 11 6.9 0.001 0.003 8.496 8.498 multiply_cannon_metrocomm3 16440 15.4 0.044 0.045 4.356 7.081 init_scf_run 11 5.9 0.000 0.001 6.835 6.836 scf_env_initial_rho_setup 11 6.9 0.003 0.018 6.835 6.836 prepare_preconditioner 11 7.9 0.000 0.000 6.706 6.722 make_preconditioner 11 8.9 0.000 0.002 6.706 6.722 sum_up_and_integrate 110 10.3 0.060 0.061 6.573 6.587 integrate_v_rspace 110 11.3 0.002 0.003 6.513 6.527 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.022 6.380 dbcsr_mm_accdrv_process 34862 16.1 4.724 5.377 5.752 6.344 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.041 6.055 calculate_rho_elec 110 8.6 0.058 0.059 6.041 6.054 qs_ot_get_p 110 10.4 0.001 0.001 5.245 5.394 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.934 5.365 apply_single 110 13.6 0.000 0.000 4.934 5.365 make_m2s 4110 13.4 0.049 0.051 4.179 4.526 make_images 4110 14.4 0.391 0.512 4.064 4.411 ot_diis_step 99 11.5 0.010 0.011 4.379 4.379 density_rs2pw 110 9.6 0.004 0.004 3.143 4.371 multiply_cannon_sync_h2d 16440 15.4 3.681 4.263 3.681 4.263 rs_pw_transfer 902 11.9 0.010 0.011 2.702 3.878 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.840 3.842 calculate_first_density_matrix 1 7.0 0.001 0.003 3.780 3.783 pw_transfer 1331 11.6 0.066 0.073 3.728 3.735 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.049 3.697 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.684 3.688 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.620 3.630 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.431 3.478 grid_integrate_task_list 110 12.3 3.165 3.443 3.165 3.443 cp_dbcsr_syevd 48 12.0 0.003 0.004 3.351 3.352 fft_wrap_pw1pw2_140 451 13.1 0.213 0.217 3.134 3.146 wfi_extrapolate 11 7.9 0.001 0.001 2.968 2.968 fft3d_ps 1111 14.6 1.087 1.100 2.920 2.930 make_images_data 4110 15.4 0.043 0.048 2.475 2.919 hybrid_alltoall_any 4261 16.3 0.105 0.377 2.187 2.874 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.852 2.854 cp_fm_redistribute_end 48 14.0 1.779 2.827 1.782 2.827 cp_fm_diag_elpa_base 48 14.0 0.984 2.687 1.042 2.790 calculate_dm_sparse 110 9.5 0.001 0.001 2.581 2.608 cp_fm_cholesky_invert 11 10.9 2.589 2.595 2.589 2.595 mp_waitany 17072 13.8 1.227 2.476 1.227 2.476 grid_collocate_task_list 110 9.6 2.082 2.457 2.082 2.457 multiply_cannon_metrocomm4 14385 15.4 0.045 0.048 0.892 2.380 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.289 2.353 potential_pw2rs 110 12.3 0.011 0.011 2.306 2.313 mp_irecv_dv 48980 15.7 0.821 2.248 0.821 2.248 acc_transpose_blocks 16440 15.4 0.070 0.072 2.069 2.227 mp_alltoall_d11v 2046 13.8 1.754 2.224 1.754 2.224 rs_pw_transfer_RS2PW_140 121 11.5 0.177 0.193 0.937 2.107 mp_sum_l 6514 12.8 1.484 2.038 1.484 2.038 qs_energies_init_hamiltonians 11 5.9 0.003 0.012 2.020 2.022 dbcsr_complete_redistribute 325 12.2 0.356 0.437 1.538 2.000 cp_fm_upper_to_full 70 13.6 1.388 1.848 1.388 1.848 acc_transpose_blocks_kernels 16440 16.4 0.208 0.215 1.650 1.803 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.681 1.695 jit_kernel_transpose 5 15.6 1.442 1.595 1.442 1.595 cp_fm_cholesky_decompose 22 10.9 1.570 1.590 1.570 1.590 mp_allgather_i34 2055 14.4 0.466 1.586 0.466 1.586 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.550 1.565 jit_kernel_multiply 8 16.4 0.642 1.525 0.642 1.525 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.368 1.524 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.066 1.518 rs_gather_matrices 110 12.3 0.234 0.259 0.966 1.406 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.228 1.236 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=61.584000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=625.818182, yerr=9.437047 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/48320402b71dba81277f3b122ba0d11842227d22_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 729.518080E+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 0.063 0.138 65.289 65.291 qs_mol_dyn_low 1 2.0 0.003 0.004 64.922 64.930 qs_forces 11 3.9 0.005 0.013 64.585 64.587 qs_energies 11 4.9 0.002 0.005 61.159 61.164 scf_env_do_scf 11 5.9 0.001 0.001 52.625 52.628 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 41.117 41.118 velocity_verlet 10 3.0 0.005 0.006 37.040 37.053 dbcsr_multiply_generic 2055 12.4 0.116 0.118 29.604 29.807 qs_scf_new_mos 99 7.5 0.001 0.001 25.939 26.053 qs_scf_loop_do_ot 99 8.5 0.002 0.008 25.939 26.053 ot_scf_mini 99 9.5 0.003 0.005 24.334 24.438 multiply_cannon 2055 13.4 0.242 0.259 22.736 24.103 multiply_cannon_loop 2055 14.4 0.887 0.905 21.332 21.933 ot_mini 99 10.5 0.001 0.001 14.058 14.182 multiply_cannon_multrec 24660 15.4 4.204 6.660 12.840 13.923 rebuild_ks_matrix 110 8.3 0.000 0.000 12.321 12.433 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.321 12.433 init_scf_loop 11 6.9 0.001 0.002 11.469 11.469 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.883 10.980 qs_ot_get_derivative 99 11.5 0.001 0.001 9.918 10.022 prepare_preconditioner 11 7.9 0.000 0.000 9.693 9.711 make_preconditioner 11 8.9 0.000 0.001 9.693 9.711 dbcsr_mm_accdrv_process 52304 16.0 6.863 8.229 8.490 9.420 make_full_inverse_cholesky 11 9.9 0.000 0.000 7.933 9.377 sum_up_and_integrate 110 10.3 0.067 0.070 6.539 6.555 integrate_v_rspace 110 11.3 0.003 0.003 6.472 6.491 mp_waitall_1 121746 16.5 4.401 6.314 4.401 6.314 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.162 6.170 calculate_rho_elec 110 8.6 0.077 0.081 6.162 6.169 init_scf_run 11 5.9 0.000 0.001 5.752 5.752 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.751 5.752 make_m2s 4110 13.4 0.059 0.063 5.308 5.688 make_images 4110 14.4 0.572 0.692 5.168 5.543 qs_ot_get_p 110 10.4 0.001 0.001 5.273 5.442 cp_fm_upper_to_full 70 13.8 3.250 4.621 3.250 4.621 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.019 4.111 apply_single 110 13.6 0.000 0.000 4.018 4.111 ot_diis_step 99 11.5 0.011 0.011 4.105 4.105 pw_transfer 1331 11.6 0.065 0.075 3.910 3.939 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.804 3.837 dbcsr_complete_redistribute 325 12.2 0.438 0.492 2.658 3.784 density_rs2pw 110 9.6 0.004 0.004 3.079 3.586 qs_ot_p2m_diag 48 11.0 0.054 0.063 3.558 3.571 grid_integrate_task_list 110 12.3 3.248 3.450 3.248 3.450 multiply_cannon_metrocomm3 24660 15.4 0.036 0.037 1.539 3.409 multiply_cannon_sync_h2d 24660 15.4 3.167 3.359 3.167 3.359 fft_wrap_pw1pw2_140 451 13.1 0.203 0.211 3.286 3.323 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.235 3.291 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.126 3.233 fft3d_ps 1111 14.6 1.091 1.127 3.096 3.126 hybrid_alltoall_any 4261 16.3 0.120 0.456 2.265 3.105 cp_dbcsr_syevd 48 12.0 0.006 0.010 3.088 3.088 make_images_data 4110 15.4 0.047 0.051 2.720 3.081 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.949 3.002 calculate_dm_sparse 110 9.5 0.001 0.001 2.960 2.989 wfi_extrapolate 11 7.9 0.001 0.001 2.969 2.969 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.964 2.967 rs_pw_transfer 902 11.9 0.010 0.011 2.353 2.841 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.752 2.839 mp_alltoall_i22 605 13.7 1.628 2.796 1.628 2.796 calculate_first_density_matrix 1 7.0 0.001 0.002 2.691 2.694 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.581 2.626 cp_fm_cholesky_invert 11 10.9 2.597 2.605 2.597 2.605 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.570 2.574 cp_fm_redistribute_end 48 14.0 1.285 2.544 1.286 2.545 cp_fm_diag_elpa_base 48 14.0 1.179 2.413 1.256 2.525 grid_collocate_task_list 110 9.6 2.182 2.433 2.182 2.433 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 2.285 2.287 jit_kernel_multiply 11 15.7 1.300 2.222 1.300 2.222 potential_pw2rs 110 12.3 0.012 0.013 2.195 2.205 mp_alltoall_d11v 2046 13.8 1.745 2.054 1.745 2.054 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.673 1.705 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.683 1.696 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.599 1.694 cp_fm_cholesky_decompose 22 10.9 1.629 1.688 1.629 1.688 mp_allgather_i34 2055 14.4 0.506 1.634 0.506 1.634 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.002 1.592 1.609 acc_transpose_blocks 24660 15.4 0.105 0.107 1.502 1.530 mp_waitany 13376 13.8 1.054 1.522 1.054 