=== 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: d649f8f97645802a3cd6acffbbd85227349dba89 ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, Cray-FFTW 3.3.8.10, # COSMA 2.6.2, ELPA 2022.11.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.1.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.1, # SIRIUS 7.4.3, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Author: Matthias Krack (03.02.2023) # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.1.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.1 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta_prefixed_scalapack.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d649f8f97645802a3cd6acffbbd85227349dba89_performance_tests/01 job id: 45019247 --- 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/d649f8f97645802a3cd6acffbbd85227349dba89_performance_tests/02 job id: 45019248 --- 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/d649f8f97645802a3cd6acffbbd85227349dba89_performance_tests/03 job id: 45019249 --- 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/d649f8f97645802a3cd6acffbbd85227349dba89_performance_tests/04 job id: 45019252 --- 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/d649f8f97645802a3cd6acffbbd85227349dba89_performance_tests/05 job id: 45019253 --- 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/d649f8f97645802a3cd6acffbbd85227349dba89_performance_tests/06 job id: 45019254 --- 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/d649f8f97645802a3cd6acffbbd85227349dba89_performance_tests/07 job id: 45019255 --- 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/d649f8f97645802a3cd6acffbbd85227349dba89_performance_tests/08 job id: 45019261 --- 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/d649f8f97645802a3cd6acffbbd85227349dba89_performance_tests/09 job id: 45019266 --- 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/d649f8f97645802a3cd6acffbbd85227349dba89_performance_tests/10 job id: 45019269 --- 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/d649f8f97645802a3cd6acffbbd85227349dba89_performance_tests/11 job id: 45019270 --- 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/d649f8f97645802a3cd6acffbbd85227349dba89_performance_tests/12 job id: 45019271 --- 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/d649f8f97645802a3cd6acffbbd85227349dba89_performance_tests/13 job id: 45019272 --- 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/d649f8f97645802a3cd6acffbbd85227349dba89_performance_tests/14 job id: 45019273 --- 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/d649f8f97645802a3cd6acffbbd85227349dba89_performance_tests/15 job id: 45019274 --- 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/d649f8f97645802a3cd6acffbbd85227349dba89_performance_tests/16 job id: 45019275 --- 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/d649f8f97645802a3cd6acffbbd85227349dba89_performance_tests/17 job id: 45019276 --- 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/d649f8f97645802a3cd6acffbbd85227349dba89_performance_tests/18 job id: 45019277 --- 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/d649f8f97645802a3cd6acffbbd85227349dba89_performance_tests/19 job id: 45019278 --- 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/d649f8f97645802a3cd6acffbbd85227349dba89_performance_tests/20 job id: 45019279 --- 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/d649f8f97645802a3cd6acffbbd85227349dba89_performance_tests/21 job id: 45019280 --- 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/d649f8f97645802a3cd6acffbbd85227349dba89_performance_tests/22 job id: 45019281 --- 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/d649f8f97645802a3cd6acffbbd85227349dba89_performance_tests/23 job id: 45019282 --- 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/d649f8f97645802a3cd6acffbbd85227349dba89_performance_tests/24 job id: 45019283 --- 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/d649f8f97645802a3cd6acffbbd85227349dba89_performance_tests/25 job id: 45019284 --- 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/d649f8f97645802a3cd6acffbbd85227349dba89_performance_tests/26 job id: 45019285 --- 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/d649f8f97645802a3cd6acffbbd85227349dba89_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 0.024 0.035 135.887 135.888 farming_run 1 2.0 135.377 135.379 135.851 135.855 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.456558E+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 229 1108280. MP_Allreduce 485 2282278. MP_Sync 27 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 8 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.025 116.517 116.517 qs_energies 1 2.0 0.000 0.000 116.300 116.300 mp2_main 1 3.0 0.000 0.000 114.181 114.182 mp2_gpw_main 1 4.0 0.019 0.026 113.242 113.242 mp2_ri_gpw_compute_in 1 5.0 0.172 0.173 94.024 94.523 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 55.406 55.907 mp2_eri_3c_integrate_gpw 272 7.0 0.152 0.166 41.716 47.111 get_2c_integrals 1 6.0 0.000 0.001 37.524 38.444 integrate_v_rspace 273 8.0 0.435 0.450 25.129 30.303 pw_transfer 6555 10.6 0.371 0.380 27.449 28.026 fft_wrap_pw1pw2 5465 11.4 0.045 0.046 26.104 26.724 grid_integrate_task_list 273 9.0 20.937 26.589 20.937 26.589 fft_wrap_pw1pw2_100 2178 12.4 1.189 1.426 23.606 24.228 compute_2c_integrals 1 7.0 0.003 0.003 19.647 19.648 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.909 19.340 mp2_eri_2c_integrate_gpw 1 9.0 2.381 2.437 18.906 19.337 rpa_ri_compute_en 1 5.0 0.011 0.016 19.111 19.302 cp_fm_cholesky_decompose 12 8.2 17.902 18.825 17.902 18.825 cholesky_decomp 1 7.0 0.000 0.000 16.720 17.642 fft3d_s 5443 13.4 16.183 16.716 16.205 16.737 ao_to_mo_and_store_B_mult_1 272 7.0 10.851 15.559 10.851 15.559 calculate_wavefunction 272 8.0 5.382 5.536 12.509 13.083 rpa_num_int 1 6.0 0.001 0.009 10.729 10.729 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.637 10.674 calc_mat_Q 8 8.0 0.000 0.000 9.414 9.522 contract_S_to_Q 8 9.0 0.000 0.000 8.836 8.943 calc_potential_gpw 544 9.5 0.005 0.006 8.282 8.682 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.236 8.539 parallel_gemm_fm 14 9.1 0.000 0.000 8.417 8.499 parallel_gemm_fm_cosma 14 10.1 8.417 8.499 8.417 8.499 potential_pw2rs 545 10.0 0.108 0.111 7.690 8.259 create_integ_mat 1 6.0 0.006 0.008 7.773 7.773 collocate_single_gaussian 272 10.0 0.039 0.042 7.471 7.702 array2fm 1 7.0 0.000 0.000 6.699 7.141 pw_scatter_s 2720 13.7 4.442 4.696 4.442 4.696 pw_gather_s 2722 13.2 3.861 4.192 3.861 4.192 array2fm_buffer_send 1 8.0 2.878 3.336 2.878 3.336 pw_poisson_solve 545 10.5 1.119 1.158 2.172 2.429 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=113.242250, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2728.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d649f8f97645802a3cd6acffbbd85227349dba89_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 0.028 0.038 397.695 397.697 farming_run 1 2.0 396.946 396.950 397.658 397.662 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827128576 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788821 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.225921E+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 704 407793. MP_Allreduce 1821 23730. MP_Sync 38 MP_Alltoall 77 9998443. 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 0.014 0.046 210.117 210.118 qs_energies 1 2.0 0.000 0.000 209.880 209.888 scf_env_do_scf 1 3.0 0.000 0.000 106.454 106.455 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.466 105.473 rebuild_ks_matrix 4 6.0 0.000 0.000 105.465 105.472 qs_ks_build_kohn_sham_matrix 4 7.0 0.057 0.064 105.465 105.472 hfx_ks_matrix 4 8.0 0.001 0.001 105.071 105.076 integrate_four_center 4 9.0 0.144 0.459 105.070 105.075 mp2_main 1 3.0 0.000 0.000 103.116 103.123 mp2_gpw_main 1 4.0 0.032 0.046 102.185 102.193 integrate_four_center_main 4 10.0 0.095 0.601 96.814 99.668 integrate_four_center_bin 265 11.0 96.719 99.646 96.719 99.646 init_scf_loop 1 4.0 0.000 0.000 92.006 92.007 mp2_ri_gpw_compute_in 1 5.0 0.064 0.064 74.982 76.182 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.535 55.735 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.165 42.182 47.341 integrate_v_rspace 95 8.0 0.397 0.561 28.551 33.721 pw_transfer 2240 10.6 0.142 0.160 29.984 30.345 fft_wrap_pw1pw2 1868 11.4 0.018 0.020 28.988 29.359 grid_integrate_task_list 95 9.0 23.826 29.111 23.826 29.111 mp2_ri_gpw_compute_en 1 5.0 0.055 0.063 27.057 28.819 ao_to_mo_and_store_B_mult_1 91 7.0 10.668 28.347 10.668 28.347 fft_wrap_pw1pw2_100 730 12.4 1.287 1.466 26.697 27.112 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.832 1.907 25.205 25.217 get_2c_integrals 1 6.0 0.000 0.000 20.351 20.384 compute_2c_integrals 1 7.0 0.002 0.003 19.325 19.329 compute_2c_integrals_loop_lm 1 8.0 0.001 0.002 18.900 19.192 mp2_eri_2c_integrate_gpw 1 9.0 1.742 1.853 18.899 19.191 fft3d_s 1823 13.4 18.461 18.895 18.474 18.908 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.446 14.446 calculate_wavefunction 91 8.0 2.009 2.038 9.747 9.968 mp2_ri_gpw_compute_en_expansio 172 7.0 0.555 0.577 8.808 9.688 potential_pw2rs 186 10.0 0.034 0.035 8.675 9.200 local_gemm 172 8.0 8.253 9.121 8.253 9.121 mp2_ri_gpw_compute_en_comm 22 7.0 0.489 0.514 8.178 8.727 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.259 8.657 collocate_single_gaussian 91 10.0 0.016 0.018 7.906 8.165 calc_potential_gpw 182 9.5 0.002 0.002 7.912 8.145 mp_sendrecv_dm3 2068 8.0 6.239 6.815 6.239 6.815 mp2_ri_gpw_compute_en_ener 172 7.0 6.342 6.424 6.342 6.424 mp_sync 38 10.4 2.905 5.577 2.905 5.577 pw_gather_s 912 13.2 4.920 5.381 4.920 5.381 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=102.185301, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1511.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d649f8f97645802a3cd6acffbbd85227349dba89_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 451.559424E+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 578578. 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 0.021 0.061 56.948 56.953 qs_mol_dyn_low 1 2.0 0.005 0.014 56.254 56.264 qs_forces 11 3.9 0.005 0.026 56.089 56.091 qs_energies 11 4.9 0.005 0.024 54.507 54.523 scf_env_do_scf 11 5.9 0.000 0.001 47.816 47.816 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 45.459 45.459 dbcsr_multiply_generic 2286 12.5 0.094 0.097 35.025 35.509 qs_scf_new_mos 108 7.5 0.000 0.001 34.120 34.435 qs_scf_loop_do_ot 108 8.5 0.000 0.001 34.119 34.434 ot_scf_mini 108 9.5 0.003 0.011 32.380 32.578 multiply_cannon 2286 13.5 0.183 0.187 26.604 28.329 velocity_verlet 10 3.0 0.002 0.006 28.163 28.165 multiply_cannon_loop 2286 14.5 1.498 1.571 25.678 27.407 ot_mini 108 10.5 0.001 0.001 20.447 20.691 qs_ot_get_derivative 108 11.5 0.001 0.002 17.406 17.584 mp_waitall_1 245248 16.5 9.307 15.542 9.307 15.542 multiply_cannon_metrocomm3 54864 15.5 0.068 0.072 6.222 13.503 multiply_cannon_multrec 54864 15.5 4.165 6.359 7.681 11.111 rebuild_ks_matrix 119 8.3 0.000 0.000 8.640 8.778 qs_ks_build_kohn_sham_matrix 119 9.3 0.011 0.016 8.639 8.777 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.660 7.779 mp_sum_l 7207 12.9 5.902 7.770 5.902 7.770 multiply_cannon_sync_h2d 54864 15.5 5.810 7.408 5.810 7.408 qs_ot_get_p 119 10.4 0.001 0.001 7.067 7.390 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 6.032 6.494 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.721 5.882 init_scf_run 11 5.9 0.000 0.001 5.342 5.342 scf_env_initial_rho_setup 11 6.9 0.002 0.014 5.341 5.342 sum_up_and_integrate 119 10.3 0.012 0.016 5.052 5.061 integrate_v_rspace 119 11.3 0.002 0.002 5.039 5.050 qs_rho_update_rho_low 119 7.7 0.002 0.009 4.901 5.021 calculate_rho_elec 119 8.7 0.011 0.018 4.899 5.021 dbcsr_mm_accdrv_process 76910 16.1 1.192 1.853 3.438 4.811 qs_ot_p2m_diag 50 11.0 0.004 0.008 3.910 3.965 multiply_cannon_metrocomm1 54864 15.5 0.053 0.059 2.013 3.703 rs_pw_transfer 974 11.9 0.011 0.013 3.355 3.479 calculate_dm_sparse 119 9.5 0.000 0.000 3.012 3.157 apply_preconditioner_dbcsr 119 12.6 0.000 0.001 2.928 3.123 apply_single 119 13.6 0.000 0.000 2.928 3.123 cp_dbcsr_syevd 50 12.0 0.002 0.003 3.038 3.039 calculate_first_density_matrix 1 7.0 0.006 0.015 3.002 3.008 density_rs2pw 119 9.7 0.004 0.005 2.838 2.956 jit_kernel_multiply 13 15.8 2.182 2.886 2.182 2.886 ot_diis_step 108 11.5 0.006 0.009 2.731 2.732 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.702 2.702 cp_fm_redistribute_end 50 14.0 2.449 2.668 2.455 2.670 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.586 2.655 cp_fm_diag_elpa_base 50 14.0 0.212 2.567 0.213 2.576 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.542 2.544 pw_transfer 1439 11.6 0.053 0.061 2.314 2.408 potential_pw2rs 119 12.3 0.004 0.004 2.322 2.360 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 2.237 2.336 init_scf_loop 11 6.9 0.000 0.000 2.329 2.329 make_m2s 4572 13.5 0.054 0.056 2.235 2.303 wfi_extrapolate 11 7.9 0.001 0.003 2.251 2.252 mp_sum_d 4127 12.0 1.551 2.226 1.551 2.226 make_images 4572 14.5 0.132 0.138 2.152 2.218 acc_transpose_blocks 54864 15.5 0.233 0.252 1.755 2.183 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.109 2.156 fft3d_ps 1201 14.6 0.369 0.474 2.009 2.104 grid_integrate_task_list 119 12.3 2.001 2.104 2.001 2.104 mp_alltoall_d11v 2130 13.8 1.615 1.829 1.615 1.829 fft_wrap_pw1pw2_140 487 13.2 0.081 0.094 1.725 1.824 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.451 1.472 mp_waitany 12084 13.8 1.291 1.450 1.291 1.450 grid_collocate_task_list 119 9.7 1.289 1.340 1.289 1.340 dbcsr_dot_sd 1205 11.9 0.051 0.062 0.875 1.322 make_images_sizes 4572 15.5 0.004 0.004 0.938 1.192 mp_alltoall_i44 4572 16.5 0.934 1.188 0.934 1.188 prepare_preconditioner 11 7.9 0.000 0.000 1.108 1.141 make_preconditioner 11 8.9 0.000 0.000 1.108 1.141 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=56.953000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=430.181818, yerr=1.113404 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d649f8f97645802a3cd6acffbbd85227349dba89_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 488.325120E+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 1464561. 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 0.078 0.135 42.262 42.267 qs_mol_dyn_low 1 2.0 0.003 0.003 41.501 41.509 qs_forces 11 3.9 0.002 0.002 41.436 41.437 qs_energies 11 4.9 0.014 0.052 39.658 39.674 scf_env_do_scf 11 5.9 0.004 0.014 33.782 33.786 scf_env_do_scf_inner_loop 108 6.5 0.013 0.028 30.829 30.833 dbcsr_multiply_generic 2286 12.5 0.100 0.102 22.175 22.539 qs_scf_new_mos 108 7.5 0.001 0.001 20.889 21.133 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.888 21.132 ot_scf_mini 108 9.5 0.003 0.003 19.998 20.170 velocity_verlet 10 3.0 0.001 0.001 19.427 19.429 multiply_cannon 2286 13.5 0.207 0.217 16.843 18.569 multiply_cannon_loop 2286 14.5 0.896 0.965 15.651 16.963 ot_mini 108 10.5 0.001 0.001 12.263 12.497 mp_waitall_1 200699 16.5 6.184 11.405 6.184 11.405 qs_ot_get_derivative 108 11.5 0.001 0.001 9.764 9.938 multiply_cannon_metrocomm3 27432 15.5 0.068 0.072 4.255 9.773 multiply_cannon_multrec 27432 15.5 1.994 4.645 6.108 9.017 rebuild_ks_matrix 119 8.3 0.000 0.000 7.724 7.858 qs_ks_build_kohn_sham_matrix 119 9.3 0.032 0.069 7.723 7.858 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.834 6.959 dbcsr_mm_accdrv_process 47894 16.0 3.122 5.355 4.045 5.941 qs_ot_get_p 119 10.4 0.001 0.001 4.792 5.024 sum_up_and_integrate 119 10.3 0.024 0.026 4.559 4.563 integrate_v_rspace 119 11.3 0.002 0.002 4.535 4.542 init_scf_run 11 5.9 0.000 0.001 4.519 4.520 scf_env_initial_rho_setup 11 6.9 0.000 0.002 4.519 4.519 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.679 4.517 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.320 4.366 calculate_rho_elec 119 8.7 0.021 0.024 4.320 4.365 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.069 4.136 apply_single 119 13.6 0.000 0.000 3.068 4.136 mp_sum_l 7207 12.9 2.169 4.036 2.169 4.036 rs_pw_transfer 974 11.9 0.010 0.011 3.215 3.651 qs_ot_p2m_diag 50 11.0 0.008 0.012 3.205 3.224 density_rs2pw 119 9.7 0.004 0.004 2.613 3.050 make_m2s 4572 13.5 0.052 0.053 2.729 2.955 init_scf_loop 11 6.9 0.012 0.050 2.918 2.922 make_images 4572 14.5 0.200 0.237 2.640 2.865 multiply_cannon_sync_h2d 27432 15.5 2.186 2.825 2.186 2.825 calculate_first_density_matrix 1 7.0 0.006 0.023 2.795 2.810 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.750 2.751 ot_diis_step 108 11.5 0.011 0.011 2.435 2.436 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.330 2.331 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.232 2.320 potential_pw2rs 119 12.3 0.006 0.006 2.277 2.297 cp_fm_redistribute_end 50 14.0 1.921 2.296 1.926 2.297 pw_transfer 1439 11.6 0.064 0.072 2.216 2.263 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.244 2.246 cp_fm_diag_elpa_base 50 14.0 0.355 2.178 0.368 2.228 calculate_dm_sparse 119 9.5 0.000 0.001 2.154 2.226 jit_kernel_multiply 10 16.2 0.870 2.203 0.870 2.203 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.124 2.176 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.886 1.925 grid_integrate_task_list 119 12.3 1.822 1.911 1.822 1.911 fft3d_ps 1201 14.6 0.511 0.563 1.826 1.873 prepare_preconditioner 11 7.9 0.000 0.001 1.805 1.840 make_preconditioner 11 8.9 0.007 0.029 1.805 1.840 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.688 1.764 make_images_data 4572 15.5 0.044 0.050 1.206 1.637 fft_wrap_pw1pw2_140 487 13.2 0.078 0.085 1.577 1.628 wfi_extrapolate 11 7.9 0.001 0.001 1.600 1.600 hybrid_alltoall_any 4725 16.4 0.050 0.112 1.056 1.556 acc_transpose_blocks 27432 15.5 0.111 0.116 1.228 1.551 mp_alltoall_d11v 2130 13.8 1.338 1.548 1.338 1.548 mp_allgather_i34 2286 14.5 0.623 1.402 0.623 1.402 grid_collocate_task_list 119 9.7 1.237 1.396 1.237 1.396 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.382 1.390 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.212 1.258 mp_sum_d 4127 12.0 0.677 1.159 0.677 1.159 mp_waitany 5720 13.7 0.583 1.054 0.583 1.054 rs_pw_transfer_RS2PW_140 130 11.5 0.144 0.162 0.608 1.034 make_images_sizes 4572 15.5 0.005 0.005 0.715 1.003 mp_alltoall_i44 4572 16.5 0.711 0.998 0.711 0.998 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 0.977 0.978 mp_alltoall_z22v 1201 16.6 0.852 0.978 0.852 0.978 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.959 0.973 acc_transpose_blocks_kernels 27432 16.5 0.182 0.270 0.694 0.928 rs_pw_transfer_PW2RS_50 119 14.3 0.587 0.606 0.812 0.900 dbcsr_complete_redistribute 329 12.2 0.136 0.205 0.805 0.878 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=42.267000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=464.545455, yerr=1.157084 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d649f8f97645802a3cd6acffbbd85227349dba89_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 519.401472E+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 658590. 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 0.051 0.077 34.510 34.511 qs_mol_dyn_low 1 2.0 0.003 0.005 34.147 34.154 qs_forces 11 3.9 0.002 0.006 34.075 34.075 qs_energies 11 4.9 0.002 0.009 32.406 32.409 scf_env_do_scf 11 5.9 0.001 0.001 27.344 27.344 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 24.479 24.481 dbcsr_multiply_generic 2286 12.5 0.094 0.096 16.927 17.011 velocity_verlet 10 3.0 0.001 0.001 16.387 16.388 qs_scf_new_mos 108 7.5 0.001 0.001 15.601 15.617 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.601 15.616 ot_scf_mini 108 9.5 0.002 0.003 14.858 14.868 multiply_cannon 2286 13.5 0.195 0.201 13.330 14.087 multiply_cannon_loop 2286 14.5 0.633 0.668 12.466 13.238 ot_mini 108 10.5 0.001 0.001 9.036 9.053 qs_ot_get_derivative 108 11.5 0.001 0.001 7.505 7.514 multiply_cannon_multrec 18288 15.5 1.926 2.828 6.942 7.193 rebuild_ks_matrix 119 8.3 0.000 0.000 6.810 6.831 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.016 6.810 6.830 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.986 6.004 dbcsr_mm_accdrv_process 38222 16.0 3.963 5.552 4.932 5.826 sum_up_and_integrate 119 10.3 0.030 0.031 4.312 4.320 integrate_v_rspace 119 11.3 0.002 0.003 4.282 4.292 mp_waitall_1 158411 16.6 2.905 4.037 2.905 4.037 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.856 3.863 calculate_rho_elec 119 8.7 0.030 0.031 3.855 3.862 init_scf_run 11 5.9 0.000 0.001 3.822 3.822 scf_env_initial_rho_setup 11 6.9 0.000 0.001 3.822 3.822 qs_ot_get_p 119 10.4 0.001 0.001 3.606 3.636 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.836 3.438 rs_pw_transfer 974 11.9 0.009 0.010 2.755 2.972 init_scf_loop 11 6.9 0.001 0.006 2.842 2.843 density_rs2pw 119 9.7 0.004 0.004 2.353 2.599 calculate_first_density_matrix 1 7.0 0.001 0.005 2.459 2.459 multiply_cannon_metrocomm3 18288 15.5 0.044 0.046 1.416 2.420 qs_ot_p2m_diag 50 11.0 0.012 0.012 2.399 2.404 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.037 2.368 apply_single 119 13.6 0.000 0.000 2.037 2.368 pw_transfer 1439 11.6 0.065 0.072 2.200 2.228 make_m2s 4572 13.5 0.044 0.045 2.044 2.192 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.107 2.133 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.122 2.123 make_images 4572 14.5 0.190 0.203 1.959 2.106 potential_pw2rs 119 12.3 0.007 0.008 2.057 2.076 jit_kernel_multiply 10 16.1 0.917 2.051 0.917 2.051 prepare_preconditioner 11 7.9 0.000 0.000 1.976 1.979 make_preconditioner 11 8.9 0.001 0.006 1.976 1.979 calculate_dm_sparse 119 9.5 0.000 0.000 1.909 1.921 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.822 1.908 grid_integrate_task_list 119 12.3 1.805 1.884 1.805 1.884 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.858 1.860 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.832 1.834 fft3d_ps 1201 14.6 0.522 0.541 1.788 1.816 cp_fm_redistribute_end 50 14.0 1.360 1.801 1.362 1.802 cp_fm_diag_elpa_base 50 14.0 0.422 1.710 0.437 1.756 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.672 1.680 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.663 1.671 fft_wrap_pw1pw2_140 487 13.2 0.088 0.092 1.623 1.651 multiply_cannon_sync_h2d 18288 15.5 1.399 1.550 1.399 1.550 mp_sum_l 7207 12.9 1.190 1.527 1.190 1.527 ot_diis_step 108 11.5 0.011 0.011 1.505 1.505 grid_collocate_task_list 119 9.7 1.211 1.364 1.211 1.364 wfi_extrapolate 11 7.9 0.001 0.001 1.312 1.312 acc_transpose_blocks 18288 15.5 0.077 0.079 1.280 1.308 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.264 1.269 make_images_data 4572 15.5 0.044 0.048 0.901 1.092 hybrid_alltoall_any 4725 16.4 0.055 0.113 0.782 0.992 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 0.988 0.989 mp_alltoall_z22v 1201 16.6 0.849 0.965 0.849 0.965 mp_alltoall_d11v 2130 13.8 0.787 0.936 0.787 0.936 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.911 0.932 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.900 0.901 cp_fm_cholesky_invert 11 10.9 0.878 0.884 0.878 0.884 acc_transpose_blocks_kernels 18288 16.5 0.210 0.218 0.838 0.857 mp_waitany 9880 13.7 0.567 0.834 0.567 0.834 rs_pw_transfer_RS2PW_140 130 11.5 0.121 0.138 0.552 0.796 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.661 0.724 mp_allgather_i34 2286 14.5 0.329 0.715 0.329 0.715 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=34.511000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=494.909091, yerr=1.311110 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d649f8f97645802a3cd6acffbbd85227349dba89_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 546.828288E+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 0.014 0.033 37.547 37.548 qs_mol_dyn_low 1 2.0 0.003 0.003 37.356 37.365 qs_forces 11 3.9 0.002 0.002 37.292 37.292 qs_energies 11 4.9 0.001 0.001 35.478 35.483 scf_env_do_scf 11 5.9 0.000 0.001 30.236 30.238 scf_env_do_scf_inner_loop 108 6.5 0.005 0.011 26.454 26.454 velocity_verlet 10 3.0 0.001 0.001 19.105 19.107 dbcsr_multiply_generic 2286 12.5 0.100 0.102 18.752 18.857 qs_scf_new_mos 108 7.5 0.001 0.001 17.331 17.378 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.330 17.377 ot_scf_mini 108 9.5 0.002 0.003 16.359 16.412 multiply_cannon 2286 13.5 0.230 0.271 14.582 15.138 multiply_cannon_loop 2286 14.5 0.937 0.967 13.515 13.919 ot_mini 108 10.5 0.001 0.001 9.746 9.815 multiply_cannon_multrec 27432 15.5 2.327 3.014 8.664 9.018 qs_ot_get_derivative 108 11.5 0.001 0.001 7.894 7.947 dbcsr_mm_accdrv_process 47916 15.9 5.200 6.999 6.245 7.401 rebuild_ks_matrix 119 8.3 0.000 0.000 6.984 7.044 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 6.983 7.044 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.182 6.229 sum_up_and_integrate 119 10.3 0.035 0.038 4.135 4.142 integrate_v_rspace 119 11.3 0.002 0.003 4.100 4.109 qs_ot_get_p 119 10.4 0.001 0.001 3.917 3.996 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.908 3.939 calculate_rho_elec 119 8.7 0.040 0.046 3.907 3.939 init_scf_run 11 5.9 0.000 0.001 3.853 3.853 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.853 3.853 init_scf_loop 11 6.9 0.000 0.000 3.758 3.759 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.850 3.349 prepare_preconditioner 11 7.9 0.000 0.000 2.879 2.888 make_preconditioner 11 8.9 0.000 0.000 2.879 2.888 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.468 2.807 make_m2s 4572 13.5 0.054 0.055 2.527 2.689 rs_pw_transfer 974 11.9 0.009 0.009 2.424 2.614 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.168 2.592 apply_single 119 13.6 0.000 0.000 2.168 2.591 make_images 4572 14.5 0.271 0.334 2.420 2.580 density_rs2pw 119 9.7 0.004 0.004 2.306 2.488 mp_waitall_1 137007 16.6 2.097 2.471 2.097 2.471 qs_ot_p2m_diag 50 11.0 0.015 0.022 2.437 2.447 pw_transfer 1439 11.6 0.066 0.072 2.309 2.356 calculate_first_density_matrix 1 7.0 0.001 0.001 2.353 2.355 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.216 2.268 calculate_dm_sparse 119 9.5 0.000 0.000 2.112 2.164 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.122 2.148 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.088 2.089 fft3d_ps 1201 14.6 0.560 0.612 1.889 1.933 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.911 1.912 grid_integrate_task_list 119 12.3 1.827 1.891 1.827 1.891 potential_pw2rs 119 12.3 0.009 0.009 1.875 1.882 ot_diis_step 108 11.5 0.012 0.012 1.811 1.812 fft_wrap_pw1pw2_140 487 13.2 0.088 0.095 1.756 1.812 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.763 1.766 jit_kernel_multiply 9 15.8 0.983 1.733 0.983 1.733 cp_fm_redistribute_end 50 14.0 1.165 1.728 1.167 1.729 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.694 1.708 mp_sum_l 7207 12.9 1.218 1.683 1.218 1.683 cp_fm_diag_elpa_base 50 14.0 0.536 1.639 0.556 1.678 acc_transpose_blocks 27432 15.5 0.114 0.117 1.495 1.524 wfi_extrapolate 11 7.9 0.001 0.001 1.447 1.447 grid_collocate_task_list 119 9.7 1.221 1.350 1.221 1.350 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.263 1.274 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.170 1.185 dbcsr_complete_redistribute 329 12.2 0.122 0.150 0.894 1.172 make_images_data 4572 15.5 0.045 0.048 0.999 1.165 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.720 1.162 cp_fm_upper_to_full 72 13.5 0.812 1.138 0.812 1.138 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.101 1.101 multiply_cannon_sync_h2d 27432 15.5 0.974 1.049 0.974 1.049 hybrid_alltoall_any 4725 16.4 0.062 0.151 0.864 1.047 mp_alltoall_d11v 2130 13.8 0.859 0.992 0.859 0.992 cp_fm_cholesky_invert 11 10.9 0.973 0.977 0.973 0.977 mp_alltoall_z22v 1201 16.6 0.914 0.953 0.914 0.953 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.894 0.904 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.629 0.902 acc_transpose_blocks_kernels 27432 16.5 0.269 0.277 0.865 0.884 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.792 0.872 mp_alltoall_i22 627 13.8 0.456 0.756 0.456 0.756 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=37.548000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=517.636364, yerr=4.395715 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d649f8f97645802a3cd6acffbbd85227349dba89_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 607.866880E+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 0.021 0.036 29.934 29.935 qs_mol_dyn_low 1 2.0 0.003 0.003 