1.522 mp_sum_l 6514 12.8 0.909 1.502 0.909 1.502 multiply_cannon_metrocomm4 20550 15.4 0.057 0.060 0.870 1.474 mp_irecv_dv 62702 16.1 0.769 1.391 0.769 1.391 qs_env_update_s_mstruct 11 6.9 0.000 0.001 1.234 1.330 rs_pw_transfer_RS2PW_140 121 11.5 0.169 0.175 0.822 1.317 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=65.291000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=691.272727, yerr=9.583043 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/48320402b71dba81277f3b122ba0d11842227d22_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 842.969088E+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 0.037 0.069 55.887 55.894 qs_mol_dyn_low 1 2.0 0.003 0.004 55.595 55.604 qs_forces 11 3.9 0.002 0.002 55.491 55.491 qs_energies 11 4.9 0.009 0.009 51.783 51.788 scf_env_do_scf 11 5.9 0.000 0.001 43.013 43.013 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 35.354 35.354 velocity_verlet 10 3.0 0.007 0.008 31.124 31.129 dbcsr_multiply_generic 2055 12.4 0.105 0.107 23.202 23.331 qs_scf_new_mos 99 7.5 0.001 0.001 20.118 20.158 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.117 20.158 multiply_cannon 2055 13.4 0.244 0.257 17.846 19.227 ot_scf_mini 99 9.5 0.002 0.002 18.880 18.900 multiply_cannon_loop 2055 14.4 0.322 0.334 16.491 16.791 rebuild_ks_matrix 110 8.3 0.000 0.000 12.055 12.078 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 12.055 12.078 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.708 10.728 ot_mini 99 10.5 0.001 0.001 10.265 10.281 multiply_cannon_multrec 8220 15.4 3.270 4.625 7.905 8.917 init_scf_loop 11 6.9 0.000 0.000 7.613 7.614 mp_waitall_1 103326 16.6 5.864 7.433 5.864 7.433 sum_up_and_integrate 110 10.3 0.079 0.081 6.585 6.598 integrate_v_rspace 110 11.3 0.003 0.003 6.506 6.519 qs_ot_get_derivative 99 11.5 0.001 0.001 6.494 6.513 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.366 6.380 calculate_rho_elec 110 8.6 0.115 0.116 6.365 6.380 prepare_preconditioner 11 7.9 0.000 0.000 5.944 5.948 make_preconditioner 11 8.9 0.000 0.000 5.944 5.948 init_scf_run 11 5.9 0.000 0.001 5.623 5.624 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.623 5.623 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.528 5.599 dbcsr_mm_accdrv_process 17442 15.9 2.989 4.284 4.506 5.548 qs_ot_get_p 110 10.4 0.001 0.001 4.725 4.742 make_m2s 4110 13.4 0.038 0.039 4.157 4.426 make_images 4110 14.4 0.636 0.689 4.027 4.294 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 2.933 4.276 pw_transfer 1331 11.6 0.066 0.072 4.159 4.175 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.051 4.071 ot_diis_step 99 11.5 0.012 0.012 3.751 3.752 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.671 3.697 apply_single 110 13.6 0.000 0.000 3.671 3.697 fft_wrap_pw1pw2_140 451 13.1 0.216 0.219 3.513 3.535 grid_integrate_task_list 110 12.3 3.358 3.515 3.358 3.515 density_rs2pw 110 9.6 0.004 0.004 3.117 3.490 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.431 3.434 fft3d_ps 1111 14.6 1.152 1.189 3.293 3.314 cp_dbcsr_syevd 48 12.0 0.004 0.004 3.132 3.132 multiply_cannon_sync_h2d 8220 15.4 2.920 3.009 2.920 3.009 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.884 2.886 cp_fm_cholesky_invert 11 10.9 2.840 2.844 2.840 2.844 calculate_first_density_matrix 1 7.0 0.000 0.000 2.808 2.810 hybrid_alltoall_any 4261 16.3 0.200 0.867 2.315 2.765 make_images_data 4110 15.4 0.038 0.043 2.377 2.762 wfi_extrapolate 11 7.9 0.001 0.001 2.718 2.718 qs_energies_init_hamiltonians 11 5.9 0.002 0.002 2.691 2.692 calculate_dm_sparse 110 9.5 0.001 0.001 2.641 2.685 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.628 2.628 cp_fm_redistribute_end 48 14.0 0.668 2.602 0.671 2.602 rs_pw_transfer 902 11.9 0.010 0.010 2.181 2.575 cp_fm_diag_elpa_base 48 14.0 1.747 2.415 1.923 2.573 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.548 2.562 grid_collocate_task_list 110 9.6 2.278 2.539 2.278 2.539 jit_kernel_multiply 10 15.6 1.210 2.166 1.210 2.166 potential_pw2rs 110 12.3 0.015 0.016 2.141 2.146 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.771 1.997 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.939 1.948 mp_alltoall_d11v 2046 13.8 1.594 1.800 1.594 1.800 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.755 1.767 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.682 1.686 cp_fm_cholesky_decompose 22 10.9 1.660 1.674 1.660 1.674 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.526 1.650 mp_allgather_i34 2055 14.4 0.537 1.626 0.537 1.626 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.604 1.613 mp_waitany 9240 13.8 1.077 1.507 1.077 1.507 dbcsr_complete_redistribute 325 12.2 0.558 0.585 1.413 1.504 qs_create_task_list 11 7.9 0.001 0.001 1.227 1.330 generate_qs_task_list 11 8.9 0.383 0.448 1.226 1.330 rs_pw_transfer_RS2PW_140 121 11.5 0.162 0.165 0.801 1.207 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.124 1.138 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=55.894000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=794.727273, yerr=14.384909 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/48320402b71dba81277f3b122ba0d11842227d22_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 1.317978E+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 0.039 0.080 88.467 88.475 qs_mol_dyn_low 1 2.0 0.003 0.004 88.121 88.129 qs_forces 11 3.9 0.002 0.002 87.953 87.954 qs_energies 11 4.9 0.002 0.002 83.714 83.716 scf_env_do_scf 11 5.9 0.001 0.001 73.652 73.653 velocity_verlet 10 3.0 0.002 0.002 56.599 56.617 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 44.997 44.997 dbcsr_multiply_generic 2055 12.4 0.120 0.122 28.520 28.608 init_scf_loop 11 6.9 0.000 0.000 28.583 28.584 prepare_preconditioner 11 7.9 0.000 0.000 26.537 26.546 make_preconditioner 11 8.9 0.000 0.000 26.537 26.546 qs_scf_new_mos 99 7.5 0.001 0.001 26.329 26.395 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.328 26.395 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.700 26.011 ot_scf_mini 99 9.5 0.002 0.002 24.553 24.611 multiply_cannon 2055 13.4 0.349 0.371 21.531 22.225 multiply_cannon_loop 2055 14.4 0.342 0.344 19.745 20.062 cp_fm_upper_to_full 70 14.2 12.697 18.252 12.697 18.252 rebuild_ks_matrix 110 8.3 0.001 0.001 14.387 14.427 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 14.387 14.426 ot_mini 99 10.5 0.001 0.001 13.580 13.633 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.022 13.058 dbcsr_complete_redistribute 325 12.2 1.017 1.047 7.435 10.743 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.389 9.699 multiply_cannon_multrec 8220 15.4 4.358 4.554 9.572 9.678 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.822 9.110 qs_ot_get_derivative 99 11.5 0.001 0.001 9.018 9.065 mp_alltoall_i22 605 13.7 5.452 8.745 5.452 8.745 mp_waitall_1 84994 16.7 7.551 8.365 7.551 8.365 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.932 7.966 calculate_rho_elec 110 8.6 0.227 0.227 7.932 7.966 sum_up_and_integrate 110 10.3 0.149 0.151 7.513 7.527 integrate_v_rspace 110 11.3 0.004 0.004 7.363 7.378 qs_ot_get_p 110 10.4 0.001 0.001 5.765 5.827 init_scf_run 11 5.9 0.000 0.001 5.702 5.702 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.702 5.702 make_m2s 4110 13.4 0.043 0.043 5.207 5.701 make_images 4110 14.4 0.880 0.936 5.019 5.513 pw_transfer 1331 11.6 0.075 0.075 5.433 5.444 cp_fm_cholesky_invert 11 10.9 5.405 5.409 5.405 5.409 dbcsr_mm_accdrv_process 11614 15.7 3.228 3.639 5.071 5.336 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 5.316 5.327 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 4.830 5.164 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.672 5.144 apply_single 110 13.6 0.000 0.000 4.672 5.144 fft_wrap_pw1pw2_140 451 13.1 0.221 0.224 4.613 4.628 ot_diis_step 99 11.5 0.015 0.015 4.535 4.536 fft3d_ps 1111 14.6 1.295 1.306 4.512 4.526 qs_ot_p2m_diag 48 11.0 0.151 0.155 4.205 4.211 density_rs2pw 110 9.6 0.004 0.004 3.946 3.987 multiply_cannon_sync_h2d 8220 15.4 3.950 3.957 3.950 3.957 cp_dbcsr_syevd 48 12.0 0.012 0.019 3.760 3.760 qs_energies_init_hamiltonians 11 5.9 0.008 0.015 3.758 3.758 grid_integrate_task_list 110 12.3 3.678 3.738 3.678 3.738 hybrid_alltoall_any 