29.579 29.589 qs_forces 11 3.9 0.002 0.002 29.518 29.518 qs_energies 11 4.9 0.001 0.001 27.771 27.773 scf_env_do_scf 11 5.9 0.000 0.001 22.814 22.814 scf_env_do_scf_inner_loop 108 6.5 0.007 0.013 20.109 20.110 velocity_verlet 10 3.0 0.001 0.001 15.258 15.261 dbcsr_multiply_generic 2286 12.5 0.092 0.095 12.569 12.652 qs_scf_new_mos 108 7.5 0.001 0.001 11.566 11.588 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.566 11.587 ot_scf_mini 108 9.5 0.002 0.002 10.859 10.883 multiply_cannon 2286 13.5 0.230 0.240 9.742 10.272 multiply_cannon_loop 2286 14.5 0.328 0.339 8.809 9.008 rebuild_ks_matrix 119 8.3 0.000 0.000 6.280 6.304 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 6.280 6.303 multiply_cannon_multrec 9144 15.5 1.583 1.827 5.916 6.164 ot_mini 108 10.5 0.001 0.001 5.903 5.936 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.605 5.626 qs_ot_get_derivative 108 11.5 0.001 0.001 4.591 4.616 dbcsr_mm_accdrv_process 12550 15.8 3.200 3.923 4.233 4.302 sum_up_and_integrate 119 10.3 0.038 0.042 3.904 3.908 integrate_v_rspace 119 11.3 0.002 0.003 3.866 3.870 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.847 3.854 calculate_rho_elec 119 8.7 0.060 0.061 3.846 3.853 init_scf_run 11 5.9 0.000 0.001 3.497 3.497 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.497 3.497 qs_ot_get_p 119 10.4 0.001 0.001 3.136 3.178 init_scf_loop 11 6.9 0.000 0.000 2.681 2.683 pw_transfer 1439 11.6 0.066 0.069 2.334 2.345 density_rs2pw 119 9.7 0.004 0.004 2.143 2.287 calculate_first_density_matrix 1 7.0 0.001 0.001 2.271 2.271 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.240 2.253 mp_waitall_1 115863 16.7 1.648 2.215 1.648 2.215 rs_pw_transfer 974 11.9 0.008 0.008 1.956 2.120 make_m2s 4572 13.5 0.034 0.035 1.946 2.112 qs_ot_p2m_diag 50 11.0 0.022 0.023 2.087 2.090 make_images 4572 14.5 0.269 0.300 1.857 2.023 jit_kernel_multiply 10 15.8 0.995 1.933 0.995 1.933 grid_integrate_task_list 119 12.3 1.851 1.931 1.851 1.931 fft3d_ps 1201 14.6 0.563 0.574 1.899 1.911 prepare_preconditioner 11 7.9 0.000 0.000 1.903 1.908 make_preconditioner 11 8.9 0.000 0.000 1.903 1.908 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.878 1.878 calculate_dm_sparse 119 9.5 0.000 0.000 1.848 1.869 fft_wrap_pw1pw2_140 487 13.2 0.087 0.090 1.809 1.824 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.775 1.811 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.645 1.646 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.629 1.643 potential_pw2rs 119 12.3 0.010 0.011 1.609 1.616 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.556 1.557 cp_fm_redistribute_end 50 14.0 0.775 1.528 0.777 1.529 cp_fm_diag_elpa_base 50 14.0 0.707 1.450 0.750 1.506 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.459 1.467 grid_collocate_task_list 119 9.7 1.275 1.390 1.275 1.390 ot_diis_step 108 11.5 0.013 0.013 1.297 1.297 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.239 1.245 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.227 1.238 qs_energies_init_hamiltonians 11 5.9 0.002 0.002 1.231 1.232 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.164 1.192 apply_single 119 13.6 0.000 0.000 1.164 1.192 wfi_extrapolate 11 7.9 0.001 0.001 1.167 1.167 hybrid_alltoall_any 4725 16.4 0.062 0.174 0.863 1.129 make_images_data 4572 15.5 0.038 0.042 0.900 1.124 mp_alltoall_d11v 2130 13.8 0.877 0.982 0.877 0.982 cp_fm_cholesky_invert 11 10.9 0.975 0.978 0.975 0.978 acc_transpose_blocks 9144 15.5 0.039 0.039 0.943 0.947 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.867 0.917 mp_alltoall_z22v 1201 16.6 0.785 0.839 0.785 0.839 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.798 0.801 multiply_cannon_sync_h2d 9144 15.5 0.707 0.775 0.707 0.775 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.752 0.762 qs_env_update_s_mstruct 11 6.9 0.000 0.001 0.686 0.735 acc_transpose_blocks_kernels 9144 16.5 0.116 0.119 0.702 0.704 mp_allgather_i34 2286 14.5 0.266 0.700 0.266 0.700 yz_to_x 606 15.1 0.264 0.275 0.678 0.695 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.291 0.668 x_to_yz 595 16.2 0.276 0.285 0.647 0.654 dbcsr_complete_redistribute 329 12.2 0.164 0.176 0.607 0.636 mp_waitany 5200 13.7 0.465 0.626 0.465 0.626 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=29.935000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=572.181818, yerr=7.825578 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d649f8f97645802a3cd6acffbbd85227349dba89_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 762.114048E+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 0.017 0.039 43.869 43.870 qs_mol_dyn_low 1 2.0 0.003 0.003 43.649 43.658 qs_forces 11 3.9 0.002 0.002 43.574 43.574 qs_energies 11 4.9 0.001 0.002 41.557 41.560 scf_env_do_scf 11 5.9 0.001 0.001 35.718 35.718 scf_env_do_scf_inner_loop 108 6.5 0.003 0.008 27.355 27.356 velocity_verlet 10 3.0 0.001 0.001 24.740 24.748 dbcsr_multiply_generic 2286 12.5 0.101 0.102 18.041 18.269 qs_scf_new_mos 108 7.5 0.001 0.001 16.931 17.027 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.930 17.026 ot_scf_mini 108 9.5 0.002 0.002 15.808 15.915 multiply_cannon 2286 13.5 0.306 0.320 13.835 14.725 multiply_cannon_loop 2286 14.5 0.345 0.352 12.547 13.377 ot_mini 108 10.5 0.001 0.001 9.231 9.355 multiply_cannon_multrec 9144 15.5 3.337 4.653 8.625 8.720 init_scf_loop 11 6.9 0.000 0.000 8.329 8.331 rebuild_ks_matrix 119 8.3 0.000 0.000 7.675 7.829 qs_ks_build_kohn_sham_matrix 119 9.3 0.014 0.014 7.674 7.828 prepare_preconditioner 11 7.9 0.000 0.000 7.325 7.339 make_preconditioner 11 8.9 0.000 0.000 7.325 7.339 qs_ot_get_derivative 108 11.5 0.001 0.001 7.134 7.241 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.848 7.217 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.947 7.086 dbcsr_mm_accdrv_process 12550 15.8 4.183 5.717 5.168 6.458 cp_fm_upper_to_full 72 14.2 3.263 4.683 3.263 4.683 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.591 4.595 calculate_rho_elec 119 8.7 0.118 0.121 4.590 4.595 sum_up_and_integrate 119 10.3 0.065 0.067 4.280 4.285 integrate_v_rspace 119 11.3 0.003 0.003 4.214 4.219 qs_ot_get_p 119 10.4 0.001 0.001 3.762 3.906 init_scf_run 11 5.9 0.000 0.001 3.774 3.774 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.773 3.774 mp_waitall_1 94719 16.7 2.753 3.595 2.753 3.595 pw_transfer 1439 11.6 0.068 0.069 3.068 3.073 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.639 3.028 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.970 2.975 dbcsr_complete_redistribute 329 12.2 0.284 0.289 2.108 2.942 make_m2s 4572 13.5 0.037 0.038 2.532 2.701 fft3d_ps 1201 14.6 0.595 0.604 2.592 2.596 make_images 4572 14.5 0.357 0.389 2.410 2.579 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.735 2.561 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.206 2.532 apply_single 119 13.6 0.000 0.000 2.206 2.532 density_rs2pw 119 9.7 0.004 0.004 2.470 2.486 fft_wrap_pw1pw2_140 487 13.2 0.095 0.096 2.468 2.476 calculate_dm_sparse 119 9.5 0.000 0.000 2.313 2.333 qs_ot_p2m_diag 50 11.0 0.042 0.043 2.327 2.329 mp_alltoall_i22 627 13.8 1.482 2.319 1.482 2.319 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 1.463 2.292 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.472 2.289 calculate_first_density_matrix 1 7.0 0.001 0.001 2.224 2.226 grid_integrate_task_list 119 12.3 2.074 2.092 2.074 2.092 ot_diis_step 108 11.5 0.014 0.015 2.063 2.063 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.035 2.035 mp_sum_l 7207 12.9 1.130 1.911 1.130 1.911 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.866 1.867 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.794 1.845 jit_kernel_multiply 10 15.7 0.958 1.844 0.958 1.844 qs_energies_init_hamiltonians 11 5.9 0.002 0.004 1.778 1.779 cp_fm_cholesky_invert 11 10.9 1.722 1.726 1.722 1.726 rs_pw_transfer 974 11.9 0.009 0.009 1.680 1.699 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.688 1.689 cp_fm_diag_elpa_base 50 14.0 1.536 1.589 1.686 1.686 potential_pw2rs 119 12.3 0.014 0.014 1.616 1.619 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.567 1.581 grid_collocate_task_list 119 9.7 1.498 1.516 1.498 1.516 hybrid_alltoall_any 4725 16.4 0.088 0.147 1.259 1.506 make_images_data 4572 15.5 0.043 0.046 1.232 1.481 wfi_extrapolate 11 7.9 0.001 0.001 1.477 1.477 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.347 1.356 mp_alltoall_d11v 2130 13.8 1.225 1.242 1.225 1.242 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.192 1.217 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.111 1.127 mp_alltoall_z22v 1201 16.6 1.050 1.092 1.050 1.092 yz_to_x 606 15.1 0.458 0.470 1.070 1.082 multiply_cannon_sync_h2d 9144 15.5 1.042 1.046 1.042 1.046 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 0.985 1.024 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.942 0.956 qs_create_task_list 11 7.9 0.003 0.004 0.943 0.954 generate_qs_task_list 11 8.9 0.370 0.391 0.940 0.951 acc_transpose_blocks 9144 15.5 0.039 0.039 0.914 0.921 x_to_yz 595 16.2 0.478 0.488 0.916 0.919 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=43.870000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=708.545455, yerr=15.257568 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d649f8f97645802a3cd6acffbbd85227349dba89_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 501.334016E+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 1378927. 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 0.021 0.049 86.430 86.432 qs_mol_dyn_low 1 2.0 0.007 0.042 85.915 85.927 qs_forces 11 3.9 0.002 0.003 85.799 85.801 qs_energies 11 4.9 0.001 0.002 82.835 82.848 scf_env_do_scf 11 5.9 0.000 0.001 73.690 73.692 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 67.820 67.821 dbcsr_multiply_generic 2055 12.4 0.107 0.110 53.004 53.325 qs_scf_new_mos 99 7.5 0.000 0.001 49.458 49.560 qs_scf_loop_do_ot 99 8.5 0.000 0.001 49.458 49.560 ot_scf_mini 99 9.5 0.002 0.002 46.982 47.089 multiply_cannon 2055 13.4 0.177 0.182 42.954 44.044 velocity_verlet 10 3.0 0.001 0.001 43.576 43.577 multiply_cannon_loop 2055 14.4 1.553 1.603 41.888 43.033 ot_mini 99 10.5 0.001 0.001 28.301 28.380 qs_ot_get_derivative 99 11.5 0.001 0.001 21.397 21.504 multiply_cannon_multrec 49320 15.4 12.100 12.859 17.313 18.049 rebuild_ks_matrix 110 8.3 0.000 0.001 15.153 15.349 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.022 15.152 15.348 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.283 13.465 mp_waitall_1 220248 16.4 12.029 12.930 12.029 12.930 multiply_cannon_sync_h2d 49320 15.4 9.932 10.545 9.932 10.545 qs_ot_get_p 110 10.4 0.001 0.001 10.034 10.123 multiply_cannon_metrocomm3 49320 15.4 0.079 0.084 7.049 8.230 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.143 7.681 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.213 7.638 apply_single 110 13.6 0.000 0.000 7.212 7.638 sum_up_and_integrate 110 10.3 0.036 0.043 7.552 7.564 integrate_v_rspace 110 11.3 0.003 0.003 7.516 7.535 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.180 7.310 calculate_rho_elec 110 8.6 0.021 0.026 7.179 7.310 init_scf_run 11 5.9 0.000 0.001 7.028 7.028 scf_env_initial_rho_setup 11 6.9 0.000 0.001 7.028 7.028 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.941 7.011 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.560 6.604 ot_diis_step 99 11.5 0.006 0.006 6.575 6.576 mp_sum_l 6514 12.8 5.423 6.154 5.423 6.154 init_scf_loop 11 6.9 0.000 0.000 5.835 5.835 cp_dbcsr_syevd 48 12.0 0.002 0.002 5.528 5.529 dbcsr_mm_accdrv_process 87628 16.1 2.087 2.207 5.092 5.403 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.951 4.953 cp_fm_redistribute_end 48 14.0 4.312 4.915 4.316 4.916 cp_fm_diag_elpa_base 48 14.0 0.592 4.756 0.596 4.779 rs_pw_transfer 902 11.9 0.012 0.013 4.231 4.719 density_rs2pw 110 9.6 0.004 0.005 3.909 4.392 make_m2s 4110 13.4 0.061 0.065 4.183 4.279 make_images 4110 14.4 0.177 0.191 4.088 4.187 wfi_extrapolate 11 7.9 0.001 0.001 4.175 4.175 calculate_dm_sparse 110 9.5 0.000 0.001 3.951 4.055 multiply_cannon_metrocomm1 49320 15.4 0.060 0.064 2.768 3.801 pw_transfer 1331 11.6 0.055 0.064 3.637 3.744 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.549 3.661 prepare_preconditioner 11 7.9 0.000 0.000 3.627 3.648 make_preconditioner 11 8.9 0.000 0.000 3.627 3.648 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.624 3.628 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.394 3.438 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.363 3.431 grid_integrate_task_list 110 12.3 3.226 3.411 3.226 3.411 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.236 3.314 fft3d_ps 1111 14.6 0.779 0.881 3.078 3.178 potential_pw2rs 110 12.3 0.006 0.007 3.036 3.083 fft_wrap_pw1pw2_140 451 13.1 0.170 0.190 2.922 3.036 mp_alltoall_d11v 2046 13.8 2.283 2.762 2.283 2.762 calculate_first_density_matrix 1 7.0 0.001 0.001 2.756 2.760 jit_kernel_multiply 13 15.9 2.718 2.737 2.718 2.737 mp_waitany 14300 13.8 1.839 2.328 1.839 2.328 acc_transpose_blocks 49320 15.4 0.227 0.239 2.204 2.293 grid_collocate_task_list 110 9.6 2.084 2.262 2.084 2.262 mp_sum_d 3881 11.9 1.602 2.036 1.602 2.036 cp_fm_cholesky_invert 11 10.9 2.024 2.028 2.024 2.028 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.013 2.028 make_images_data 4110 15.4 0.042 0.045 1.867 2.000 mp_alltoall_z22v 1111 16.6 1.597 1.909 1.597 1.909 hybrid_alltoall_any 4261 16.3 0.082 0.483 1.609 1.862 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.769 1.793 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=86.432000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.363636, yerr=2.739367 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d649f8f97645802a3cd6acffbbd85227349dba89_performance_tests/10/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 390.715586E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 5019072 0.0% 0.0% 100.0% average stack size 0.0 0.0 196.1 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 591.540224E+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 1931029. 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 0.034 0.057 73.832 73.834 qs_mol_dyn_low 1 2.0 0.003 0.003 73.486 73.511 qs_forces 11 3.9 0.002 0.004 73.413 73.414 qs_energies 11 4.9 0.004 0.015 69.948 69.957 scf_env_do_scf 11 5.9 0.001 0.001 60.194 60.197 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 51.779 51.781 dbcsr_multiply_generic 2055 12.4 0.114 0.117 39.431 39.644 velocity_verlet 10 3.0 0.001 0.001 38.122 38.123 qs_scf_new_mos 99 7.5 0.001 0.001 34.522 34.660 qs_scf_loop_do_ot 99 8.5 0.001 0.001 34.521 34.660 multiply_cannon 2055 13.4 0.221 0.241 32.247 33.555 ot_scf_mini 99 9.5 0.003 0.004 32.820 32.955 multiply_cannon_loop 2055 14.4 0.924 0.948 30.808 31.679 ot_mini 99 10.5 0.001 0.001 19.021 19.168 multiply_cannon_multrec 24660 15.4 7.606 9.303 14.540 16.453 rebuild_ks_matrix 110 8.3 0.000 0.001 14.386 14.516 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.018 14.386 14.516 qs_ot_get_derivative 99 11.5 0.001 0.001 13.125 13.268 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.643 12.754 mp_waitall_1 176588 16.5 8.442 11.301 8.442 11.301 multiply_cannon_metrocomm3 24660 15.4 0.070 0.073 5.552 8.837 init_scf_loop 11 6.9 0.001 0.006 8.364 8.372 multiply_cannon_sync_h2d 24660 15.4 6.935 7.946 6.935 7.946 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.585 7.371 apply_single 110 13.6 0.000 0.001 6.585 7.371 init_scf_run 11 5.9 0.000 0.001 7.262 7.263 scf_env_initial_rho_setup 11 6.9 0.000 0.002 7.262 7.263 qs_ot_get_p 110 10.4 0.001 0.002 7.082 7.257 dbcsr_mm_accdrv_process 52282 16.1 4.852 6.372 6.776 7.255 sum_up_and_integrate 110 10.3 0.052 0.059 7.122 7.137 integrate_v_rspace 110 11.3 0.002 0.003 7.070 7.088 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.683 6.697 calculate_rho_elec 110 8.6 0.039 0.047 6.682 6.696 prepare_preconditioner 11 7.9 0.000 0.000 6.230 6.248 make_preconditioner 11 8.9 0.000 0.002 6.230 6.247 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.786 5.951 ot_diis_step 99 11.5 0.010 0.010 5.832 5.832 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.893 5.636 make_m2s 4110 13.4 0.057 0.059 4.570 5.056 qs_ot_p2m_diag 48 11.0 0.028 0.044 4.960 4.984 make_images 4110 14.4 0.399 0.441 4.460 4.942 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.473 4.474 pw_transfer 1331 11.6 0.066 0.073 4.184 4.350 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 4.260 4.262 density_rs2pw 110 9.6 0.004 0.005 3.757 4.252 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 4.077 4.247 rs_pw_transfer 902 11.9 0.012 0.013 3.405 3.913 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.831 3.871 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.830 3.832 cp_fm_redistribute_end 48 14.0 2.853 3.787 2.856 3.787 cp_fm_diag_elpa_base 48 14.0 0.893 3.628 0.926 3.716 wfi_extrapolate 11 7.9 0.001 0.001 3.665 3.665 fft3d_ps 1111 14.6 1.104 1.334 3.430 3.585 calculate_first_density_matrix 1 7.0 0.001 0.003 3.504 3.506 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.429 3.494 fft_wrap_pw1pw2_140 451 13.1 0.201 0.218 3.308 3.476 grid_integrate_task_list 110 12.3 3.154 3.378 3.154 3.378 calculate_dm_sparse 110 9.5 0.001 0.001 3.031 3.066 make_images_data 4110 15.4 0.046 0.050 2.507 2.997 cp_fm_cholesky_invert 11 10.9 2.966 2.974 2.966 2.974 potential_pw2rs 110 12.3 0.008 0.009 2.854 2.885 hybrid_alltoall_any 4261 16.3 0.102 0.444 2.166 2.845 mp_sum_l 6514 12.8 2.013 2.759 2.013 2.759 jit_kernel_multiply 10 16.4 1.572 2.583 1.572 2.583 grid_collocate_task_list 110 9.6 2.055 2.483 2.055 2.483 mp_alltoall_d11v 2046 13.8 1.961 2.345 1.961 2.345 multiply_cannon_metrocomm4 22605 15.4 0.074 0.079 0.804 2.193 qs_ot_get_orbitals 99 10.5 0.001 0.001 2.045 2.063 mp_irecv_dv 57340 16.2 0.679 1.974 0.679 1.974 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.911 1.926 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 1.889 1.890 mp_waitany 10164 13.8 1.268 1.825 1.268 1.825 mp_allgather_i34 2055 14.4 0.770 1.815 0.770 1.815 mp_alltoall_z22v 1111 16.6 1.621 1.744 1.621 1.744 cp_fm_cholesky_decompose 22 10.9 1.606 1.612 1.606 1.612 dbcsr_complete_redistribute 325 12.2 0.292 0.335 1.340 1.612 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.569 1.582 acc_transpose_blocks 24660 15.4 0.113 0.116 1.540 1.561 mp_sum_d 3881 11.9 1.029 1.491 1.029 1.491 rs_pw_transfer_RS2PW_140 121 11.5 0.207 0.217 0.976 1.478 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=73.834000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=556.909091, yerr=8.554154 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d649f8f97645802a3cd6acffbbd85227349dba89_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 659.689472E+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 0.058 0.129 64.197 64.206 qs_mol_dyn_low 1 2.0 0.003 0.003 63.686 63.697 qs_forces 11 3.9 0.002 0.002 63.616 63.617 qs_energies 11 4.9 0.003 0.009 60.315 60.318 scf_env_do_scf 11 5.9 0.001 0.002 51.932 51.932 scf_env_do_scf_inner_loop 99 6.5 0.003 0.010 42.636 42.637 velocity_verlet 10 3.0 0.001 0.001 34.362 34.364 dbcsr_multiply_generic 2055 12.4 0.108 0.111 29.394 29.687 qs_scf_new_mos 99 7.5 0.001 0.001 26.790 26.902 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.790 26.901 ot_scf_mini 99 9.5 0.003 0.003 25.538 25.659 multiply_cannon 2055 13.4 0.210 0.220 22.679 24.134 multiply_cannon_loop 2055 14.4 0.613 0.632 21.404 22.518 ot_mini 99 10.5 0.001 0.001 14.270 14.395 rebuild_ks_matrix 110 8.3 0.000 0.000 12.929 13.081 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.016 12.929 13.081 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.379 11.517 mp_waitall_1 139946 16.5 7.544 10.939 7.544 10.939 multiply_cannon_multrec 16440 15.4 3.835 4.798 9.909 10.743 qs_ot_get_derivative 99 11.5 0.001 0.001 9.671 9.792 init_scf_loop 11 6.9 0.001 0.006 9.246 9.249 multiply_cannon_metrocomm3 16440 15.4 0.043 0.044 4.594 7.786 prepare_preconditioner 11 7.9 0.000 0.001 7.397 7.419 make_preconditioner 11 8.9 0.000 0.002 7.397 7.419 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.710 7.078 sum_up_and_integrate 110 10.3 0.059 0.060 7.015 7.030 integrate_v_rspace 110 11.3 0.002 0.003 6.955 6.970 qs_ot_get_p 110 10.4 0.001 0.002 6.382 6.528 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.362 6.372 calculate_rho_elec 110 8.6 0.058 0.059 6.362 6.371 dbcsr_mm_accdrv_process 34862 16.1 4.814 5.624 5.930 6.083 init_scf_run 11 5.9 0.000 0.001 5.893 5.894 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.893 5.893 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.088 5.544 apply_single 110 13.6 0.000 0.000 5.088 5.544 make_m2s 4110 13.4 0.050 0.051 4.404 4.820 density_rs2pw 110 9.6 0.004 0.005 3.426 4.713 make_images 4110 14.4 0.394 0.514 4.289 4.703 ot_diis_step 99 11.5 0.011 0.011 4.554 4.554 qs_ot_p2m_diag 48 11.0 0.042 0.044 4.524 4.534 rs_pw_transfer 902 11.9 0.010 0.011 3.108 4.369 multiply_cannon_sync_h2d 16440 15.4 3.614 4.170 3.614 4.170 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.159 4.159 pw_transfer 1331 11.6 0.066 0.074 3.999 4.015 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.280 3.927 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.891 3.911 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.526 3.533 cp_fm_redistribute_end 48 14.0 2.196 3.492 2.200 3.493 cp_fm_diag_elpa_base 48 14.0 1.220 3.320 1.281 3.427 grid_integrate_task_list 110 12.3 3.152 3.424 3.152 3.424 fft_wrap_pw1pw2_140 451 13.1 0.212 0.218 3.265 3.286 fft3d_ps 1111 14.6 1.089 1.102 3.191 3.205 wfi_extrapolate 11 7.9 0.001 0.001 3.148 3.148 cp_fm_cholesky_invert 11 10.9 3.023 3.030 3.023 3.030 make_images_data 4110 15.4 0.043 0.047 2.508 3.006 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.896 2.898 hybrid_alltoall_any 4261 16.3 0.105 0.373 2.223 2.876 calculate_dm_sparse 110 9.5 0.001 0.001 2.694 2.719 potential_pw2rs 110 12.3 0.011 0.011 2.686 2.699 calculate_first_density_matrix 1 7.0 0.002 0.004 2.654 2.655 mp_waitany 17072 13.8 1.276 2.590 1.276 2.590 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.523 2.574 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.470 2.535 grid_collocate_task_list 110 9.6 2.082 2.476 2.082 2.476 multiply_cannon_metrocomm4 14385 15.4 0.045 0.050 0.899 2.470 mp_sum_l 6514 12.8 1.720 2.362 1.720 2.362 mp_irecv_dv 48980 15.7 0.829 2.345 0.829 2.345 mp_alltoall_d11v 2046 13.8 1.991 2.342 1.991 2.342 rs_pw_transfer_RS2PW_140 121 11.5 0.175 0.179 0.908 2.158 dbcsr_complete_redistribute 325 12.2 0.319 0.355 1.613 2.086 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 2.013 2.015 cp_fm_upper_to_full 70 13.6 1.422 1.890 1.422 1.890 cp_fm_cholesky_decompose 22 10.9 1.804 1.823 1.804 1.823 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.753 1.770 jit_kernel_multiply 8 16.5 0.725 1.730 0.725 1.730 mp_allgather_i34 2055 14.4 0.604 1.685 0.604 1.685 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.642 1.656 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.048 1.513 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.359 1.480 mp_alltoall_z22v 1111 16.6 1.337 1.368 1.337 1.368 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.333 1.346 rs_gather_matrices 110 12.3 0.232 0.258 1.036 1.340 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=64.206000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=625.363636, yerr=7.376532 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d649f8f97645802a3cd6acffbbd85227349dba89_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 732.684288E+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 0.024 0.054 67.731 67.734 qs_mol_dyn_low 1 2.0 0.003 0.005 67.419 67.429 qs_forces 11 3.9 0.003 0.005 67.138 67.140 qs_energies 11 4.9 0.003 0.009 63.665 63.670 scf_env_do_scf 11 5.9 0.002 0.005 55.117 55.121 scf_env_do_scf_inner_loop 99 6.5 0.004 0.010 42.997 42.998 velocity_verlet 10 3.0 0.001 0.001 38.251 38.262 dbcsr_multiply_generic 2055 12.4 0.115 0.117 29.965 30.192 qs_scf_new_mos 99 7.5 0.001 0.001 27.298 27.411 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.297 27.411 ot_scf_mini 99 9.5 0.003 0.004 25.668 25.769 multiply_cannon 2055 13.4 0.243 0.261 22.666 23.623 multiply_cannon_loop 2055 14.4 0.882 0.899 21.184 21.834 ot_mini 99 10.5 0.001 0.001 14.445 14.556 multiply_cannon_multrec 24660 15.4 4.231 6.886 12.656 14.048 rebuild_ks_matrix 110 8.3 0.000 0.000 12.647 12.755 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.017 12.647 12.755 init_scf_loop 11 6.9 0.003 0.011 12.072 12.075 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.187 11.279 qs_ot_get_derivative 99 11.5 0.001 0.001 10.316 10.417 prepare_preconditioner 11 7.9 0.000 0.001 10.247 10.264 make_preconditioner 11 8.9 0.000 0.002 10.247 10.264 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.477 9.911 dbcsr_mm_accdrv_process 52304 16.0 6.777 8.171 8.279 9.170 sum_up_and_integrate 110 10.3 0.068 0.071 6.795 6.821 integrate_v_rspace 110 11.3 0.003 0.005 6.727 6.750 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.436 6.446 calculate_rho_elec 110 8.6 0.077 0.081 6.435 6.445 mp_waitall_1 121746 16.5 4.659 6.382 4.659 6.382 qs_ot_get_p 110 10.4 0.001 0.001 6.150 6.291 make_m2s 4110 13.4 0.059 0.062 5.645 5.987 make_images 4110 14.4 0.575 0.694 5.505 5.843 init_scf_run 11 5.9 0.000 0.001 5.733 5.734 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.733 5.733 cp_fm_upper_to_full 70 13.8 3.263 4.638 3.263 4.638 qs_ot_p2m_diag 48 11.0 0.055 0.064 4.248 4.263 pw_transfer 1331 11.6 0.065 0.073 4.105 4.140 ot_diis_step 99 11.5 0.011 0.012 4.093 4.093 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.002 4.080 apply_single 110 13.6 0.000 0.000 4.002 4.080 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.000 4.037 dbcsr_complete_redistribute 325 12.2 0.416 0.460 2.731 3.878 density_rs2pw 110 9.6 0.004 0.004 3.320 3.830 cp_dbcsr_syevd 48 12.0 0.003 0.005 3.762 3.762 grid_integrate_task_list 110 12.3 3.264 3.436 3.264 3.436 fft_wrap_pw1pw2_140 451 13.1 0.202 0.212 3.374 3.412 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.344 3.390 multiply_cannon_metrocomm3 24660 15.4 0.035 0.036 1.515 3.368 multiply_cannon_sync_h2d 24660 15.4 3.172 3.328 3.172 3.328 fft3d_ps 1111 14.6 1.092 1.129 3.292 3.316 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.150 3.277 hybrid_alltoall_any 4261 16.3 0.119 0.458 2.439 3.246 make_images_data 4110 15.4 0.046 0.050 2.888 3.246 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.108 3.169 cp_fm_diag_elpa 48 13.0 0.001 0.004 3.145 3.148 rs_pw_transfer 902 11.9 0.010 0.011 2.642 3.137 cp_fm_redistribute_end 48 14.0 1.568 3.103 1.570 3.104 wfi_extrapolate 11 7.9 0.001 0.001 3.091 3.091 cp_fm_diag_elpa_base 48 14.0 1.451 2.956 1.530 3.069 cp_fm_cholesky_invert 11 10.9 2.988 2.999 2.988 2.999 calculate_dm_sparse 110 9.5 0.001 0.001 2.940 2.961 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.919 2.921 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.761 2.875 mp_alltoall_i22 605 13.7 1.634 2.784 1.634 2.784 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.541 2.579 calculate_first_density_matrix 1 7.0 0.002 0.010 2.543 2.545 grid_collocate_task_list 110 9.6 2.182 2.440 2.182 2.440 potential_pw2rs 110 12.3 0.012 0.013 2.410 2.423 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 2.287 2.288 mp_alltoall_d11v 2046 13.8 1.923 2.218 1.923 2.218 jit_kernel_multiply 11 15.8 1.168 2.023 1.168 2.023 cp_fm_cholesky_decompose 22 10.9 1.751 1.791 1.751 1.791 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.719 1.758 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.712 1.725 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.600 1.700 acc_transpose_blocks 24660 15.4 0.108 0.110 1.600 1.636 multiply_cannon_metrocomm4 20550 15.4 0.058 0.062 0.846 1.596 mp_waitany 13376 13.8 1.098 1.570 1.098 1.570 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.550 1.564 mp_sum_l 6514 12.8 0.977 1.549 0.977 1.549 mp_irecv_dv 62702 16.1 0.747 1.520 0.747 1.520 mp_alltoall_z22v 1111 16.6 1.341 1.446 1.341 1.446 mp_allgather_i34 2055 14.4 0.583 1.409 0.583 1.409 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=67.734000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=693.090909, yerr=9.737038 