4261 16.3 0.257 0.556 2.864 3.534 make_images_data 4110 15.4 0.042 0.045 2.807 3.458 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.887 3.337 wfi_extrapolate 11 7.9 0.001 0.001 3.325 3.325 calculate_dm_sparse 110 9.5 0.001 0.001 3.135 3.161 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.137 3.145 cp_fm_diag_elpa_base 48 14.0 2.592 2.787 3.127 3.128 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.857 2.861 grid_collocate_task_list 110 9.6 2.646 2.710 2.646 2.710 potential_pw2rs 110 12.3 0.021 0.021 2.576 2.580 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.405 2.434 rs_pw_transfer 902 11.9 0.011 0.011 2.374 2.415 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.368 2.381 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.278 2.336 calculate_first_density_matrix 1 7.0 0.000 0.000 2.253 2.253 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.087 2.169 cp_fm_cholesky_decompose 22 10.9 2.069 2.092 2.069 2.092 mp_alltoall_d11v 2046 13.8 1.998 2.056 1.998 2.056 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.972 1.975 qs_create_task_list 11 7.9 0.001 0.001 1.923 1.973 generate_qs_task_list 11 8.9 0.760 0.814 1.923 1.972 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.766 1.806 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=88.475000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1195.181818, yerr=37.790856 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/48320402b71dba81277f3b122ba0d11842227d22_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410023282688 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444707676160 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796579E+12 0.0% 0.0% 100.0% flops max/rank 1.086553E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500928 0.0% 0.0% 100.0% number of processed stacks 11851392 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.8 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 628.756480E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10258080 MPI messages size (bytes): total size 4.456716E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.459062E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3542056 94711185408 32768 < size <= 131072 1282176 73356279808 131072 < size <= 4194304 5107038 3151762635464 4194304 < size <= 16777216 259842 1136842803272 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4083 56784. MP_Allreduce 11248 786. MP_Sync 168 MP_Alltoall 2210 2522944. MP_SendRecv 24130 18752. MP_ISendRecv 24130 18752. MP_Wait 42150 MP_comm_split 82 MP_ISend 15900 108037. MP_IRecv 15900 108037. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.052 204.490 204.492 qs_mol_dyn_low 1 2.0 0.005 0.014 203.840 203.854 qs_forces 11 3.9 0.006 0.026 203.684 203.686 qs_energies 11 4.9 0.002 0.003 197.995 198.012 scf_env_do_scf 11 5.9 0.001 0.001 181.357 181.361 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 160.408 160.410 dbcsr_multiply_generic 2485 12.5 0.178 0.183 123.333 123.901 velocity_verlet 10 3.0 0.013 0.052 122.399 122.402 qs_scf_new_mos 116 7.6 0.001 0.001 120.177 120.379 qs_scf_loop_do_ot 116 8.6 0.001 0.001 120.176 120.378 ot_scf_mini 116 9.6 0.003 0.004 113.627 113.810 multiply_cannon 2485 13.5 0.234 0.244 100.001 102.383 multiply_cannon_loop 2485 14.5 2.094 2.159 97.798 99.695 ot_mini 116 10.6 0.002 0.012 65.347 65.560 multiply_cannon_multrec 59640 15.5 33.011 35.451 41.306 43.514 qs_ot_get_derivative 116 11.6 0.001 0.002 40.658 40.876 rebuild_ks_matrix 127 8.3 0.001 0.001 33.806 34.187 qs_ks_build_kohn_sham_matrix 127 9.3 0.015 0.017 33.805 34.186 mp_waitall_1 264810 16.5 28.071 31.202 28.071 31.202 qs_ks_update_qs_env 127 7.6 0.001 0.001 30.322 30.659 multiply_cannon_sync_h2d 59640 15.5 27.063 28.783 27.063 28.783 qs_ot_get_p 127 10.4 0.001 0.001 26.578 26.761 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 24.202 24.755 apply_single 127 13.6 0.001 0.001 24.202 24.755 ot_diis_step 116 11.6 0.012 0.041 24.355 24.356 init_scf_loop 11 6.9 0.000 0.001 20.870 20.871 qs_ot_p2m_diag 82 11.4 0.076 0.090 19.907 20.006 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 18.897 19.050 multiply_cannon_metrocomm3 59640 15.5 0.115 0.119 15.502 18.550 cp_dbcsr_syevd 82 12.4 0.005 0.006 17.369 17.371 prepare_preconditioner 11 7.9 0.000 0.000 16.237 16.273 make_preconditioner 11 8.9 0.000 0.000 16.237 16.273 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.478 15.643 cp_fm_diag_elpa 82 13.4 0.000 0.000 14.485 14.490 cp_fm_redistribute_end 82 14.4 11.446 14.404 11.458 14.407 sum_up_and_integrate 127 10.3 0.088 0.106 14.392 14.407 integrate_v_rspace 127 11.3 0.003 0.004 14.304 14.323 make_m2s 4970 13.5 0.104 0.113 13.800 14.209 cp_fm_diag_elpa_base 82 14.4 2.907 14.033 2.937 14.140 qs_rho_update_rho_low 127 7.7 0.001 0.002 13.940 14.058 calculate_rho_elec 127 8.7 0.045 0.064 13.939 14.057 make_images 4970 14.5 0.398 0.417 13.619 14.038 init_scf_run 11 5.9 0.000 0.001 12.432 12.432 scf_env_initial_rho_setup 11 6.9 0.001 0.002 12.432 12.432 density_rs2pw 127 9.7 0.006 0.007 7.368 10.522 mp_sum_l 7804 13.0 8.285 9.312 8.285 9.312 cp_fm_cholesky_invert 11 10.9 9.272 9.279 9.272 9.279 wfi_extrapolate 11 7.9 0.001 0.002 9.193 9.193 rs_pw_transfer 1038 11.9 0.016 0.018 5.975 9.163 dbcsr_mm_accdrv_process 123452 16.2 3.267 3.415 7.860 8.448 calculate_dm_sparse 127 9.5 0.001 0.001 8.347 8.421 pw_transfer 1535 11.6 0.076 0.092 8.092 8.337 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 7.888 8.139 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.881 8.011 qs_ot_get_orbitals 116 10.6 0.001 0.001 7.744 7.856 make_images_data 4970 15.5 0.067 0.073 6.721 7.692 multiply_cannon_metrocomm1 59640 15.5 0.089 0.092 5.983 7.689 grid_integrate_task_list 127 12.3 6.973 7.495 6.973 7.495 fft_wrap_pw1pw2_140 519 13.2 0.439 0.509 6.781 7.100 hybrid_alltoall_any 5155 16.4 0.288 2.238 5.870 7.054 fft3d_ps 1281 14.7 2.091 2.882 6.687 6.853 cp_dbcsr_sm_fm_multiply 37 9.5 0.003 0.003 6.710 6.717 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.784 5.923 mp_waitany 15900 13.9 2.628 5.792 2.628 5.792 grid_collocate_task_list 127 9.7 4.531 5.761 4.531 5.761 rs_pw_transfer_RS2PW_140 138 11.5 0.278 0.293 2.090 5.261 mp_alltoall_d11v 2401 14.1 4.237 5.155 4.237 5.155 potential_pw2rs 127 12.3 0.009 0.011 4.849 4.881 cp_fm_cholesky_decompose 22 10.9 4.785 4.800 4.785 4.800 mp_sum_d 4441 12.1 3.557 4.158 3.557 4.158 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=204.492000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=594.818182, yerr=6.886326 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/48320402b71dba81277f3b122ba0d11842227d22_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963542011904 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019182452736 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019182452736 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796564E+12 0.0% 0.0% 100.0% flops max/rank 2.166473E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499488 0.0% 0.0% 100.0% number of processed stacks 5925696 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.6 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 827.551744E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2385600 MPI messages size (bytes): total size 4.069299E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.705776E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70188 2295595008 32768 < size <= 131072 716032 54973693952 131072 < size <= 4194304 1363760 1386318135296 4194304 < size <= 16777216 153648 1453842592400 16777216 < size 67056 1171888537600 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4077 57311. MP_Allreduce 11232 947. MP_Sync 168 MP_Alltoall 1955 5616456. MP_SendRecv 11938 47072. MP_ISendRecv 11938 47072. MP_Wait 25718 MP_comm_split 82 MP_ISend 11660 212488. MP_IRecv 11660 212488. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.086 0.181 191.819 191.820 qs_mol_dyn_low 1 2.0 0.003 0.004 191.221 191.234 qs_forces 11 3.9 0.005 0.015 191.123 191.125 qs_energies 11 4.9 0.002 0.005 184.267 184.282 scf_env_do_scf 11 5.9 0.002 0.015 167.769 167.778 scf_env_do_scf_inner_loop 116 6.6 0.009 0.054 134.581 134.585 velocity_verlet 10 3.0 0.012 0.024 119.928 119.930 dbcsr_multiply_generic 2485 12.5 0.184 0.190 96.780 97.986 qs_scf_new_mos 116 7.6 0.001 0.001 94.224 94.768 qs_scf_loop_do_ot 116 8.6 0.001 0.002 94.224 94.767 ot_scf_mini 116 9.6 0.004 0.005 89.476 89.915 multiply_cannon 2485 13.5 0.473 0.528 76.917 81.063 multiply_cannon_loop 2485 14.5 1.240 1.279 73.688 76.245 ot_mini 116 10.6 0.001 0.001 49.586 50.096 mp_waitall_1 212858 16.6 24.590 39.024 24.590 39.024 multiply_cannon_multrec 29820 15.5 21.841 27.322 31.352 37.041 rebuild_ks_matrix 127 8.3 0.001 0.001 33.523 33.928 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.020 33.522 33.928 init_scf_loop 11 6.9 0.001 0.002 33.089 33.090 qs_ks_update_qs_env 127 7.6 0.001 0.001 30.152 30.523 multiply_cannon_metrocomm3 29820 15.5 0.093 0.098 15.728 28.763 prepare_preconditioner 11 7.9 0.000 0.000 28.658 28.713 make_preconditioner 11 8.9 0.000 0.002 28.658 28.713 qs_ot_get_derivative 116 11.6 0.001 0.002 27.678 28.130 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.375 27.913 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 22.019 23.252 apply_single 127 13.6 0.001 0.001 22.018 23.251 qs_ot_get_p 127 10.4 0.001 0.001 21.733 22.236 ot_diis_step 116 11.6 0.014 0.015 21.731 21.733 multiply_cannon_sync_h2d 29820 15.5 19.276 21.220 19.276 21.220 qs_ot_p2m_diag 82 11.4 0.185 0.213 17.059 17.100 cp_fm_cholesky_invert 11 10.9 16.815 16.829 16.815 16.829 cp_dbcsr_syevd 82 12.4 0.006 0.006 15.908 15.910 make_m2s 4970 13.5 0.090 0.095 14.245 15.751 make_images 4970 14.5 1.138 1.324 14.034 15.544 sum_up_and_integrate 127 10.3 0.113 0.133 14.910 14.939 integrate_v_rspace 127 11.3 0.003 0.004 14.797 14.829 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.237 14.270 calculate_rho_elec 127 8.7 0.087 0.105 14.236 14.269 cp_fm_diag_elpa 82 13.4 0.000 0.001 12.783 12.796 cp_fm_redistribute_end 82 14.4 7.461 12.713 7.472 12.715 cp_fm_diag_elpa_base 82 14.4 5.007 12.214 5.219 12.558 init_scf_run 11 5.9 0.000 0.001 11.615 11.617 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.615 11.617 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 10.787 11.163 multiply_cannon_metrocomm4 27335 15.5 0.093 0.106 3.664 10.529 make_images_data 4970 15.5 0.064 0.070 8.649 10.512 mp_irecv_dv 68888 16.3 3.474 10.145 3.474 10.145 density_rs2pw 127 9.7 0.006 0.006 7.644 9.921 hybrid_alltoall_any 5155 16.4 0.338 1.494 7.444 9.650 dbcsr_mm_accdrv_process 61726 16.2 4.492 5.313 8.979 9.559 pw_transfer 1535 11.6 0.087 0.105 9.115 9.199 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.890 8.978 wfi_extrapolate 11 7.9 0.001 0.001 8.458 8.458 rs_pw_transfer 1038 11.9 0.014 0.016 5.912 8.222 fft_wrap_pw1pw2_140 519 13.2 0.469 0.528 7.790 7.901 grid_integrate_task_list 127 12.3 7.102 7.484 7.102 7.484 fft3d_ps 1281 14.7 2.733 2.904 7.269 7.315 cp_fm_cholesky_decompose 22 10.9 6.981 7.081 6.981 7.081 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.238 7.048 calculate_dm_sparse 127 9.5 0.001 0.001 6.399 6.533 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.120 6.129 mp_sum_l 7804 13.0 3.886 5.914 3.886 5.914 grid_collocate_task_list 127 9.7 4.664 5.814 4.664 5.814 qs_ot_get_orbitals 116 10.6 0.001 0.001 5.322 5.389 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.267 5.387 potential_pw2rs 127 12.3 0.015 0.017 5.233 5.255 mp_alltoall_d11v 2401 14.1 4.163 5.230 4.163 5.230 mp_allgather_i34 2485 14.5 1.816 4.948 1.816 4.948 mp_waitany 11660 13.9 2.487 4.836 2.487 4.836 rs_pw_transfer_RS2PW_140 138 11.5 0.350 0.373 2.085 4.373 mp_sum_d 4438 12.1 2.664 4.001 2.664 4.001 dbcsr_complete_redistribute 393 12.7 0.769 0.850 3.152 3.996 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=191.820000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=787.909091, yerr=4.521756 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/48320402b71dba81277f3b122ba0d11842227d22_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420241154048 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 11528896499712 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 93.514757E+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 6755939872 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 932.806656E+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 931531083728 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 57335. MP_Allreduce 11226 986. 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 0.021 0.041 176.456 176.456 qs_mol_dyn_low 1 2.0 0.003 0.004 175.800 175.813 qs_forces 11 3.9 0.004 0.010 175.692 175.698 qs_energies 11 4.9 0.001 0.002 169.053 169.065 scf_env_do_scf 11 5.9 0.001 0.001 153.332 153.333 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 118.564 118.565 velocity_verlet 10 3.0 0.001 0.002 112.942 112.945 dbcsr_multiply_generic 2507 12.6 0.182 0.187 81.479 82.821 qs_scf_new_mos 117 7.6 0.001 0.001 80.248 80.606 qs_scf_loop_do_ot 117 8.6 0.001 0.001 80.247 80.605 ot_scf_mini 117 9.6 0.004 0.004 76.118 76.501 multiply_cannon 2507 13.6 0.502 0.524 61.664 66.836 multiply_cannon_loop 2507 14.6 0.857 0.896 58.656 61.835 ot_mini 117 10.6 0.001 0.001 42.594 43.034 init_scf_loop 11 6.9 0.000 0.000 34.669 34.670 mp_waitall_1 170520 16.6 24.416 33.442 24.416 33.442 rebuild_ks_matrix 128 8.3 0.001 0.001 30.741 31.225 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.018 30.740 31.224 prepare_preconditioner 11 7.9 0.000 0.000 30.648 30.695 make_preconditioner 11 8.9 0.000 0.000 30.648 30.695 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.311 29.738 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.693 28.135 multiply_cannon_multrec 20056 15.6 13.654 17.087 22.565 26.113 multiply_cannon_metrocomm3 20056 15.6 0.060 0.064 14.790 24.680 qs_ot_get_derivative 117 11.6 0.001 0.002 22.994 23.395 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.773 20.794 apply_single 128 13.6 0.001 0.001 19.773 20.794 qs_ot_get_p 128 10.4 0.001 0.001 19.079 19.620 ot_diis_step 117 11.6 0.018 0.018 19.504 19.505 multiply_cannon_sync_h2d 20056 15.6 14.414 16.213 14.414 16.213 make_m2s 5014 13.6 0.081 0.087 14.625 15.534 make_images 5014 14.6 1.166 1.260 14.390 15.299 sum_up_and_integrate 128 10.3 0.131 0.143 14.736 14.763 qs_ot_p2m_diag 83 11.4 0.265 0.272 14.710 14.720 integrate_v_rspace 128 11.3 0.003 0.004 14.604 14.634 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.410 14.434 calculate_rho_elec 128 8.7 0.132 0.147 14.409 14.433 cp_fm_cholesky_invert 11 10.9 14.207 14.215 14.207 14.215 cp_dbcsr_syevd 83 12.4 0.006 0.006 13.712 13.713 init_scf_run 11 5.9 0.000 0.001 10.676 10.676 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.676 10.676 cp_fm_diag_elpa 83 13.4 0.000 0.001 10.594 10.595 cp_fm_redistribute_end 83 14.4 4.007 10.539 4.021 10.542 cp_fm_diag_elpa_base 83 14.4 6.095 9.943 6.501 10.423 make_images_data 5014 15.6 0.061 0.069 9.012 10.327 hybrid_alltoall_any 5200 16.5 0.433 1.982 7.876 9.510 pw_transfer 1547 11.6 0.087 0.108 9.166 9.283 density_rs2pw 128 9.7 0.006 0.006 7.278 9.265 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 8.985 9.242 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 8.941 9.064 multiply_cannon_metrocomm4 17549 15.6 0.061 0.072 3.358 8.900 mp_irecv_dv 50230 16.2 3.236 8.658 3.236 8.658 dbcsr_mm_accdrv_process 41502 16.2 4.529 5.487 8.356 8.501 fft_wrap_pw1pw2_140 523 13.2 0.478 0.523 7.923 8.053 grid_integrate_task_list 128 12.3 7.295 7.837 7.295 7.837 cp_fm_upper_to_full 105 14.5 5.842 7.489 5.842 7.489 wfi_extrapolate 11 7.9 0.001 0.001 7.393 7.393 fft3d_ps 1291 14.7 2.665 2.905 7.205 7.294 cp_fm_cholesky_decompose 22 10.9 7.238 7.270 7.238 7.270 rs_pw_transfer 1046 11.9 0.014 0.015 5.095 7.071 dbcsr_complete_redistribute 395 12.7 1.169 1.205 4.530 6.304 calculate_dm_sparse 128 9.5 0.001 0.001 5.877 5.984 grid_collocate_task_list 128 9.7 4.905 5.794 4.905 5.794 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.505 5.510 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.653 5.360 mp_sum_l 7870 13.0 3.365 5.279 3.365 5.279 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.370 5.142 mp_alltoall_d11v 2415 14.1 4.280 5.138 4.280 5.138 potential_pw2rs 