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d649f8f97645802a3cd6acffbbd85227349dba89_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 831.041536E+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 0.041 0.068 58.579 58.582 qs_mol_dyn_low 1 2.0 0.003 0.003 58.243 58.253 qs_forces 11 3.9 0.002 0.002 58.005 58.005 qs_energies 11 4.9 0.001 0.002 54.252 54.257 scf_env_do_scf 11 5.9 0.000 0.001 45.402 45.402 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 37.065 37.066 velocity_verlet 10 3.0 0.001 0.001 32.557 32.571 dbcsr_multiply_generic 2055 12.4 0.105 0.106 23.200 23.323 qs_scf_new_mos 99 7.5 0.001 0.001 21.381 21.426 qs_scf_loop_do_ot 99 8.5 0.001 0.001 21.380 21.426 ot_scf_mini 99 9.5 0.002 0.002 20.126 20.145 multiply_cannon 2055 13.4 0.249 0.267 17.475 18.815 multiply_cannon_loop 2055 14.4 0.319 0.331 16.126 16.475 rebuild_ks_matrix 110 8.3 0.000 0.000 12.308 12.336 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 12.308 12.336 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.936 10.960 ot_mini 99 10.5 0.001 0.001 10.457 10.471 multiply_cannon_multrec 8220 15.4 3.217 4.475 7.653 8.508 init_scf_loop 11 6.9 0.000 0.000 8.284 8.285 mp_waitall_1 103326 16.6 6.063 7.717 6.063 7.717 sum_up_and_integrate 110 10.3 0.079 0.080 6.772 6.785 qs_ot_get_derivative 99 11.5 0.001 0.001 6.694 6.712 integrate_v_rspace 110 11.3 0.003 0.003 6.693 6.705 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.651 6.666 calculate_rho_elec 110 8.6 0.115 0.116 6.650 6.665 prepare_preconditioner 11 7.9 0.000 0.000 6.569 6.571 make_preconditioner 11 8.9 0.000 0.000 6.569 6.571 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.138 6.223 qs_ot_get_p 110 10.4 0.001 0.001 5.716 5.732 init_scf_run 11 5.9 0.000 0.001 5.684 5.684 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.684 5.684 dbcsr_mm_accdrv_process 17442 15.9 2.805 3.888 4.307 5.245 make_m2s 4110 13.4 0.038 0.040 4.368 4.618 make_images 4110 14.4 0.638 0.688 4.239 4.487 pw_transfer 1331 11.6 0.066 0.072 4.364 4.386 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.256 4.282 multiply_cannon_metrocomm3 8220 15.4 0.017 0.018 2.901 4.232 qs_ot_p2m_diag 48 11.0 0.081 0.084 4.226 4.231 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.900 3.901 ot_diis_step 99 11.5 0.012 0.012 3.738 3.738 density_rs2pw 110 9.6 0.004 0.004 3.343 3.717 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.648 3.690 apply_single 110 13.6 0.000 0.000 3.648 3.690 fft_wrap_pw1pw2_140 451 13.1 0.214 0.218 3.633 3.666 grid_integrate_task_list 110 12.3 3.344 3.560 3.344 3.560 fft3d_ps 1111 14.6 1.146 1.184 3.502 3.523 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.308 3.309 cp_fm_cholesky_invert 11 10.9 3.269 3.274 3.269 3.274 cp_fm_redistribute_end 48 14.0 0.846 3.263 0.854 3.265 cp_fm_diag_elpa_base 48 14.0 2.221 3.044 2.398 3.213 multiply_cannon_sync_h2d 8220 15.4 2.916 2.995 2.916 2.995 wfi_extrapolate 11 7.9 0.001 0.001 2.870 2.870 hybrid_alltoall_any 4261 16.3 0.199 0.862 2.406 2.864 make_images_data 4110 15.4 0.038 0.043 2.463 2.858 rs_pw_transfer 902 11.9 0.010 0.010 2.370 2.759 calculate_first_density_matrix 1 7.0 0.001 0.001 2.712 2.713 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.709 2.710 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.688 2.689 calculate_dm_sparse 110 9.5 0.001 0.001 2.571 2.614 grid_collocate_task_list 110 9.6 2.278 2.540 2.278 2.540 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.341 2.352 potential_pw2rs 110 12.3 0.015 0.015 2.283 2.295 mp_alltoall_d11v 2046 13.8 1.817 2.165 1.817 2.165 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.995 2.011 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.772 1.994 jit_kernel_multiply 10 15.8 1.190 1.952 1.190 1.952 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.878 1.887 cp_fm_cholesky_decompose 22 10.9 1.830 1.847 1.830 1.847 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.708 1.712 qs_env_update_s_mstruct 11 6.9 0.001 0.003 1.522 1.654 mp_allgather_i34 2055 14.4 0.536 1.645 0.536 1.645 dbcsr_complete_redistribute 325 12.2 0.548 0.573 1.511 1.622 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.524 1.541 mp_waitany 9240 13.8 1.072 1.497 1.072 1.497 qs_create_task_list 11 7.9 0.001 0.001 1.230 1.337 generate_qs_task_list 11 8.9 0.378 0.445 1.229 1.336 rs_gather_matrices 110 12.3 0.323 0.367 0.972 1.293 mp_alltoall_z22v 1111 16.6 1.248 1.277 1.248 1.277 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.204 1.233 rs_pw_transfer_RS2PW_140 121 11.5 0.161 0.166 0.798 1.201 x_to_yz 550 16.1 0.572 0.583 1.177 1.189 yz_to_x 561 15.1 0.524 0.536 1.167 1.183 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=58.582000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=782.454545, yerr=12.999046 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d649f8f97645802a3cd6acffbbd85227349dba89_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 1.372025E+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 0.032 0.059 92.320 92.330 qs_mol_dyn_low 1 2.0 0.003 0.003 91.943 91.953 qs_forces 11 3.9 0.002 0.002 91.869 91.869 qs_energies 11 4.9 0.001 0.002 87.583 87.584 scf_env_do_scf 11 5.9 0.001 0.001 77.081 77.082 velocity_verlet 10 3.0 0.001 0.001 58.461 58.468 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 47.243 47.244 init_scf_loop 11 6.9 0.000 0.000 29.753 29.754 dbcsr_multiply_generic 2055 12.4 0.120 0.122 29.497 29.606 qs_scf_new_mos 99 7.5 0.001 0.001 28.155 28.204 qs_scf_loop_do_ot 99 8.5 0.001 0.001 28.155 28.203 prepare_preconditioner 11 7.9 0.000 0.000 27.646 27.656 make_preconditioner 11 8.9 0.000 0.000 27.646 27.656 make_full_inverse_cholesky 11 9.9 0.000 0.000 21.683 27.090 ot_scf_mini 99 9.5 0.002 0.002 26.356 26.397 multiply_cannon 2055 13.4 0.344 0.368 21.847 22.425 multiply_cannon_loop 2055 14.4 0.342 0.344 20.041 20.329 cp_fm_upper_to_full 70 14.2 13.036 18.694 13.036 18.694 rebuild_ks_matrix 110 8.3 0.000 0.001 14.661 14.720 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.015 14.660 14.720 ot_mini 99 10.5 0.001 0.001 14.311 14.357 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.283 13.336 dbcsr_complete_redistribute 325 12.2 1.014 1.048 7.629 10.908 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.544 9.831 multiply_cannon_multrec 8220 15.4 4.359 4.577 9.593 9.695 qs_ot_get_derivative 99 11.5 0.001 0.001 9.628 9.666 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.947 9.207 mp_waitall_1 84994 16.7 8.253 9.174 8.253 9.174 mp_alltoall_i22 605 13.7 5.572 8.891 5.572 8.891 qs_rho_update_rho_low 110 7.6 0.001 0.001 8.142 8.173 calculate_rho_elec 110 8.6 0.227 0.227 8.142 8.173 sum_up_and_integrate 110 10.3 0.150 0.153 7.657 7.670 integrate_v_rspace 110 11.3 0.003 0.003 7.507 7.520 qs_ot_get_p 110 10.4 0.001 0.001 6.717 6.758 make_m2s 4110 13.4 0.043 0.043 5.709 6.206 init_scf_run 11 5.9 0.000 0.001 6.144 6.144 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.144 6.144 make_images 4110 14.4 0.878 0.915 5.522 6.018 cp_fm_cholesky_invert 11 10.9 5.906 5.910 5.906 5.910 pw_transfer 1331 11.6 0.075 0.075 5.629 5.635 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 5.513 5.519 dbcsr_mm_accdrv_process 11614 15.7 3.242 3.645 5.091 5.363 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 5.071 5.330 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.745 5.251 apply_single 110 13.6 0.000 0.000 4.745 5.251 qs_ot_p2m_diag 48 11.0 0.151 0.155 4.918 4.925 fft_wrap_pw1pw2_140 451 13.1 0.222 0.223 4.761 4.770 fft3d_ps 1111 14.6 1.295 1.300 4.707 4.714 ot_diis_step 99 11.5 0.016 0.017 4.650 4.650 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.452 4.453 density_rs2pw 110 9.6 0.004 0.004 4.119 4.156 multiply_cannon_sync_h2d 8220 15.4 3.947 3.953 3.947 3.953 make_images_data 4110 15.4 0.041 0.044 3.198 3.872 hybrid_alltoall_any 4261 16.3 0.257 0.560 3.191 3.829 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.786 3.787 cp_fm_diag_elpa_base 48 14.0 3.229 3.427 3.784 3.785 grid_integrate_task_list 110 12.3 3.674 3.730 3.674 3.730 qs_energies_init_hamiltonians 11 5.9 0.006 0.014 3.729 3.730 wfi_extrapolate 11 7.9 0.001 0.001 3.509 3.509 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.027 3.480 calculate_dm_sparse 110 9.5 0.001 0.001 3.233 3.263 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.901 2.904 potential_pw2rs 110 12.3 0.021 0.021 2.695 2.701 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.654 2.674 grid_collocate_task_list 110 9.6 2.639 2.664 2.639 2.664 rs_pw_transfer 902 11.9 0.011 0.011 2.505 2.541 calculate_first_density_matrix 1 7.0 0.001 0.001 2.506 2.506 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.376 2.397 qs_env_update_s_mstruct 11 6.9 0.001 0.001 2.235 2.297 mp_alltoall_d11v 2046 13.8 2.202 2.274 2.202 2.274 cp_fm_cholesky_decompose 22 10.9 2.244 2.260 2.244 2.260 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.100 2.205 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.986 1.992 qs_create_task_list 11 7.9 0.001 0.001 1.904 1.950 generate_qs_task_list 11 8.9 0.735 0.788 1.903 1.950 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.864 1.907 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=92.330000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1229.090909, yerr=60.209414 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d649f8f97645802a3cd6acffbbd85227349dba89_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 628.834304E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175955383376 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4085 56760. MP_Allreduce 11253 785. MP_Sync 170 MP_Alltoall 2226 2520945. MP_SendRecv 24320 18752. MP_ISendRecv 24320 18752. MP_Wait 42476 MP_comm_split 83 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.064 0.069 207.429 207.431 qs_mol_dyn_low 1 2.0 0.003 0.003 206.937 206.951 qs_forces 11 3.9 0.003 0.003 206.846 206.848 qs_energies 11 4.9 0.002 0.002 201.195 201.214 scf_env_do_scf 11 5.9 0.001 0.001 184.590 184.594 scf_env_do_scf_inner_loop 117 6.6 0.005 0.013 163.386 163.388 dbcsr_multiply_generic 2507 12.6 0.178 0.182 124.139 124.855 velocity_verlet 10 3.0 0.001 0.001 124.837 124.838 qs_scf_new_mos 117 7.6 0.001 0.001 122.537 122.821 qs_scf_loop_do_ot 117 8.6 0.001 0.001 122.536 122.820 ot_scf_mini 117 9.6 0.003 0.003 115.959 116.178 multiply_cannon 2507 13.6 0.236 0.244 100.551 103.182 multiply_cannon_loop 2507 14.6 2.115 2.172 98.240 100.520 ot_mini 117 10.6 0.001 0.001 65.871 66.094 multiply_cannon_multrec 60168 15.6 33.090 35.015 41.478 43.145 qs_ot_get_derivative 117 11.6 0.001 0.001 40.971 41.198 rebuild_ks_matrix 128 8.3 0.001 0.001 34.159 34.535 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.018 34.158 34.534 mp_waitall_1 267128 16.5 28.199 31.128 28.199 31.128 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.696 30.993 multiply_cannon_sync_h2d 60168 15.6 27.277 28.639 27.277 28.639 qs_ot_get_p 128 10.4 0.001 0.001 28.219 28.516 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.307 25.136 apply_single 128 13.6 0.001 0.001 24.307 25.136 ot_diis_step 117 11.6 0.008 0.008 24.482 24.483 qs_ot_p2m_diag 83 11.4 0.077 0.091 21.386 21.485 init_scf_loop 11 6.9 0.000 0.000 21.123 21.124 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.040 19.215 cp_dbcsr_syevd 83 12.4 0.004 0.005 18.807 18.808 multiply_cannon_metrocomm3 60168 15.6 0.117 0.122 15.561 18.421 prepare_preconditioner 11 7.9 0.000 0.000 16.446 16.490 make_preconditioner 11 8.9 0.000 0.000 16.446 16.490 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.650 15.825 cp_fm_diag_elpa 83 13.4 0.000 0.001 15.792 15.798 cp_fm_redistribute_end 83 14.4 12.475 15.707 12.490 15.710 cp_fm_diag_elpa_base 83 14.4 3.170 15.313 3.206 15.430 sum_up_and_integrate 128 10.3 0.089 0.108 14.617 14.633 integrate_v_rspace 128 11.3 0.003 0.004 14.527 14.544 make_m2s 5014 13.6 0.104 0.113 14.044 14.350 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.214 14.303 calculate_rho_elec 128 8.7 0.045 0.065 14.213 14.302 make_images 5014 14.6 0.401 0.422 13.862 14.182 init_scf_run 11 5.9 0.000 0.001 12.425 12.426 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.425 12.426 density_rs2pw 128 9.7 0.007 0.007 7.582 10.802 mp_sum_l 7870 13.0 8.278 9.622 8.278 9.622 cp_fm_cholesky_invert 11 10.9 9.439 9.446 9.439 9.446 rs_pw_transfer 1046 11.9 0.016 0.019 6.103 9.367 wfi_extrapolate 11 7.9 0.001 0.001 9.185 9.185 pw_transfer 1547 11.6 0.076 0.093 8.364 8.611 calculate_dm_sparse 128 9.5 0.001 0.001 8.387 8.566 dbcsr_mm_accdrv_process 124484 16.2 3.320 3.467 7.946 8.465 fft_wrap_pw1pw2 1291 12.7 0.011 0.012 8.159 8.412 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.911 8.028 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.846 8.001 make_images_data 5014 15.6 0.065 0.071 6.912 7.769 grid_integrate_task_list 128 12.3 7.052 7.609 7.052 7.609 multiply_cannon_metrocomm1 60168 15.6 0.089 0.094 5.838 7.342 fft_wrap_pw1pw2_140 523 13.2 0.445 0.513 7.008 7.324 hybrid_alltoall_any 5200 16.5 0.291 2.265 6.030 7.196 fft3d_ps 1291 14.7 2.108 2.893 6.941 7.104 cp_dbcsr_sm_fm_multiply 37 9.5 0.003 0.003 6.646 6.654 mp_waitany 16020 13.9 2.647 5.885 2.647 5.885 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.785 5.861 grid_collocate_task_list 128 9.7 4.558 5.798 4.558 5.798 mp_alltoall_d11v 2415 14.1 4.301 5.417 4.301 5.417 rs_pw_transfer_RS2PW_140 139 11.5 0.279 0.299 2.107 5.358 potential_pw2rs 128 12.3 0.009 0.011 4.983 5.020 cp_fm_cholesky_decompose 22 10.9 4.751 4.764 4.751 4.764 mp_sum_d 4462 12.2 3.641 4.228 3.641 4.228 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=207.431000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=595.090909, yerr=6.707588 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d649f8f97645802a3cd6acffbbd85227349dba89_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 2.199914E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 6022464 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.2 marketing flops 145.647559E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 832.659456E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2425920 MPI messages size (bytes): total size 4.132350E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703416E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 71436 2336489472 32768 < size <= 131072 728832 55956209664 131072 < size <= 4194304 1386864 1409906900992 4194304 < size <= 16777216 155760 1473826772352 16777216 < size 68112 1190343475200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4107 56904. MP_Allreduce 11307 944. MP_Sync 170 MP_Alltoall 1983 5621256. MP_SendRecv 12126 47072. MP_ISendRecv 12126 47072. MP_Wait 26114 MP_comm_split 83 MP_ISend 11836 212447. MP_IRecv 11836 212447. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.039 198.009 198.012 qs_mol_dyn_low 1 2.0 0.003 0.005 197.548 197.562 qs_forces 11 3.9 0.005 0.017 197.120 197.122 qs_energies 11 4.9 0.004 0.019 190.220 190.232 scf_env_do_scf 11 5.9 0.002 0.015 173.172 173.182 scf_env_do_scf_inner_loop 118 6.6 0.003 0.010 139.408 139.411 velocity_verlet 10 3.0 0.001 0.001 124.201 124.209 dbcsr_multiply_generic 2527 12.6 0.187 0.192 99.677 100.798 qs_scf_new_mos 118 7.6 0.001 0.001 97.874 98.420 qs_scf_loop_do_ot 118 8.6 0.001 0.001 97.873 98.419 ot_scf_mini 118 9.6 0.004 0.005 93.025 93.692 multiply_cannon 2527 13.6 0.482 0.536 78.534 83.682 multiply_cannon_loop 2527 14.6 1.259 1.304 75.100 77.671 ot_mini 118 10.6 0.001 0.001 51.322 51.914 mp_waitall_1 216438 16.6 25.147 40.868 25.147 40.868 multiply_cannon_multrec 30324 15.6 22.076 27.242 32.152 37.711 rebuild_ks_matrix 129 8.3 0.001 0.001 34.267 34.915 qs_ks_build_kohn_sham_matrix 129 9.3 0.021 0.026 34.267 34.914 init_scf_loop 11 6.9 0.001 0.005 33.655 33.656 qs_ks_update_qs_env 129 7.6 0.001 0.001 30.865 31.465 multiply_cannon_metrocomm3 30324 15.6 0.096 0.104 15.990 29.833 qs_ot_get_derivative 118 11.6 0.001 0.002 29.167 29.831 prepare_preconditioner 11 7.9 0.000 0.001 29.159 29.231 make_preconditioner 11 8.9 0.000 0.002 29.159 29.231 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.851 28.383 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 22.705 23.884 apply_single 129 13.6 0.001 0.001 22.704 23.884 qs_ot_get_p 129 10.4 0.001 0.002 23.225 23.841 ot_diis_step 118 11.6 0.014 0.015 21.976 21.978 multiply_cannon_sync_h2d 30324 15.6 19.323 21.658 19.323 21.658 qs_ot_p2m_diag 83 11.4 0.187 0.215 18.193 18.232 cp_fm_cholesky_invert 11 10.9 17.192 17.207 17.192 17.207 cp_dbcsr_syevd 83 12.4 0.006 0.008 17.021 17.022 make_m2s 5054 13.6 0.090 0.094 14.682 16.362 make_images 5054 14.6 1.163 1.359 14.470 16.149 sum_up_and_integrate 129 10.3 0.117 0.133 15.319 15.349 integrate_v_rspace 129 11.3 0.003 0.003 15.201 15.235 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.753 14.789 calculate_rho_elec 129 8.7 0.089 0.106 14.753 14.789 cp_fm_diag_elpa 83 13.4 0.000 0.001 13.763 13.771 cp_fm_redistribute_end 83 14.4 8.028 13.674 8.040 13.676 cp_fm_diag_elpa_base 83 14.4 5.392 13.151 5.614 13.503 init_scf_run 11 5.9 0.000 0.001 12.041 12.042 scf_env_initial_rho_setup 11 6.9 0.001 0.002 12.041 12.042 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 11.156 11.639 make_images_data 5054 15.6 0.063 0.072 8.860 10.918 multiply_cannon_metrocomm4 27797 15.6 0.097 0.113 3.841 10.689 dbcsr_mm_accdrv_process 62734 16.2 4.732 6.074 9.532 10.425 density_rs2pw 129 9.7 0.006 0.007 8.021 10.380 mp_irecv_dv 70031 16.3 3.644 10.299 3.644 10.299 hybrid_alltoall_any 5240 16.5 0.346 1.540 7.514 10.067 pw_transfer 1559 11.6 0.087 0.109 9.462 9.550 fft_wrap_pw1pw2 1301 12.7 0.011 0.012 9.235 9.328 wfi_extrapolate 11 7.9 0.001 0.001 8.615 8.615 rs_pw_transfer 1054 12.0 0.014 0.017 6.235 8.584 fft_wrap_pw1pw2_140 527 13.2 0.476 0.530 8.024 8.149 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 6.520 7.859 fft3d_ps 1301 14.7 2.774 2.944 7.589 7.637 grid_integrate_task_list 129 12.3 7.206 7.610 7.206 7.610 cp_fm_cholesky_decompose 22 10.9 7.112 7.187 7.112 7.187 calculate_dm_sparse 129 9.5 0.001 0.001 6.618 6.755 mp_sum_l 7930 13.1 4.673 6.518 4.673 6.518 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.396 6.409 grid_collocate_task_list 129 9.7 4.741 5.904 4.741 5.904 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 5.486 5.639 potential_pw2rs 129 12.3 0.015 0.018 5.461 5.502 qs_ot_get_orbitals 118 10.6 0.001 0.001 5.397 5.464 mp_allgather_i34 2527 14.6 2.002 5.397 2.002 5.397 mp_alltoall_d11v 2423 14.1 4.352 5.241 4.352 5.241 mp_waitany 11836 13.9 2.551 4.938 2.551 4.938 rs_pw_transfer_RS2PW_140 140 11.5 0.355 0.390 2.131 4.478 mp_sum_d 4491 12.2 2.747 4.211 2.747 4.211 dbcsr_complete_redistribute 395 12.7 0.765 0.846 3.224 4.073 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=198.012000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=792.090909, yerr=6.126760 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d649f8f97645802a3cd6acffbbd85227349dba89_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 2.928533E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 3984192 0.0% 0.0% 100.0% average stack size 0.0 0.0 1695.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 935.571456E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1042912 MPI messages size (bytes): total size 2.716210E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.604448E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 281856 36943429632 131072 < size <= 4194304 660064 996105256960 4194304 < size <= 16777216 65632 931530938576 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 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 0.022 0.035 183.702 183.704 qs_mol_dyn_low 1 2.0 0.003 0.006 183.209 183.222 qs_forces 11 3.9 0.003 0.003 183.100 183.104 qs_energies 11 4.9 0.002 0.007 176.327 176.338 scf_env_do_scf 11 5.9 0.001 0.002 160.351 160.352 scf_env_do_scf_inner_loop 117 6.6 0.003 0.011 123.488 123.491 velocity_verlet 10 3.0 0.001 0.001 117.588 117.591 qs_scf_new_mos 117 7.6 0.001 0.001 83.958 84.309 qs_scf_loop_do_ot 117 8.6 0.001 0.001 83.957 84.308 dbcsr_multiply_generic 2507 12.6 0.181 0.184 82.966 84.024 ot_scf_mini 117 9.6 0.004 0.005 79.776 80.127 multiply_cannon 2507 13.6 0.496 0.519 62.418 66.882 multiply_cannon_loop 2507 14.6 0.851 0.883 59.284 61.697 ot_mini 117 10.6 0.001 0.001 43.418 43.793 init_scf_loop 11 6.9 0.001 0.006 36.741 36.746 mp_waitall_1 170520 16.6 26.160 35.721 26.160 35.721 prepare_preconditioner 11 7.9 0.000 0.000 32.533 32.588 make_preconditioner 11 8.9 0.001 0.010 32.533 32.588 rebuild_ks_matrix 128 8.3 0.001 0.001 31.701 32.157 qs_ks_build_kohn_sham_matrix 128 9.3 0.018 0.026 31.701 32.156 make_full_inverse_cholesky 11 9.9 0.000 0.000 30.094 31.559 qs_ks_update_qs_env 128 7.6 0.001 0.001 28.542 28.957 multiply_cannon_multrec 20056 15.6 13.338 16.930 22.283 25.783 multiply_cannon_metrocomm3 20056 15.6 0.059 0.063 15.831 25.071 qs_ot_get_derivative 117 11.6 0.001 0.002 23.459 23.830 qs_ot_get_p 128 10.4 0.001 0.002 21.598 22.025 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 20.064 21.014 apply_single 128 13.6 0.001 0.001 20.064 21.014 ot_diis_step 117 11.6 0.018 0.018 19.843 19.844 qs_ot_p2m_diag 83 11.4 0.265 0.272 16.991 17.005 multiply_cannon_sync_h2d 20056 15.6 14.138 16.514 14.138 16.514 make_m2s 5014 13.6 0.081 0.084 15.288 16.233 make_images 5014 14.6 1.173 1.262 15.055 16.000 cp_dbcsr_syevd 83 12.4 0.005 0.005 15.939 15.941 cp_fm_cholesky_invert 11 10.9 15.410 15.421 15.410 15.421 sum_up_and_integrate 128 10.3 0.134 0.145 15.280 15.307 integrate_v_rspace 128 11.3 0.003 0.004 15.146 15.174 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.906 14.947 calculate_rho_elec 128 8.7 0.132 0.147 14.905 14.946 cp_fm_diag_elpa 83 13.4 0.000 0.001 12.689 12.691 cp_fm_redistribute_end 83 14.4 4.799 12.605 4.816 12.608 cp_fm_diag_elpa_base 83 14.4 7.362 11.982 7.768 12.452 init_scf_run 11 5.9 0.000 0.001 10.908 10.909 scf_env_initial_rho_setup 11 6.9 0.000 0.003 10.908 10.908 make_images_data 5014 15.6 0.060 0.068 9.456 10.750 hybrid_alltoall_any 5200 16.5 0.436 2.009 8.133 9.770 pw_transfer 1547 11.6 0.086 0.111 9.605 9.727 density_rs2pw 128 9.7 0.006 0.006 7.722 9.721 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.269 9.522 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 9.380 9.508 multiply_cannon_metrocomm4 17549 15.6 0.062 0.072 3.504 9.375 mp_irecv_dv 50230 16.2 3.382 9.120 3.382 9.120 dbcsr_mm_accdrv_process 41502 16.2 4.468 5.446 8.408 8.574 fft_wrap_pw1pw2_140 523 13.2 0.477 0.531 8.171 8.303 cp_fm_cholesky_decompose 22 10.9 7.906 7.938 7.906 7.938 grid_integrate_task_list 128 12.3 7.262 7.911 7.262 7.911 fft3d_ps 1291 14.7 2.666 2.895 7.645 7.724 wfi_extrapolate 11 7.9 0.001 0.001 7.673 7.674 rs_pw_transfer 1046 11.9 0.014 0.015 5.653 7.666 cp_fm_upper_to_full 105 14.5 5.732 7.431 5.732 7.431 dbcsr_complete_redistribute 395 12.7 1.165 1.194 4.728 6.515 calculate_dm_sparse 128 9.5 0.001 0.001 5.903 5.983 grid_collocate_task_list 128 9.7 4.902 5.771 4.902 5.771 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.577 5.583 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.663 5.369 potential_pw2rs 128 12.3 0.020 0.022 5.294 5.316 mp_alltoall_d11v 2415 14.1 4.605 5.277 4.605 5.277 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.437 5.217 mp_allgather_i34 2507 14.6 1.668 4.835 1.668 4.835 mp_sum_l 7870 13.0 3.452 4.807 3.452 4.807 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.648 4.803 mp_waitany 11748 13.9 2.414 4.399 2.414 4.399 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.114 4.153 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.418 4.145 rs_pw_transfer_RS2PW_140 139 11.5 0.330 0.351 1.920 3.906 mp_alltoall_i22 716 14.1 1.972 3.857 1.972 3.857 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.774 3.775 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.679 3.718 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=183.704000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=887.636364, yerr=11.941317 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d649f8f97645802a3cd6acffbbd85227349dba89_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 4.353791E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 5977344 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.178698E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1143192 MPI messages size (bytes): total size 2.023815E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.770320E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 319024 36042702848 131072 < size <= 4194304 715736 785529176064 4194304 < size <= 16777216 70320 665379475120 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4093 57084. MP_Allreduce 11271 1067. 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 0.079 0.095 190.972 190.979 qs_mol_dyn_low 1 2.0 0.004 0.005 190.472 190.486 qs_forces 11 3.9 0.004 0.006 190.359 190.363 qs_energies 11 4.9 0.004 0.009 183.082 183.094 scf_env_do_scf 11 5.9 0.001 0.002 166.285 166.291 velocity_verlet 10 3.0 0.001 0.001 126.133 126.135 scf_env_do_scf_inner_loop 117 6.6 0.004 0.010 119.498 119.500 qs_scf_new_mos 117 7.6 0.001 0.001 81.377 81.687 qs_scf_loop_do_ot 117 8.6 0.001 0.001 81.377 81.687 dbcsr_multiply_generic 2507 12.6 0.188 0.195 79.462 80.205 ot_scf_mini 117 9.6 0.003 0.004 76.891 77.205 multiply_cannon 2507 13.6 0.554 0.584 54.765 58.147 multiply_cannon_loop 2507 14.6 1.177 1.203 50.920 52.850 init_scf_loop 11 6.9 0.002 0.010 46.658 46.660 ot_mini 117 10.6 0.001 0.001 42.575 42.895 prepare_preconditioner 11 7.9 0.000 0.001 42.483 42.503 make_preconditioner 11 8.9 0.000 0.002 42.483 42.503 make_full_inverse_cholesky 11 9.9 0.000 0.000 36.187 41.120 multiply_cannon_multrec 30084 15.6 14.130 19.744 26.075 31.118 rebuild_ks_matrix 128 8.3 0.001 0.001 30.295 30.632 qs_ks_build_kohn_sham_matrix 128 9.3 0.018 0.020 30.294 30.631 mp_waitall_1 147882 16.7 17.379 27.683 17.379 27.683 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.265 27.572 qs_ot_get_derivative 117 11.6 0.001 0.002 22.940 23.257 make_m2s 5014 13.6 0.096 0.099 20.271 21.242 make_images 5014 14.6 1.947 2.231 19.964 20.937 qs_ot_get_p 128 10.4 0.001 0.002 20.573 20.929 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 19.006 19.648 apply_single 128 13.6 0.001 0.001 19.006 19.648 ot_diis_step 117 11.6 0.017 0.019 19.516 19.519 cp_fm_cholesky_invert 11 10.9 16.371 16.380 16.371 16.380 qs_ot_p2m_diag 83 11.4 0.342 0.389 16.248 16.300 cp_fm_upper_to_full 105 14.7 11.025 16.283 11.025 16.283 sum_up_and_integrate 128 10.3 0.141 0.154 15.016 15.040 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.916 14.952 calculate_rho_elec 128 8.7 0.175 0.190 14.915 14.951 cp_dbcsr_syevd 83 12.4 0.006 0.008 14.937 14.939 integrate_v_rspace 128 11.3 0.003 0.004 14.874 14.904 multiply_cannon_metrocomm3 30084 15.6 0.046 0.049 6.449 14.769 multiply_cannon_sync_h2d 30084 15.6 11.674 12.772 11.674 12.772 dbcsr_complete_redistribute 395 12.7 1.504 1.623 8.906 12.591 make_images_data 5014 15.6 0.063 0.068 10.920 12.510 dbcsr_mm_accdrv_process 62264 16.2 7.520 8.351 11.521 12.078 cp_fm_diag_elpa 83 13.4 0.000 0.000 11.713 11.717 cp_fm_redistribute_end 83 14.4 2.031 11.619 2.053 11.625 cp_fm_diag_elpa_base 83 14.4 8.954 10.989 9.531 11.460 hybrid_alltoall_any 5200 16.5 0.522 2.196 9.865 11.436 copy_fm_to_dbcsr 209 11.7 0.002 0.002 7.505 11.217 init_scf_run 11 5.9 0.000 0.001 10.850 10.851 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.850 10.851 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.273 9.903 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.543 9.774 pw_transfer 1547 11.6 0.088 0.107 9.697 9.764 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 9.468 9.542 mp_alltoall_i22 716 14.1 5.427 9.134 5.427 9.134 density_rs2pw 128 9.7 0.006 0.006 7.350 8.629 fft_wrap_pw1pw2_140 523 13.2 0.480 0.493 8.348 8.435 grid_integrate_task_list 128 12.3 7.524 7.926 7.524 7.926 cp_fm_cholesky_decompose 22 10.9 7.710 7.804 7.710 7.804 fft3d_ps 1291 14.7 2.766 2.853 7.694 7.762 wfi_extrapolate 11 7.9 0.001 0.001 7.698 7.698 multiply_cannon_metrocomm4 25070 15.6 0.076 0.087 2.774 7.117 mp_irecv_dv 76098 16.2 2.631 6.852 2.631 6.852 calculate_dm_sparse 128 9.5 0.001 0.001 6.178 6.272 rs_pw_transfer 1046 11.9 0.013 0.014 4.838 6.144 grid_collocate_task_list 128 9.7 5.087 5.755 5.087 5.755 mp_alltoall_d11v 2415 14.1 4.959 5.740 4.959 5.740 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.414 5.452 potential_pw2rs 128 12.3 0.023 0.023 4.807 4.837 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.411 4.500 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 4.456 4.458 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.353 4.438 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.184 4.236 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=190.979000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1093.000000, yerr=25.743490 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d649f8f97645802a3cd6acffbbd85227349dba89_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 5.865088E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 1960712 0.0% 0.0% 100.0% average stack size 0.0 0.0 3445.