128 12.3 0.020 0.022 4.813 4.823 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.619 4.749 mp_waitany 11748 13.9 2.335 4.289 2.335 4.289 mp_allgather_i34 2507 14.6 1.534 4.231 1.534 4.231 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.317 4.052 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.021 4.049 mp_alltoall_i22 716 14.1 1.968 3.896 1.968 3.896 rs_pw_transfer_RS2PW_140 139 11.5 0.332 0.356 1.928 3.888 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.789 3.791 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.561 3.603 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=176.456000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=885.090909, yerr=9.069976 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/48320402b71dba81277f3b122ba0d11842227d22_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 4.353788E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 5977344 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.167507E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1143192 MPI messages size (bytes): total size 2.023815E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.770319E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 319024 36042702848 131072 < size <= 4194304 715736 785529176064 4194304 < size <= 16777216 70320 665379241840 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 57332. MP_Allreduce 11226 1068. 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 0.053 0.088 187.479 187.489 qs_mol_dyn_low 1 2.0 0.003 0.004 186.882 186.894 qs_forces 11 3.9 0.004 0.005 186.771 186.775 qs_energies 11 4.9 0.002 0.005 179.531 179.541 scf_env_do_scf 11 5.9 0.002 0.011 162.717 162.724 velocity_verlet 10 3.0 0.008 0.020 123.894 123.896 scf_env_do_scf_inner_loop 117 6.6 0.005 0.025 116.517 116.528 dbcsr_multiply_generic 2507 12.6 0.224 0.232 78.757 79.340 qs_scf_new_mos 117 7.6 0.001 0.001 78.934 79.190 qs_scf_loop_do_ot 117 8.6 0.001 0.002 78.933 79.189 ot_scf_mini 117 9.6 0.005 0.013 74.453 74.729 multiply_cannon 2507 13.6 0.558 0.591 54.503 56.581 multiply_cannon_loop 2507 14.6 1.184 1.211 50.878 52.168 init_scf_loop 11 6.9 0.001 0.002 46.071 46.112 ot_mini 117 10.6 0.001 0.001 41.820 42.098 prepare_preconditioner 11 7.9 0.000 0.001 41.774 41.796 make_preconditioner 11 8.9 0.000 0.001 41.774 41.796 make_full_inverse_cholesky 11 9.9 0.000 0.000 35.454 40.394 multiply_cannon_multrec 30084 15.6 14.229 19.241 26.197 30.633 rebuild_ks_matrix 128 8.3 0.001 0.001 30.047 30.223 qs_ks_build_kohn_sham_matrix 128 9.3 0.020 0.031 30.046 30.223 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.010 27.175 mp_waitall_1 147882 16.7 16.772 26.931 16.772 26.931 qs_ot_get_derivative 117 11.6 0.001 0.002 22.284 22.564 make_m2s 5014 13.6 0.095 0.101 19.915 20.920 make_images 5014 14.6 1.939 2.240 19.607 20.612 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 18.996 19.476 apply_single 128 13.6 0.001 0.001 18.995 19.476 ot_diis_step 117 11.6 0.017 0.018 19.414 19.416 qs_ot_get_p 128 10.4 0.002 0.008 19.080 19.394 cp_fm_upper_to_full 105 14.7 10.962 16.195 10.962 16.195 cp_fm_cholesky_invert 11 10.9 15.956 15.964 15.956 15.964 qs_ot_p2m_diag 83 11.4 0.343 0.390 15.059 15.110 sum_up_and_integrate 128 10.3 0.140 0.153 14.798 14.833 multiply_cannon_metrocomm3 30084 15.6 0.047 0.050 6.168 14.811 integrate_v_rspace 128 11.3 0.004 0.004 14.657 14.697 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.614 14.673 calculate_rho_elec 128 8.7 0.175 0.190 14.614 14.673 cp_dbcsr_syevd 83 12.4 0.006 0.006 13.788 13.789 multiply_cannon_sync_h2d 30084 15.6 11.781 12.736 11.781 12.736 dbcsr_complete_redistribute 395 12.7 1.512 1.659 9.032 12.708 make_images_data 5014 15.6 0.064 0.068 10.702 12.191 dbcsr_mm_accdrv_process 62264 16.2 7.445 8.430 11.542 12.066 hybrid_alltoall_any 5200 16.5 0.525 2.209 9.651 11.385 copy_fm_to_dbcsr 209 11.7 0.002 0.002 7.703 11.377 init_scf_run 11 5.9 0.000 0.001 10.929 10.931 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.929 10.930 cp_fm_diag_elpa 83 13.4 0.000 0.000 10.633 10.635 cp_fm_redistribute_end 83 14.4 1.834 10.563 1.847 10.569 cp_fm_diag_elpa_base 83 14.4 8.114 9.948 8.691 10.447 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.296 9.880 pw_transfer 1547 11.6 0.088 0.106 9.437 9.503 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.174 9.390 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 9.209 9.281 mp_alltoall_i22 716 14.1 5.512 9.204 5.512 9.204 density_rs2pw 128 9.7 0.006 0.006 7.051 8.367 fft_wrap_pw1pw2_140 523 13.2 0.484 0.499 8.164 8.259 grid_integrate_task_list 128 12.3 7.521 7.859 7.521 7.859 wfi_extrapolate 11 7.9 0.001 0.001 7.642 7.642 cp_fm_cholesky_decompose 22 10.9 7.434 7.524 7.434 7.524 fft3d_ps 1291 14.7 2.775 2.853 7.432 7.510 multiply_cannon_metrocomm4 25070 15.6 0.075 0.085 2.753 7.432 mp_irecv_dv 76098 16.2 2.606 7.159 2.606 7.159 calculate_dm_sparse 128 9.5 0.001 0.001 6.231 6.307 rs_pw_transfer 1046 11.9 0.013 0.015 4.611 5.995 grid_collocate_task_list 128 9.7 5.089 5.791 5.089 5.791 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.516 5.558 mp_alltoall_d11v 2415 14.1 4.815 5.538 4.815 5.538 potential_pw2rs 128 12.3 0.023 0.023 4.676 4.703 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.471 4.520 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 4.451 4.452 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.320 4.404 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.171 4.230 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=187.489000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1081.000000, yerr=27.820855 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/48320402b71dba81277f3b122ba0d11842227d22_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019187724288 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019187724288 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796577E+12 0.0% 0.0% 100.0% flops max/rank 5.820059E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500544 0.0% 0.0% 100.0% number of processed stacks 1944496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3448.5 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.488876E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 238560 MPI messages size (bytes): total size 1.321104E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.537828E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 112800 59139686400 4194304 < size <= 16777216 104112 545846722560 16777216 < size 20064 716108613552 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8852 52. MP_Alltoall 9584 804353. MP_ISend 39716 2104723. MP_IRecv 39716 2103824. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3949 58975. MP_Allreduce 10870 1179. MP_Sync 86 MP_Alltoall 1700 18828209. MP_SendRecv 3810 122880. MP_ISendRecv 3810 122880. MP_Wait 16000 MP_ISend 10600 423612. MP_IRecv 10600 423612. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.024 0.071 172.852 172.858 qs_mol_dyn_low 1 2.0 0.003 0.003 172.418 172.430 qs_forces 11 3.9 0.003 0.003 172.306 172.312 qs_energies 11 4.9 0.002 0.002 164.704 164.712 scf_env_do_scf 11 5.9 0.001 0.013 147.440 147.450 velocity_verlet 10 3.0 0.013 0.014 113.156 113.160 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 112.068 112.069 dbcsr_multiply_generic 2485 12.5 0.181 0.186 74.159 74.724 qs_scf_new_mos 116 7.6 0.001 0.001 73.473 73.595 qs_scf_loop_do_ot 116 8.6 0.001 0.001 73.473 73.594 ot_scf_mini 116 9.6 0.003 0.004 69.117 69.260 multiply_cannon 2485 13.5 0.580 0.608 54.348 58.587 multiply_cannon_loop 2485 14.5 0.443 0.467 49.622 50.796 ot_mini 116 10.6 0.001 0.001 39.606 39.763 init_scf_loop 11 6.9 0.000 0.000 35.214 35.216 mp_waitall_1 124680 16.7 26.100 33.923 26.100 33.923 prepare_preconditioner 11 7.9 0.000 0.000 31.232 31.258 make_preconditioner 11 8.9 0.000 0.000 31.232 31.258 rebuild_ks_matrix 127 8.3 0.001 0.001 29.958 30.099 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.018 29.957 30.098 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.176 29.425 qs_ks_update_qs_env 127 7.6 0.001 0.001 27.157 27.288 multiply_cannon_multrec 9940 15.5 10.298 14.446 17.587 21.897 multiply_cannon_metrocomm3 