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.480753E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 240672 MPI messages size (bytes): total size 1.331455E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.532237E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 113904 59718500352 4194304 < size <= 16777216 104976 550376570880 16777216 < size 20208 721350092304 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8931 51. MP_Alltoall 9654 799394. MP_ISend 40068 2102572. MP_IRecv 40068 2101675. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58352. MP_Allreduce 10977 1175. MP_Sync 87 MP_Alltoall 1712 18838210. MP_SendRecv 3840 122880. MP_ISendRecv 3840 122880. MP_Wait 16122 MP_ISend 10680 423556. MP_IRecv 10680 423556. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.038 178.685 178.686 qs_mol_dyn_low 1 2.0 0.003 0.003 178.231 178.244 qs_forces 11 3.9 0.003 0.003 178.117 178.125 qs_energies 11 4.9 0.002 0.002 170.519 170.525 scf_env_do_scf 11 5.9 0.001 0.001 153.037 153.040 scf_env_do_scf_inner_loop 117 6.6 0.003 0.010 116.733 116.734 velocity_verlet 10 3.0 0.001 0.001 115.754 115.758 qs_scf_new_mos 117 7.6 0.001 0.001 77.393 77.481 qs_scf_loop_do_ot 117 8.6 0.001 0.001 77.393 77.481 dbcsr_multiply_generic 2507 12.6 0.182 0.187 74.718 75.083 ot_scf_mini 117 9.6 0.003 0.004 72.993 73.060 multiply_cannon 2507 13.6 0.590 0.617 54.700 58.878 multiply_cannon_loop 2507 14.6 0.442 0.451 50.080 50.869 ot_mini 117 10.6 0.001 0.001 39.868 39.929 init_scf_loop 11 6.9 0.000 0.000 36.155 36.156 mp_waitall_1 125778 16.7 26.164 33.862 26.164 33.862 prepare_preconditioner 11 7.9 0.000 0.000 32.174 32.200 make_preconditioner 11 8.9 0.000 0.000 32.174 32.200 rebuild_ks_matrix 128 8.3 0.001 0.001 30.316 30.406 qs_ks_build_kohn_sham_matrix 128 9.3 0.018 0.019 30.315 30.406 make_full_inverse_cholesky 11 9.9 0.000 0.000 30.061 30.329 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.566 27.649 multiply_cannon_multrec 10028 15.6 10.428 14.681 18.010 21.947 qs_ot_get_p 128 10.4 0.001 0.001 20.209 20.289 qs_ot_get_derivative 117 11.6 0.001 0.002 19.985 20.059 ot_diis_step 117 11.6 0.020 0.020 19.816 19.817 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.471 19.686 apply_single 128 13.6 0.001 0.001 19.470 19.686 multiply_cannon_metrocomm3 10028 15.6 0.023 0.025 12.397 19.056 make_m2s 5014 13.6 0.065 0.070 16.428 18.543 cp_fm_cholesky_invert 11 10.9 18.450 18.457 18.450 18.457 make_images 5014 14.6 2.276 2.755 16.122 18.241 qs_ot_p2m_diag 83 11.4 0.496 0.503 16.337 16.353 qs_rho_update_rho_low 128 7.7 0.001 0.001 15.732 15.763 calculate_rho_elec 128 8.7 0.259 0.268 15.731 15.763 sum_up_and_integrate 128 10.3 0.180 0.189 15.270 15.329 cp_dbcsr_syevd 83 12.4 0.005 0.005 15.169 15.170 integrate_v_rspace 128 11.3 0.004 0.004 15.090 15.158 make_images_data 5014 15.6 0.051 0.059 10.050 12.644 hybrid_alltoall_any 5200 16.5 0.843 3.820 9.830 12.305 multiply_cannon_sync_h2d 10028 15.6 11.591 12.054 11.591 12.054 cp_fm_diag_elpa 83 13.4 0.000 0.000 11.992 12.006 cp_fm_diag_elpa_base 83 14.4 11.743 11.824 11.981 11.994 init_scf_run 11 5.9 0.000 0.001 10.661 10.662 scf_env_initial_rho_setup 11 6.9 0.000 0.001 10.661 10.661 pw_transfer 1547 11.6 0.085 0.094 10.089 10.139 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 9.866 9.921 fft_wrap_pw1pw2_140 523 13.2 0.496 0.516 8.660 8.710 cp_fm_cholesky_decompose 22 10.9 8.082 8.185 8.082 8.185 grid_integrate_task_list 128 12.3 7.765 8.184 7.765 8.184 density_rs2pw 128 9.7 0.005 0.006 7.262 8.137 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 8.063 8.122 fft3d_ps 1291 14.7 2.711 2.783 8.023 8.060 dbcsr_mm_accdrv_process 20762 16.1 2.825 3.567 7.210 7.883 wfi_extrapolate 11 7.9 0.001 0.001 7.482 7.482 mp_allgather_i34 2507 14.6 2.834 7.091 2.834 7.091 multiply_cannon_metrocomm1 10028 15.6 0.028 0.029 4.411 6.958 calculate_dm_sparse 128 9.5 0.001 0.001 6.141 6.212 mp_alltoall_d11v 2415 14.1 5.118 6.107 5.118 6.107 grid_collocate_task_list 128 9.7 5.392 6.051 5.392 6.051 dbcsr_complete_redistribute 395 12.7 2.117 2.204 5.188 5.581 qs_energies_init_hamiltonians 11 5.9 0.002 0.002 5.305 5.305 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.256 5.265 rs_pw_transfer 1046 11.9 0.012 0.013 4.258 5.162 multiply_cannon_metrocomm4 7521 15.6 0.024 0.027 1.857 4.964 mp_irecv_dv 28860 15.9 1.821 4.884 1.821 4.884 potential_pw2rs 128 12.3 0.026 0.027 4.691 4.719 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.188 4.215 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.562 3.884 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.708 3.738 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.373 3.680 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.655 3.663 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.572 3.652 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=178.686000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1387.000000, yerr=36.022720 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d649f8f97645802a3cd6acffbbd85227349dba89_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1399806074880 0.0% 0.0% 100.0% flops 32 x 32 x 32 1906965479424 0.0% 0.0% 100.0% flops 22 x 9 x 32 1943685697536 0.0% 0.0% 100.0% flops 9 x 22 x 32 1949309632512 0.0% 0.0% 100.0% flops 22 x 22 x 32 2694948675584 0.0% 0.0% 100.0% flops 32 x 32 x 9 4338990710784 0.0% 0.0% 100.0% flops 32 x 32 x 22 5303210868736 0.0% 0.0% 100.0% flops 9 x 32 x 32 5347334946816 0.0% 0.0% 100.0% flops 22 x 32 x 32 6535631601664 0.0% 0.0% 100.0% flops 9 x 32 x 9 11360521838592 0.0% 0.0% 100.0% flops 22 x 32 x 9 14909195169792 0.0% 0.0% 100.0% flops 9 x 32 x 22 14909195169792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19481731219456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.080527E+12 0.0% 0.0% 100.0% flops max/rank 11.516854E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6655092288 0.0% 0.0% 100.0% number of processed stacks 1931936 0.0% 0.0% 100.0% average stack size 0.0 0.0 3444.8 marketing flops 142.438295E+12 ------------------------------------------------------------------------------- # multiplications 2464 max memory usage/rank 2.980286E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 98560 MPI messages size (bytes): total size 1.118682E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.350266E+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 44368 34435235840 4194304 < size <= 16777216 43616 373376942080 16777216 < size 9960 710866914816 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4018 58913. MP_Allreduce 11042 1514. MP_Sync 87 MP_Alltoall 1688 36934192. MP_SendRecv 1764 218624. MP_ISendRecv 1764 218624. MP_Wait 9654 MP_ISend 6312 1080635. MP_IRecv 6312 1080635. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.024 0.042 305.490 305.491 qs_mol_dyn_low 1 2.0 0.003 0.003 304.906 304.917 qs_forces 11 3.9 0.003 0.003 304.805 304.808 qs_energies 11 4.9 0.002 0.002 295.224 295.233 scf_env_do_scf 11 5.9 0.001 0.001 272.751 272.773 velocity_verlet 10 3.0 0.001 0.001 219.304 219.313 scf_env_do_scf_inner_loop 115 6.6 0.007 0.019 143.459 143.461 init_scf_loop 11 6.9 0.000 0.000 129.024 129.028 prepare_preconditioner 11 7.9 0.000 0.000 123.782 123.813 make_preconditioner 11 8.9 0.000 0.000 123.781 123.813 make_full_inverse_cholesky 11 9.9 0.000 0.000 98.548 120.890 qs_scf_new_mos 115 7.6 0.001 0.001 92.370 92.451 qs_scf_loop_do_ot 115 8.6 0.001 0.001 92.369 92.450 ot_scf_mini 115 9.6 0.003 0.004 87.669 87.723 dbcsr_multiply_generic 2464 12.5 0.209 0.216 81.984 82.545 cp_fm_upper_to_full 105 14.8 54.321 78.548 54.321 78.548 multiply_cannon 2464 13.5 0.693 0.770 57.334 58.091 multiply_cannon_loop 2464 14.5 0.462 0.470 53.815 55.113 dbcsr_complete_redistribute 395 12.7 3.973 4.012 31.035 44.607 ot_mini 115 10.6 0.001 0.001 43.745 43.786 copy_fm_to_dbcsr 209 11.7 0.001 0.002 27.536 41.111 rebuild_ks_matrix 126 8.3 0.001 0.001 38.720 38.755 qs_ks_build_kohn_sham_matrix 126 9.3 0.018 0.018 38.720 38.754 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 25.190 38.714 mp_alltoall_i22 712 14.1 23.008 36.533 23.008 36.533 qs_ks_update_qs_env 126 7.6 0.001 0.001 35.530 35.578 cp_fm_cholesky_invert 11 10.9 34.701 34.708 34.701 34.708 mp_waitall_1 101902 16.8 27.811 31.720 27.811 31.720 qs_ot_get_p 126 10.4 0.001 0.001 28.695 28.737 qs_ot_p2m_diag 83 11.4 0.878 0.883 24.541 24.565 qs_ot_get_derivative 115 11.6 0.002 0.002 24.098 24.150 cp_dbcsr_syevd 83 12.4 0.006 0.006 22.809 22.813 qs_rho_update_rho_low 126 7.7 0.001 0.001 20.873 20.881 calculate_rho_elec 126 8.7 0.475 0.476 20.872 20.880 make_m2s 4928 13.5 0.075 0.079 19.935 20.793 make_images 4928 14.5 3.680 3.931 19.462 20.327 sum_up_and_integrate 126 10.3 0.315 0.318 19.920 19.999 integrate_v_rspace 126 11.3 0.004 0.004 19.605 19.681 ot_diis_step 115 11.6 0.022 0.022 19.607 19.608 cp_fm_diag_elpa 83 13.4 0.000 0.000 19.411 19.411 cp_fm_diag_elpa_base 83 14.4 14.913 16.604 19.406 19.406 multiply_cannon_metrocomm3 9856 15.5 0.023 0.023 18.245 19.314 apply_preconditioner_dbcsr 126 12.6 0.000 0.000 18.856 19.015 apply_single 126 13.6 0.001 0.001 18.856 19.015 multiply_cannon_multrec 9856 15.5 10.313 12.045 17.656 17.721 multiply_cannon_sync_h2d 9856 15.5 15.413 15.436 15.413 15.436 pw_transfer 1523 11.6 0.091 0.093 12.739 12.752 hybrid_alltoall_any 5114 16.4 1.283 2.979 10.882 12.716 fft_wrap_pw1pw2 1271 12.7 0.011 0.011 12.508 12.519 make_images_data 4928 15.5 0.058 0.064 10.676 12.426 init_scf_run 11 5.9 0.000 0.001 12.403 12.404 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.403 12.404 fft_wrap_pw1pw2_140 515 13.2 0.533 0.542 11.056 11.081 fft3d_ps 1271 14.7 2.703 2.739 10.571 10.578 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.520 9.572 cp_fm_cholesky_decompose 22 10.9 9.364 9.403 9.364 9.403 mp_alltoall_d11v 2399 14.1 8.560 9.344 8.560 9.344 wfi_extrapolate 11 7.9 0.001 0.001 9.271 9.271 dbcsr_mm_accdrv_process 20422 16.0 3.679 5.644 7.109 8.894 density_rs2pw 126 9.7 0.005 0.005 8.600 8.734 grid_integrate_task_list 126 12.3 8.475 8.647 8.475 8.647 qs_energies_init_hamiltonians 11 5.9 0.003 0.005 8.041 8.042 calculate_dm_sparse 126 9.5 0.001 0.001 6.539 6.582 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.374 6.470 copy_dbcsr_to_fm 186 11.8 0.004 0.005 6.259 6.321 grid_collocate_task_list 126 9.7 6.267 6.305 6.267 6.305 rs_scatter_matrices 137 9.7 3.550 4.346 5.995 6.217 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=305.491000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2671.272727, yerr=159.923329 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d649f8f97645802a3cd6acffbbd85227349dba89_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 1.261728E+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 0.016 0.036 84.912 84.913 qs_energies 1 2.0 0.000 0.000 84.402 84.410 ls_scf 1 3.0 0.000 0.000 83.414 83.422 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.359 72.539 multiply_cannon 111 7.7 0.018 0.021 55.619 57.027 multiply_cannon_loop 111 8.7 0.212 0.229 52.237 53.680 ls_scf_main 1 4.0 0.000 0.000 52.237 52.237 density_matrix_trs4 2 5.0 0.002 0.003 46.711 46.790 ls_scf_init_scf 1 4.0 0.000 0.000 28.133 28.134 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.024 27.067 mp_waitall_1 11031 10.9 22.036 26.112 22.036 26.112 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 24.940 24.955 multiply_cannon_multrec 2664 9.7 8.154 8.931 15.585 17.380 multiply_cannon_sync_h2d 2664 9.7 13.679 15.616 13.679 15.616 make_m2s 222 7.7 0.008 0.010 13.057 13.524 make_images 222 8.7 0.099 0.110 13.035 13.503 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.314 11.844 make_images_data 222 9.7 0.004 0.005 7.604 8.135 dbcsr_mm_accdrv_process 4760 10.4 0.510 0.613 7.046 8.063 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.447 7.988 hybrid_alltoall_any 227 10.6 0.215 1.840 6.551 7.692 dbcsr_mm_accdrv_process_sort 4760 11.4 6.335 7.264 6.335 7.264 calculate_norms 4752 9.8 5.554 6.181 5.554 6.181 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.980 5.144 mp_sum_l 807 5.4 3.143 4.740 3.143 4.740 make_images_sizes 222 9.7 0.000 0.000 0.749 3.829 mp_alltoall_i44 222 10.7 0.749 3.829 0.749 3.829 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.064 3.699 mp_irecv_dv 6231 10.9 2.046 3.671 2.046 3.671 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.321 3.597 arnoldi_extremal 4 6.8 0.000 0.000 3.318 3.340 arnoldi_normal_ev 4 7.8 0.001 0.003 3.318 3.340 build_subspace 16 8.4 0.009 0.012 3.224 3.226 ls_scf_post 1 4.0 0.000 0.000 3.044 3.052 ls_scf_store_result 1 5.0 0.000 0.000 2.860 2.904 dbcsr_special_finalize 555 9.7 0.005 0.006 2.346 2.726 dbcsr_merge_single_wm 555 10.7 0.451 0.575 2.338 2.718 make_images_pack 222 9.7 2.211 2.632 2.213 2.634 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.352 2.581 dbcsr_sort_data 658 11.4 2.146 2.524 2.146 2.524 dbcsr_matrix_vector_mult_local 304 10.0 2.073 2.468 2.075 2.469 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.317 2.458 buffer_matrices_ensure_size 222 8.7 1.752 2.053 1.752 2.053 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.791 1.792 rebuild_ks_matrix 3 7.3 0.000 0.000 1.781 1.783 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.781 1.782 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=84.913000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1133.