9940 15.5 0.023 0.024 12.679 19.995 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.442 19.928 apply_single 127 13.6 0.001 0.001 19.441 19.928 ot_diis_step 116 11.6 0.020 0.020 19.888 19.889 qs_ot_get_derivative 116 11.6 0.001 0.002 19.651 19.801 make_m2s 4970 13.5 0.066 0.069 16.199 18.578 make_images 4970 14.5 2.263 2.616 15.892 18.273 cp_fm_cholesky_invert 11 10.9 17.860 17.866 17.860 17.866 qs_ot_get_p 127 10.4 0.001 0.001 16.522 16.670 qs_rho_update_rho_low 127 7.7 0.001 0.001 15.382 15.437 calculate_rho_elec 127 8.7 0.257 0.266 15.382 15.436 sum_up_and_integrate 127 10.3 0.179 0.189 15.031 15.075 integrate_v_rspace 127 11.3 0.004 0.004 14.852 14.903 qs_ot_p2m_diag 82 11.4 0.489 0.495 12.828 12.845 make_images_data 4970 15.5 0.051 0.059 9.885 12.378 multiply_cannon_sync_h2d 9940 15.5 11.434 11.986 11.434 11.986 hybrid_alltoall_any 5155 16.4 0.805 3.760 9.702 11.956 cp_dbcsr_syevd 82 12.4 0.012 0.012 11.722 11.723 init_scf_run 11 5.9 0.000 0.001 10.374 10.374 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.374 10.374 pw_transfer 1535 11.6 0.085 0.094 9.870 9.903 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 9.648 9.693 cp_fm_diag_elpa 82 13.4 0.000 0.000 8.776 8.786 cp_fm_diag_elpa_base 82 14.4 8.543 8.613 8.772 8.784 fft_wrap_pw1pw2_140 519 13.2 0.494 0.510 8.535 8.587 grid_integrate_task_list 127 12.3 7.700 8.123 7.700 8.123 cp_fm_cholesky_decompose 22 10.9 7.763 7.913 7.763 7.913 qs_ot_get_derivative_diag 76 12.4 0.002 0.003 7.765 7.870 density_rs2pw 127 9.7 0.005 0.006 6.988 7.852 fft3d_ps 1281 14.7 2.690 2.783 7.824 7.841 multiply_cannon_metrocomm1 9940 15.5 0.029 0.029 4.290 7.632 mp_allgather_i34 2485 14.5 2.954 7.586 2.954 7.586 dbcsr_mm_accdrv_process 20590 16.1 2.650 3.499 6.924 7.548 wfi_extrapolate 11 7.9 0.001 0.001 7.468 7.468 calculate_dm_sparse 127 9.5 0.001 0.001 6.002 6.098 grid_collocate_task_list 127 9.7 5.353 6.050 5.353 6.050 mp_alltoall_d11v 2401 14.1 4.928 5.861 4.928 5.861 dbcsr_complete_redistribute 393 12.7 2.116 2.171 5.054 5.458 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 5.380 5.381 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.177 5.188 multiply_cannon_metrocomm4 7455 15.5 0.024 0.027 1.863 5.149 mp_irecv_dv 28618 15.9 1.824 5.067 1.824 5.067 rs_pw_transfer 1038 11.9 0.012 0.013 4.043 4.947 potential_pw2rs 127 12.3 0.026 0.027 4.536 4.548 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.116 4.148 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.556 3.850 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.700 3.713 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.680 3.713 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.526 3.619 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.290 3.610 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.455 3.482 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=172.858000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1388.272727, yerr=41.848408 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/48320402b71dba81277f3b122ba0d11842227d22_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456039424 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613072052224 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176077312 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176077312 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.233020E+12 0.0% 0.0% 100.0% flops max/rank 11.786061E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383904 0.0% 0.0% 100.0% number of processed stacks 1980288 0.0% 0.0% 100.0% average stack size 0.0 0.0 3437.1 marketing flops 145.650931E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 2.938470E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101160 MPI messages size (bytes): total size 1.144970E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.318403E+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 45648 35433480192 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592466352 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4045 58536. MP_Allreduce 11109 1511. MP_Sync 88 MP_Alltoall 1724 36993632. MP_SendRecv 1806 218624. MP_ISendRecv 1806 218624. MP_Wait 9876 MP_ISend 6456 1080169. MP_IRecv 6456 1080169. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.032 0.071 298.839 298.855 qs_mol_dyn_low 1 2.0 0.003 0.003 298.239 298.250 qs_forces 11 3.9 0.042 0.043 298.138 298.140 qs_energies 11 4.9 0.002 0.002 288.726 288.735 scf_env_do_scf 11 5.9 0.001 0.007 266.171 266.187 velocity_verlet 10 3.0 0.022 0.023 215.677 215.685 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 142.138 142.140 init_scf_loop 11 6.9 0.000 0.000 123.772 123.776 prepare_preconditioner 11 7.9 0.000 0.000 118.563 118.601 make_preconditioner 11 8.9 0.000 0.000 118.563 118.601 make_full_inverse_cholesky 11 9.9 0.000 0.000 94.891 115.713 qs_scf_new_mos 118 7.6 0.001 0.001 90.671 90.824 qs_scf_loop_do_ot 118 8.6 0.001 0.001 90.670 90.824 ot_scf_mini 118 9.6 0.004 0.004 85.791 85.893 dbcsr_multiply_generic 2529 12.6 0.215 0.221 84.163 84.933 cp_fm_upper_to_full 106 14.8 52.711 75.319 52.711 75.319 multiply_cannon 2529 13.6 0.728 0.768 59.454 60.436 multiply_cannon_loop 2529 14.6 0.477 0.491 55.756 57.753 ot_mini 118 10.6 0.001 0.001 44.964 45.088 dbcsr_complete_redistribute 397 12.7 3.966 4.016 29.300 41.989 rebuild_ks_matrix 129 8.3 0.001 0.001 38.953 39.105 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.018 38.952 39.104 copy_fm_to_dbcsr 210 11.7 0.002 0.002 25.933 38.559 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 23.627 36.116 qs_ks_update_qs_env 129 7.6 0.001 0.001 35.911 36.045 mp_alltoall_i22 720 14.1 21.459 33.999 21.459 33.999 cp_fm_cholesky_invert 11 10.9 33.207 33.213 33.207 33.213 mp_waitall_1 104580 16.8 28.747 33.159 28.747 33.159 qs_ot_get_p 129 10.4 0.001 0.001 25.269 25.393 qs_ot_get_derivative 118 11.6 0.002 0.002 24.324 24.428 qs_ot_p2m_diag 84 11.4 0.889 0.895 21.215 21.245 multiply_cannon_metrocomm3 10116 15.6 0.024 0.025 19.311 21.101 make_m2s 5058 13.6 0.077 0.080 19.915 20.912 qs_rho_update_rho_low 129 7.7 0.001 0.001 20.788 20.805 calculate_rho_elec 129 8.7 0.487 0.487 20.788 20.804 ot_diis_step 118 11.6 0.022 0.022 20.596 20.596 make_images 5058 14.6 3.786 3.927 19.431 20.433 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.650 20.064 apply_single 129 13.6 0.001 0.001 19.650 20.064 sum_up_and_integrate 129 10.3 0.323 0.325 19.871 19.954 integrate_v_rspace 129 11.3 0.004 0.004 19.548 19.629 cp_dbcsr_syevd 84 12.4 0.011 0.011 19.482 19.484 multiply_cannon_multrec 10116 15.6 10.577 12.320 18.089 18.167 cp_fm_diag_elpa 84 13.4 0.000 0.000 16.292 16.293 cp_fm_diag_elpa_base 84 14.4 11.945 13.530 16.288 16.288 multiply_cannon_sync_h2d 10116 15.6 15.777 15.796 15.777 15.796 hybrid_alltoall_any 5245 16.5 1.311 3.060 10.895 12.794 make_images_data 5058 15.6 0.061 0.066 10.667 12.465 init_scf_run 11 5.9 0.000 0.001 12.429 12.429 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.428 12.429 pw_transfer 1559 11.6 0.094 0.095 12.412 12.423 fft_wrap_pw1pw2 1301 12.7 0.011 0.011 12.174 12.185 fft_wrap_pw1pw2_140 527 13.2 0.547 0.548 10.832 10.842 fft3d_ps 1301 14.7 2.765 2.770 10.192 10.209 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.536 9.615 wfi_extrapolate 11 7.9 0.001 0.001 9.162 9.162 dbcsr_mm_accdrv_process 20934 16.1 3.830 5.830 7.272 9.140 mp_alltoall_d11v 2429 14.1 8.143 8.937 8.143 8.937 grid_integrate_task_list 129 12.3 8.650 8.838 8.650 8.838 cp_fm_cholesky_decompose 22 10.9 8.717 8.735 8.717 8.735 density_rs2pw 129 9.7 0.005 0.005 8.331 8.465 qs_energies_init_hamiltonians 11 5.9 0.002 0.003 8.122 8.126 calculate_dm_sparse 129 9.5 0.001 0.001 6.768 6.845 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.386 6.466 grid_collocate_task_list 129 9.7 6.419 6.466 6.419 6.466 rs_scatter_matrices 140 9.7 3.673 4.499 5.991 6.206 copy_dbcsr_to_fm 187 11.8 0.004 0.004 6.058 6.123 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=298.855000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2604.909091, yerr=171.207718 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/48320402b71dba81277f3b122ba0d11842227d22_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 