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d649f8f97645802a3cd6acffbbd85227349dba89_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 2.103931E+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 0.036 0.054 91.219 91.222 qs_energies 1 2.0 0.000 0.001 90.566 90.570 ls_scf 1 3.0 0.000 0.003 89.244 89.249 dbcsr_multiply_generic 111 6.7 0.015 0.016 75.141 75.465 multiply_cannon 111 7.7 0.029 0.045 53.418 56.500 ls_scf_main 1 4.0 0.000 0.007 54.714 54.721 multiply_cannon_loop 111 8.7 0.116 0.122 50.169 53.346 density_matrix_trs4 2 5.0 0.002 0.004 48.966 49.162 ls_scf_init_scf 1 4.0 0.002 0.005 30.956 30.961 mp_waitall_1 9105 10.9 21.085 30.117 21.085 30.117 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.759 29.851 multiply_cannon_multrec 1332 9.7 13.095 17.063 22.426 27.532 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.360 27.383 multiply_cannon_metrocomm3 1332 9.7 0.006 0.007 11.877 20.963 make_m2s 222 7.7 0.006 0.007 15.172 15.803 make_images 222 8.7 1.583 1.982 15.142 15.774 dbcsr_mm_accdrv_process 4041 10.4 0.280 0.460 8.932 10.475 dbcsr_mm_accdrv_process_sort 4041 11.4 8.518 10.032 8.518 10.032 make_images_data 222 9.7 0.004 0.004 8.557 9.428 hybrid_alltoall_any 227 10.6 0.522 2.493 8.006 8.914 mp_sum_l 807 5.4 5.251 8.067 5.251 8.067 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.225 7.848 mp_irecv_dv 3311 11.0 3.205 7.793 3.205 7.793 calculate_norms 2376 9.8 5.994 6.888 5.994 6.888 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.007 6.534 multiply_cannon_sync_h2d 1332 9.7 4.778 5.858 4.778 5.858 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.060 5.260 arnoldi_extremal 4 6.8 0.000 0.000 4.764 4.791 arnoldi_normal_ev 4 7.8 0.001 0.004 4.764 4.791 build_subspace 16 8.4 0.014 0.021 4.494 4.496 ls_scf_post 1 4.0 0.001 0.003 3.573 3.578 dbcsr_matrix_vector_mult 304 9.0 0.009 0.021 3.163 3.407 ls_scf_store_result 1 5.0 0.000 0.000 3.292 3.395 dbcsr_matrix_vector_mult_local 304 10.0 2.743 3.220 2.745 3.222 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 1.283 2.903 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.619 2.703 make_images_pack 222 9.7 2.027 2.384 2.029 2.386 mp_allgather_i34 111 8.7 0.946 2.141 0.946 2.141 dbcsr_sort_data 436 11.2 1.849 2.076 1.849 2.076 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.891 1.898 rebuild_ks_matrix 3 7.3 0.000 0.000 1.878 1.885 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.007 1.878 1.885 dbcsr_data_new 4174 10.1 1.614 1.840 1.614 1.840 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.222000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1705.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d649f8f97645802a3cd6acffbbd85227349dba89_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 2.704175E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265470. MP_Allreduce 3058 11181. MP_Sync 4 MP_Alltoall 47 23526250. MP_SendRecv 93 57600. MP_ISendRecv 93 57600. MP_Wait 639 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.036 0.061 94.479 94.480 qs_energies 1 2.0 0.000 0.000 93.847 93.856 ls_scf 1 3.0 0.000 0.000 92.448 92.458 dbcsr_multiply_generic 111 6.7 0.015 0.017 77.010 77.269 ls_scf_main 1 4.0 0.000 0.000 57.895 57.899 multiply_cannon 111 7.7 0.046 0.128 53.123 56.685 multiply_cannon_loop 111 8.7 0.100 0.109 49.442 53.476 density_matrix_trs4 2 5.0 0.002 0.003 51.849 51.999 mp_waitall_1 7281 11.0 24.026 33.880 24.026 33.880 ls_scf_init_scf 1 4.0 0.000 0.000 30.931 30.934 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.722 29.801 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.289 27.304 multiply_cannon_multrec 888 9.7 12.568 15.428 21.196 24.672 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.114 22.969 make_m2s 222 7.7 0.006 0.006 16.974 18.350 make_images 222 8.7 1.971 2.268 16.936 18.311 make_images_data 222 9.7 0.003 0.004 9.759 10.897 hybrid_alltoall_any 227 10.6 0.621 2.874 9.427 10.847 dbcsr_mm_accdrv_process 3754 10.4 0.270 0.414 8.155 9.362 mp_sum_l 807 5.4 5.355 9.013 5.355 9.013 dbcsr_mm_accdrv_process_sort 3754 11.4 7.785 8.948 7.785 8.948 multiply_cannon_sync_h2d 888 9.7 6.024 7.367 6.024 7.367 multiply_cannon_metrocomm1 888 9.7 0.002 0.003 3.774 7.263 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.108 7.134 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.453 6.987 mp_irecv_dv 2335 11.1 2.437 6.936 2.437 6.936 arnoldi_extremal 4 6.8 0.000 0.000 5.203 5.222 arnoldi_normal_ev 4 7.8 0.001 0.006 5.203 5.222 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.996 5.215 build_subspace 16 8.4 0.014 0.020 4.897 4.901 calculate_norms 1584 9.8 4.263 4.604 4.263 4.604 mp_allgather_i34 111 8.7 1.519 3.906 1.519 3.906 dbcsr_matrix_vector_mult 304 9.0 0.009 0.020 3.479 3.805 ls_scf_post 1 4.0 0.000 0.000 3.623 3.634 dbcsr_matrix_vector_mult_local 304 10.0 3.021 3.601 3.023 3.603 ls_scf_store_result 1 5.0 0.000 0.000 3.341 3.436 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.883 2.979 make_images_pack 222 9.7 1.819 2.123 1.822 2.126 dbcsr_sort_data 325 11.1 1.899 2.085 1.899 2.085 make_images_sizes 222 9.7 0.000 0.000 0.872 1.954 mp_alltoall_i44 222 10.7 0.871 1.954 0.871 1.954 dbcsr_data_release 9322 10.9 1.318 1.947 1.318 1.947 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.933 1.936 rebuild_ks_matrix 3 7.3 0.000 0.000 1.915 1.917 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.915 1.917 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=94.480000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2165.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d649f8f97645802a3cd6acffbbd85227349dba89_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 3.350426E+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 0.074 96.669 96.671 qs_energies 1 2.0 0.000 0.000 95.905 95.928 ls_scf 1 3.0 0.000 0.001 94.245 94.267 dbcsr_multiply_generic 111 6.7 0.016 0.018 77.859 78.133 ls_scf_main 1 4.0 0.002 0.030 58.399 58.405 multiply_cannon 111 7.7 0.061 0.147 51.255 55.890 density_matrix_trs4 2 5.0 0.004 0.038 52.218 52.343 multiply_cannon_loop 111 8.7 0.115 0.125 46.153 48.629 ls_scf_init_scf 1 4.0 0.001 0.006 32.635 32.638 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.392 31.467 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.925 28.939 mp_waitall_1 6369 11.0 22.205 28.798 22.205 28.798 multiply_cannon_multrec 1332 9.7 14.180 17.355 22.141 24.887 make_m2s 222 7.7 0.007 0.008 21.007 22.515 make_images 222 8.7 3.137 3.602 20.957 22.466 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 8.917 16.582 make_images_data 222 9.7 0.004 0.004 11.712 13.376 hybrid_alltoall_any 227 10.6 0.799 3.814 10.973 12.734 dbcsr_mm_accdrv_process 3641 10.4 0.203 0.381 7.593 9.135 dbcsr_mm_accdrv_process_sort 3641 11.4 7.221 8.723 7.221 8.723 mp_sum_l 807 5.4 4.027 7.312 4.027 7.312 multiply_cannon_sync_h2d 1332 9.7 5.484 6.366 5.484 6.366 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.079 6.051 mp_irecv_dv 3229 10.9 2.056 5.978 2.056 5.978 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.091 5.724 arnoldi_extremal 4 6.8 0.000 0.000 5.293 5.309 arnoldi_normal_ev 4 7.8 0.001 0.005 5.293 5.309 build_subspace 16 8.4 0.014 0.021 4.951 4.959 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.547 4.770 mp_allgather_i34 111 8.7 2.227 4.763 2.227 4.763 calculate_norms 2376 9.8 4.192 4.534 4.192 4.534 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.405 4.456 dbcsr_matrix_vector_mult 304 9.0 0.010 0.020 3.602 3.916 dbcsr_matrix_vector_mult_local 304 10.0 3.186 3.688 3.188 3.690 dbcsr_sort_data 658 11.4 3.039 3.454 3.039 3.454 dbcsr_special_finalize 555 9.7 0.006 0.007 2.781 3.236 ls_scf_post 1 4.0 0.002 0.016 3.210 3.233 dbcsr_merge_single_wm 555 10.7 0.530 0.663 2.773 3.228 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.066 3.122 ls_scf_store_result 1 5.0 0.000 0.000 2.939 2.999 dbcsr_data_release 10477 10.7 1.588 2.389 1.588 2.389 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.001 2.003 dbcsr_finalize 304 7.8 0.049 0.061 1.801 1.984 rebuild_ks_matrix 3 7.3 0.000 0.000 1.978 1.980 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.978 1.980 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=96.671000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2771.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d649f8f97645802a3cd6acffbbd85227349dba89_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 4.636586E+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 0.043 0.058 99.642 99.642 qs_energies 1 2.0 0.000 0.000 98.782 98.792 ls_scf 1 3.0 0.000 0.000 96.837 96.850 dbcsr_multiply_generic 111 6.7 0.017 0.018 77.934 78.209 ls_scf_main 1 4.0 0.000 0.000 62.241 62.241 multiply_cannon 111 7.7 0.067 0.233 55.126 60.214 density_matrix_trs4 2 5.0 0.002 0.003 54.938 55.010 multiply_cannon_loop 111 8.7 0.069 0.076 50.726 52.691 mp_waitall_1 5436 11.0 26.554 31.368 26.554 31.368 ls_scf_init_scf 1 4.0 0.000 0.000 30.995 31.001 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.621 29.657 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.430 27.443 multiply_cannon_multrec 444 9.7 13.975 16.770 21.114 22.620 make_m2s 222 7.7 0.004 0.005 18.049 20.460 make_images 222 8.7 3.719 4.415 17.987 20.400 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.679 16.538 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.514 14.030 make_images_data 222 9.7 0.003 0.004 10.162 12.599 hybrid_alltoall_any 227 10.6 0.790 3.767 9.838 12.509 dbcsr_mm_accdrv_process 3003 10.4 0.178 0.368 6.843 7.992 multiply_cannon_sync_h2d 444 9.7 6.556 7.954 6.556 7.954 dbcsr_mm_accdrv_process_sort 3003 11.4 6.524 7.644 6.524 7.644 mp_allgather_i34 111 8.7 2.635 7.043 2.635 7.043 arnoldi_extremal 4 6.8 0.000 0.000 5.900 5.915 arnoldi_normal_ev 4 7.8 0.002 0.006 5.900 5.915 build_subspace 16 8.4 0.015 0.019 5.501 5.511 mp_sum_l 807 5.4 2.799 4.685 2.799 4.685 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.552 4.664 dbcsr_matrix_vector_mult 304 9.0 0.011 0.020 4.200 4.406 dbcsr_matrix_vector_mult_local 304 10.0 3.702 4.183 3.705 4.185 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.705 3.806 calculate_norms 792 9.8 3.585 3.729 3.585 3.729 make_images_sizes 222 9.7 0.000 0.000 1.244 3.726 mp_alltoall_i44 222 10.7 1.243 3.726 1.243 3.726 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.980 3.678 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.630 3.646 mp_irecv_dv 1241 11.2 1.612 3.617 1.612 3.617 ls_scf_post 1 4.0 0.000 0.000 3.601 3.610 ls_scf_store_result 1 5.0 0.000 0.000 3.331 3.378 dbcsr_finalize 304 7.8 0.062 0.078 2.205 2.254 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.232 2.233 rebuild_ks_matrix 3 7.3 0.000 0.000 2.200 2.201 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.200 2.201 dbcsr_merge_all 275 8.9 0.474 0.526 2.051 2.105 dbcsr_data_release 10123 10.8 1.327 2.018 1.327 2.018 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=99.642000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3613.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d649f8f97645802a3cd6acffbbd85227349dba89_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 8.676827E+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 0.080 0.098 108.571 108.572 qs_energies 1 2.0 0.000 0.000 107.104 107.119 ls_scf 1 3.0 0.000 0.000 104.146 104.160 dbcsr_multiply_generic 111 6.7 0.023 0.027 77.270 77.401 ls_scf_main 1 4.0 0.000 0.000 66.033 66.034 density_matrix_trs4 2 5.0 0.002 0.003 56.686 56.741 multiply_cannon 111 7.7 0.113 0.169 49.773 52.087 multiply_cannon_loop 111 8.7 0.067 0.070 46.263 46.690 ls_scf_init_scf 1 4.0 0.000 0.000 34.395 34.395 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.696 32.712 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.872 29.881 mp_waitall_1 4527 11.1 22.100 25.409 22.100 25.409 make_m2s 222 7.7 0.005 0.005 23.988 25.045 make_images 222 8.7 4.593 5.004 23.883 24.938 multiply_cannon_multrec 444 9.7 17.825 18.587 22.486 23.078 hybrid_alltoall_any 227 10.6 1.662 3.624 12.946 16.062 make_images_data 222 9.7 0.003 0.003 13.202 15.728 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.272 11.122 multiply_cannon_sync_h2d 444 9.7 8.847 8.900 8.847 8.900 arnoldi_extremal 4 6.8 0.000 0.000 7.495 7.505 arnoldi_normal_ev 4 7.8 0.004 0.009 7.495 7.505 build_subspace 16 8.4 0.026 0.036 6.933 6.945 dbcsr_matrix_vector_mult 304 9.0 0.016 0.033 5.532 5.698 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.296 5.375 dbcsr_matrix_vector_mult_local 304 10.0 5.060 5.367 5.062 5.370 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.039 5.275 dbcsr_mm_accdrv_process 1814 10.4 0.254 0.318 4.492 4.624 dbcsr_mm_accdrv_process_sort 1814 11.4 4.192 4.328 4.192 4.328 ls_scf_post 1 4.0 0.000 0.000 3.718 3.732 make_images_sizes 222 9.7 0.000 0.000 1.439 3.568 mp_alltoall_i44 222 10.7 1.438 3.567 1.438 3.567 mp_allgather_i34 111 8.7 1.165 3.558 1.165 3.558 ls_scf_store_result 1 5.0 0.000 0.000 3.426 3.437 calculate_norms 792 9.8 3.241 3.281 3.241 3.281 dbcsr_finalize 304 7.8 0.082 0.089 3.093 3.191 dbcsr_merge_all 275 8.9 0.891 0.923 2.879 2.971 qs_energies_init_hamiltonians 1 3.0 0.002 0.002 2.927 2.927 dbcsr_data_release 12724 10.6 2.337 2.871 2.337 2.871 dbcsr_complete_redistribute 5 7.6 1.429 1.465 2.769 2.864 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.623 2.624 rebuild_ks_matrix 3 7.3 0.000 0.000 2.556 2.558 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.556 2.558 matrix_ls_to_qs 2 6.0 0.000 0.000 2.437 2.551 dbcsr_sort_data 325 11.1 2.448 2.518 2.448 2.518 dbcsr_new_transposed 4 7.5 0.245 0.255 2.343 2.368 dbcsr_frobenius_norm 74 6.6 2.057 2.131 2.191 2.226 dbcsr_add_d 103 6.2 0.000 0.000 2.133 2.207 dbcsr_add_anytype 103 7.2 0.860 0.891 2.132 2.207 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=108.572000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6861.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: d649f8f97645802a3cd6acffbbd85227349dba89 Summary: empty Status: OK