1.261883E+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 0.035 0.084 85.194 85.195 qs_energies 1 2.0 0.000 0.000 84.604 84.612 ls_scf 1 3.0 0.000 0.000 83.691 83.698 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.612 72.753 multiply_cannon 111 7.7 0.018 0.021 55.885 57.041 multiply_cannon_loop 111 8.7 0.213 0.227 52.411 53.963 ls_scf_main 1 4.0 0.000 0.000 52.545 52.546 density_matrix_trs4 2 5.0 0.002 0.004 47.061 47.174 ls_scf_init_scf 1 4.0 0.000 0.000 28.126 28.127 ls_scf_init_matrix_S 1 5.0 0.000 0.000 26.997 27.043 mp_waitall_1 11031 10.9 21.822 26.071 21.822 26.071 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 24.852 24.875 multiply_cannon_multrec 2664 9.7 8.240 8.997 15.573 17.332 multiply_cannon_sync_h2d 2664 9.7 14.092 15.858 14.092 15.858 make_m2s 222 7.7 0.008 0.011 13.094 13.722 make_images 222 8.7 0.099 0.110 13.072 13.701 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.253 12.356 make_images_data 222 9.7 0.004 0.005 7.635 8.300 dbcsr_mm_accdrv_process 4760 10.4 0.511 0.621 6.948 7.903 hybrid_alltoall_any 227 10.6 0.215 1.831 6.530 7.866 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.260 7.660 dbcsr_mm_accdrv_process_sort 4760 11.4 6.238 7.115 6.238 7.115 calculate_norms 4752 9.8 5.601 6.186 5.601 6.186 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.908 5.114 mp_sum_l 807 5.4 3.103 4.544 3.103 4.544 multiply_cannon_metrocomm4 2442 9.7 0.012 0.014 2.034 3.647 mp_irecv_dv 6231 10.9 2.017 3.614 2.017 3.614 make_images_sizes 222 9.7 0.000 0.000 0.847 3.388 mp_alltoall_i44 222 10.7 0.846 3.388 0.846 3.388 arnoldi_extremal 4 6.8 0.000 0.000 3.279 3.313 arnoldi_normal_ev 4 7.8 0.001 0.004 3.278 3.313 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.270 3.308 build_subspace 16 8.4 0.009 0.012 3.171 3.173 ls_scf_post 1 4.0 0.000 0.000 3.020 3.027 ls_scf_store_result 1 5.0 0.000 0.000 2.828 2.873 dbcsr_special_finalize 555 9.7 0.005 0.006 2.259 2.738 dbcsr_merge_single_wm 555 10.7 0.456 0.586 2.252 2.729 make_images_pack 222 9.7 2.207 2.628 2.209 2.630 dbcsr_matrix_vector_mult 304 9.0 0.005 0.013 2.330 2.600 dbcsr_sort_data 658 11.4 2.056 2.497 2.056 2.497 dbcsr_matrix_vector_mult_local 304 10.0 2.069 2.457 2.071 2.459 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.318 2.406 buffer_matrices_ensure_size 222 8.7 1.750 2.061 1.750 2.061 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.815 1.816 rebuild_ks_matrix 3 7.3 0.000 0.000 1.805 1.806 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.805 1.806 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.195000, 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/48320402b71dba81277f3b122ba0d11842227d22_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 2.099126E+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 0.050 0.065 91.416 91.417 qs_energies 1 2.0 0.000 0.000 90.821 90.825 ls_scf 1 3.0 0.000 0.000 89.507 89.511 dbcsr_multiply_generic 111 6.7 0.015 0.016 75.424 75.678 multiply_cannon 111 7.7 0.028 0.043 53.341 57.188 ls_scf_main 1 4.0 0.000 0.000 55.204 55.208 multiply_cannon_loop 111 8.7 0.115 0.122 50.056 53.464 density_matrix_trs4 2 5.0 0.002 0.003 49.438 49.659 ls_scf_init_scf 1 4.0 0.000 0.000 30.846 30.848 mp_waitall_1 9105 10.9 21.145 30.580 21.145 30.580 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.619 29.717 multiply_cannon_multrec 1332 9.7 13.087 16.982 22.309 27.285 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.170 27.183 multiply_cannon_metrocomm3 1332 9.7 0.006 0.007 11.906 21.221 make_m2s 222 7.7 0.006 0.008 15.149 15.775 make_images 222 8.7 1.572 1.919 15.119 15.745 dbcsr_mm_accdrv_process 4041 10.4 0.267 0.447 8.819 10.364 dbcsr_mm_accdrv_process_sort 4041 11.4 8.413 9.917 8.413 9.917 make_images_data 222 9.7 0.004 0.004 8.602 9.526 hybrid_alltoall_any 227 10.6 0.521 2.423 8.001 8.984 mp_sum_l 807 5.4 5.636 8.664 5.636 8.664 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.208 7.803 mp_irecv_dv 3311 11.0 3.189 7.747 3.189 7.747 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.346 6.876 calculate_norms 2376 9.8 5.996 6.812 5.996 6.812 multiply_cannon_sync_h2d 1332 9.7 4.791 5.836 4.791 5.836 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.037 5.368 arnoldi_extremal 4 6.8 0.000 0.000 4.672 4.698 arnoldi_normal_ev 4 7.8 0.001 0.005 4.672 4.698 build_subspace 16 8.4 0.014 0.021 4.418 4.420 ls_scf_post 1 4.0 0.000 0.000 3.457 3.461 dbcsr_matrix_vector_mult 304 9.0 0.009 0.021 3.150 3.403 ls_scf_store_result 1 5.0 0.000 0.000 3.209 3.280 dbcsr_matrix_vector_mult_local 304 10.0 2.731 3.214 2.733 3.216 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.270 2.973 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.628 2.759 make_images_pack 222 9.7 2.018 2.410 2.021 2.412 mp_allgather_i34 111 8.7 0.980 2.247 0.980 2.247 dbcsr_sort_data 436 11.2 1.808 2.077 1.808 2.077 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.912 1.914 rebuild_ks_matrix 3 7.3 0.000 0.000 1.899 1.901 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.899 1.901 dbcsr_data_new 4174 10.1 1.611 1.863 1.611 1.863 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.417000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1743.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/48320402b71dba81277f3b122ba0d11842227d22_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 2.703081E+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 0.054 0.058 94.669 94.669 qs_energies 1 2.0 0.000 0.000 93.972 93.975 ls_scf 1 3.0 0.000 0.000 92.559 92.562 dbcsr_multiply_generic 111 6.7 0.015 0.016 77.367 77.724 ls_scf_main 1 4.0 0.000 0.000 58.173 58.179 multiply_cannon 111 7.7 0.049 0.131 52.878 56.803 multiply_cannon_loop 111 8.7 0.100 0.107 49.293 53.865 density_matrix_trs4 2 5.0 0.002 0.003 52.195 52.391 mp_waitall_1 7281 11.0 23.897 34.685 23.897 34.685 ls_scf_init_scf 1 4.0 0.000 0.000 30.796 30.798 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.589 29.675 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.181 27.196 multiply_cannon_multrec 888 9.7 12.609 15.233 21.154 24.399 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.948 24.014 make_m2s 222 7.7 0.006 0.006 17.261 18.499 make_images 222 8.7 1.969 2.300 17.221 18.460 hybrid_alltoall_any 227 10.6 0.620 2.871 9.296 11.030 make_images_data 222 9.7 0.003 0.004 9.780 10.865 mp_sum_l 807 5.4 5.667 9.258 5.667 9.258 dbcsr_mm_accdrv_process 3754 10.4 0.247 0.411 8.059 9.244 dbcsr_mm_accdrv_process_sort 3754 11.4 7.689 8.834 7.689 8.834 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.503 7.338 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.794 7.228 multiply_cannon_sync_h2d 888 9.7 6.059 7.218 6.059 7.218 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.465 7.003 mp_irecv_dv 2335 11.1 2.449 6.942 2.449 6.942 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.928 5.122 arnoldi_extremal 4 6.8 0.000 0.000 5.042 5.058 arnoldi_normal_ev 4 7.8 0.001 0.005 5.042 5.058 build_subspace 16 8.4 0.014 0.020 4.736 4.742 calculate_norms 1584 9.8 4.258 4.627 4.258 4.627 mp_allgather_i34 111 8.7 1.420 3.891 1.420 3.891 dbcsr_matrix_vector_mult 304 9.0 0.009 0.021 3.420 3.745 ls_scf_post 1 4.0 0.000 0.000 3.590 3.594 dbcsr_matrix_vector_mult_local 304 10.0 3.002 3.575 3.004 3.577 ls_scf_store_result 1 5.0 0.000 0.000 3.321 3.421 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.859 2.970 make_images_sizes 222 9.7 0.000 0.000 1.169 2.276 mp_alltoall_i44 222 10.7 1.168 2.275 1.168 2.275 dbcsr_sort_data 325 11.1 1.879 2.138 1.879 2.138 make_images_pack 222 9.7 1.816 2.134 1.819 2.137 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.925 1.927 rebuild_ks_matrix 3 7.3 0.000 0.000 1.907 1.909 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.907 1.909 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=94.669000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2186.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/48320402b71dba81277f3b122ba0d11842227d22_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 3.370992E+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 0.048 0.066 97.188 97.189 qs_energies 1 2.0 0.000 0.000 96.441 96.451 ls_scf 1 3.0 0.001 0.006 94.778 94.789 dbcsr_multiply_generic 111 6.7 0.016 0.018 78.446 78.685 ls_scf_main 1 4.0 0.000 0.000 58.844 58.846 multiply_cannon 111 7.7 0.053 0.131 51.628 56.056 density_matrix_trs4 2 5.0 0.002 0.016 52.707 52.864 multiply_cannon_loop 111 8.7 0.115 0.126 46.570 50.212 ls_scf_init_scf 1 4.0 0.000 0.002 32.671 32.673 ls_scf_init_matrix_S 1 5.0 0.000 0.002 31.323 31.392 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.849 28.861 mp_waitall_1 6369 11.0 22.522 28.221 22.522 28.221 multiply_cannon_multrec 1332 9.7 14.157 17.513 22.007 24.659 make_m2s 222 7.7 0.007 0.008 20.850 22.304 make_images 222 8.7 3.129 3.588 20.800 22.256 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.106 17.571 make_images_data 222 9.7 0.004 0.005 11.524 13.130 hybrid_alltoall_any 227 10.6 0.799 3.822 10.888 12.809 dbcsr_mm_accdrv_process 3641 10.4 0.203 0.381 7.496 9.024 dbcsr_mm_accdrv_process_sort 3641 11.4 7.123 8.605 7.123 8.605 mp_sum_l 807 5.4 4.392 8.490 4.392 8.490 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.334 6.757 multiply_cannon_sync_h2d 1332 9.7 5.524 6.145 5.524 6.145 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.728 6.054 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.077 6.014 mp_irecv_dv 3229 10.9 2.055 5.940 2.055 5.940 arnoldi_extremal 4 6.8 0.000 0.000 5.214 5.230 arnoldi_normal_ev 4 7.8 0.001 0.005 5.214 5.230 build_subspace 16 8.4 0.014 0.021 4.886 4.893 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.561 4.803 mp_allgather_i34 111 8.7 2.183 4.690 2.183 4.690 calculate_norms 2376 9.8 4.191 4.579 4.191 4.579 dbcsr_matrix_vector_mult 304 9.0 0.010 0.020 3.570 3.865 dbcsr_matrix_vector_mult_local 304 10.0 3.163 3.662 3.165 3.664 dbcsr_sort_data 658 11.4 3.070 3.386 3.070 3.386 ls_scf_post 1 4.0 0.001 0.010 3.262 3.273 dbcsr_special_finalize 555 9.7 0.006 0.007 2.820 3.231 dbcsr_merge_single_wm 555 10.7 0.538 0.665 2.811 3.223 ls_scf_dm_to_ks 2 5.0 0.001 0.011 3.015 3.069 ls_scf_store_result 1 5.0 0.000 0.000 2.991 3.069 dbcsr_data_release 10477 10.7 1.576 2.449 1.576 2.449 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.069 2.070 rebuild_ks_matrix 3 7.3 0.000 0.000 2.046 2.047 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.014 2.046 2.047 dbcsr_finalize 304 7.8 0.049 0.061 1.799 1.975 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.189000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2730.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/48320402b71dba81277f3b122ba0d11842227d22_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 4.623544E+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 0.041 0.055 99.281 99.281 qs_energies 1 2.0 0.000 0.000 98.377 98.384 ls_scf 1 3.0 0.000 0.000 96.428 96.440 dbcsr_multiply_generic 111 6.7 0.017 0.018 77.759 77.988 ls_scf_main 1 4.0 0.000 0.000 62.246 62.246 multiply_cannon 111 7.7 0.086 0.152 55.242 60.624 density_matrix_trs4 2 5.0 0.002 0.003 54.949 55.051 multiply_cannon_loop 111 8.7 0.070 0.076 50.839 52.371 mp_waitall_1 5436 11.0 26.583 32.334 26.583 32.334 ls_scf_init_scf 1 4.0 0.000 0.000 30.555 30.560 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.209 29.266 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.048 27.063 multiply_cannon_multrec 444 9.7 13.961 16.377 21.004 22.685 make_m2s 222 7.7 0.004 0.005 17.827 20.223 make_images 222 8.7 3.726 4.433 17.766 20.162 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.844 16.323 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.525 14.362 make_images_data 222 9.7 0.003 0.004 10.015 12.414 hybrid_alltoall_any 227 10.6 0.790 3.760 9.858 12.284 dbcsr_mm_accdrv_process 3003 10.4 0.164 0.345 6.742 7.863 multiply_cannon_sync_h2d 444 9.7 6.566 7.520 6.566 7.520 dbcsr_mm_accdrv_process_sort 3003 11.4 6.426 7.518 6.426 7.518 mp_allgather_i34 111 8.7 2.637 7.040 2.637 7.040 arnoldi_extremal 4 6.8 0.000 0.000 5.761 5.778 arnoldi_normal_ev 4 7.8 0.002 0.005 5.761 5.778 build_subspace 16 8.4 0.015 0.020 5.375 5.385 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.571 4.719 mp_sum_l 807 5.4 2.747 4.340 2.747 4.340 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.699 4.337 dbcsr_matrix_vector_mult 304 9.0 0.011 0.020 4.138 4.336 mp_irecv_dv 1241 11.2 1.679 4.308 1.679 4.308 dbcsr_matrix_vector_mult_local 304 10.0 3.675 4.143 3.677 4.146 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.711 3.804 calculate_norms 792 9.8 3.559 3.698 3.559 3.698 ls_scf_post 1 4.0 0.000 0.000 3.626 3.633 make_images_sizes 222 9.7 0.000 0.000 1.152 3.492 mp_alltoall_i44 222 10.7 1.152 3.492 1.152 3.492 ls_scf_store_result 1 5.0 0.000 0.000 3.406 3.453 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.784 3.042 dbcsr_finalize 304 7.8 0.062 0.078 2.200 2.286 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.217 2.218 rebuild_ks_matrix 3 7.3 0.000 0.000 2.185 2.185 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.185 2.185 dbcsr_merge_all 275 8.9 0.473 0.524 2.048 2.091 dbcsr_data_release 10123 10.8 1.321 2.033 1.321 2.033 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=99.281000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3635.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/48320402b71dba81277f3b122ba0d11842227d22_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 8.741274E+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 0.098 0.168 108.144 108.144 qs_energies 1 2.0 0.000 0.000 106.638 106.653 ls_scf 1 3.0 0.000 0.000 103.704 103.718 dbcsr_multiply_generic 111 6.7 0.023 0.028 77.211 77.334 ls_scf_main 1 4.0 0.000 0.000 65.864 65.871 density_matrix_trs4 2 5.0 0.004 0.015 56.597 56.643 multiply_cannon 111 7.7 0.152 0.242 49.847 51.802 multiply_cannon_loop 111 8.7 0.067 0.070 46.409 47.364 ls_scf_init_scf 1 4.0 0.000 0.000 34.175 34.176 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.489 32.505 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.641 29.652 mp_waitall_1 4527 11.1 22.147 25.906 22.147 25.906 make_m2s 222 7.7 0.005 0.005 23.784 24.749 make_images 222 8.7 4.588 4.966 23.678 24.640 multiply_cannon_multrec 444 9.7 17.801 18.472 22.419 23.072 make_images_data 222 9.7 0.003 0.003 13.031 15.409 hybrid_alltoall_any 227 10.6 1.660 3.622 12.827 15.391 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.512 10.897 multiply_cannon_sync_h2d 444 9.7 8.844 8.903 8.844 8.903 arnoldi_extremal 4 6.8 0.000 0.000 7.249 7.255 arnoldi_normal_ev 4 7.8 0.004 0.010 7.249 7.255 build_subspace 16 8.4 0.026 0.036 6.701 6.716 dbcsr_matrix_vector_mult 304 9.0 0.017 0.033 5.391 5.523 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.257 5.346 dbcsr_matrix_vector_mult_local 304 10.0 4.972 5.267 4.975 5.270 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.987 5.239 dbcsr_mm_accdrv_process 1814 10.4 0.234 0.322 4.434 4.553 dbcsr_mm_accdrv_process_sort 1814 11.4 4.130 4.266 4.130 4.266 ls_scf_post 1 4.0 0.000 0.000 3.665 3.679 mp_allgather_i34 111 8.7 1.093 3.512 1.093 3.512 make_images_sizes 222 9.7 0.000 0.000 1.423 3.459 mp_alltoall_i44 222 10.7 1.422 3.459 1.422 3.459 ls_scf_store_result 1 5.0 0.000 0.000 3.404 3.415 calculate_norms 792 9.8 3.245 3.280 3.245 3.280 dbcsr_finalize 304 7.8 0.082 0.089 3.093 3.168 dbcsr_merge_all 275 8.9 0.891 0.916 2.878 2.947 qs_energies_init_hamiltonians 1 3.0 0.001 0.002 2.903 2.903 dbcsr_complete_redistribute 5 7.6 1.443 1.489 2.779 2.882 dbcsr_data_release 12724 10.6 2.324 2.868 2.324 2.868 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.614 2.616 rebuild_ks_matrix 3 7.3 0.000 0.000 2.546 2.548 qs_ks_build_kohn_sham_matrix 3 8.3 0.006 0.008 2.546 2.548 matrix_ls_to_qs 2 6.0 0.000 0.000 2.413 2.516 dbcsr_sort_data 325 11.1 2.448 2.504 2.448 2.504 dbcsr_new_transposed 4 7.5 0.257 0.280 2.288 2.300 dbcsr_frobenius_norm 74 6.6 2.056 2.132 2.194 2.233 dbcsr_add_d 103 6.2 0.000 0.000 2.136 2.210 dbcsr_add_anytype 103 7.2 0.860 0.892 2.136 2.209 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=108.144000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6854.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 48320402b71dba81277f3b122ba0d11842227d22 Summary: empty Status: OK