=== 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: 7d51e9aba56ae0e0080ac2083bd6f96ee184da78 ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, Cray-FFTW 3.3.8.10, # COSMA 2.6.6, ELPA 2022.11.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.1.0, 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 (19.04.2023) # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.1.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.1 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(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.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/7d51e9aba56ae0e0080ac2083bd6f96ee184da78_performance_tests/01 job id: 46329067 --- 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/7d51e9aba56ae0e0080ac2083bd6f96ee184da78_performance_tests/02 job id: 46329068 --- 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/7d51e9aba56ae0e0080ac2083bd6f96ee184da78_performance_tests/03 job id: 46329069 --- 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/7d51e9aba56ae0e0080ac2083bd6f96ee184da78_performance_tests/04 job id: 46329070 --- 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/7d51e9aba56ae0e0080ac2083bd6f96ee184da78_performance_tests/05 job id: 46329071 --- 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/7d51e9aba56ae0e0080ac2083bd6f96ee184da78_performance_tests/06 job id: 46329072 --- 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/7d51e9aba56ae0e0080ac2083bd6f96ee184da78_performance_tests/07 job id: 46329074 --- 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/7d51e9aba56ae0e0080ac2083bd6f96ee184da78_performance_tests/08 job id: 46329075 --- 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/7d51e9aba56ae0e0080ac2083bd6f96ee184da78_performance_tests/09 job id: 46329076 --- 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/7d51e9aba56ae0e0080ac2083bd6f96ee184da78_performance_tests/10 job id: 46329077 --- 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/7d51e9aba56ae0e0080ac2083bd6f96ee184da78_performance_tests/11 job id: 46329078 --- 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/7d51e9aba56ae0e0080ac2083bd6f96ee184da78_performance_tests/12 job id: 46329079 --- 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/7d51e9aba56ae0e0080ac2083bd6f96ee184da78_performance_tests/13 job id: 46329080 --- 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/7d51e9aba56ae0e0080ac2083bd6f96ee184da78_performance_tests/14 job id: 46329082 --- 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/7d51e9aba56ae0e0080ac2083bd6f96ee184da78_performance_tests/15 job id: 46329083 --- 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/7d51e9aba56ae0e0080ac2083bd6f96ee184da78_performance_tests/16 job id: 46329084 --- 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/7d51e9aba56ae0e0080ac2083bd6f96ee184da78_performance_tests/17 job id: 46329085 --- 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/7d51e9aba56ae0e0080ac2083bd6f96ee184da78_performance_tests/18 job id: 46329086 --- 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/7d51e9aba56ae0e0080ac2083bd6f96ee184da78_performance_tests/19 job id: 46329087 --- 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/7d51e9aba56ae0e0080ac2083bd6f96ee184da78_performance_tests/20 job id: 46329089 --- 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/7d51e9aba56ae0e0080ac2083bd6f96ee184da78_performance_tests/21 job id: 46329090 --- 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/7d51e9aba56ae0e0080ac2083bd6f96ee184da78_performance_tests/22 job id: 46329091 --- 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/7d51e9aba56ae0e0080ac2083bd6f96ee184da78_performance_tests/23 job id: 46329092 --- 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/7d51e9aba56ae0e0080ac2083bd6f96ee184da78_performance_tests/24 job id: 46329094 --- 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/7d51e9aba56ae0e0080ac2083bd6f96ee184da78_performance_tests/25 job id: 46329095 --- 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/7d51e9aba56ae0e0080ac2083bd6f96ee184da78_performance_tests/26 job id: 46329096 --- Point --- name: 510 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/1r/12t) --- Point --- name: 511 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: 512 H2O (4 NVE MD steps on 64 nodes) input file: benchmarks/QS/00512_H2O/H2O-512_md.inp required files: [] output file: result.log # nodes = 64 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7d51e9aba56ae0e0080ac2083bd6f96ee184da78_performance_tests/27 job id: 46329097 --- Point --- name: 601 plot: h2o_512_md regex: CP2K label: (64n/12r/1t) --- Point --- name: 602 plot: h2o_512_md_mem regex: Estimated peak process memory label: (64n/12r/1t) ~~~~~~~ END TEST ~~~~~~~ === END TESTS (description) === ===== PLOTS (description) ===== ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2_mem", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md_mem", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md_mem", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md_mem", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls_mem", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_512_md", title="512 H2O (4 NVE MD steps on 64 nodes)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_512_md_mem", title="512 H2O (4 NVE MD steps on 64 nodes)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" === END PLOTS (description) === ============ RESULTS ============ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7d51e9aba56ae0e0080ac2083bd6f96ee184da78_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 344 9. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.036 133.432 133.433 farming_run 1 2.0 132.926 132.928 133.401 133.404 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.448231E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 230 1103589. MP_Allreduce 491 2254389. MP_Sync 25 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 10 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.007 0.027 115.142 115.143 qs_energies 1 2.0 0.000 0.000 114.952 114.953 mp2_main 1 3.0 0.000 0.000 112.923 112.925 mp2_gpw_main 1 4.0 0.019 0.025 111.941 111.943 mp2_ri_gpw_compute_in 1 5.0 0.172 0.174 92.870 93.383 mp2_ri_gpw_compute_in_loop 1 6.0 0.003 0.004 55.536 56.052 mp2_eri_3c_integrate_gpw 272 7.0 0.152 0.166 41.887 47.297 get_2c_integrals 1 6.0 0.008 0.009 36.666 37.160 integrate_v_rspace 273 8.0 0.434 0.450 25.231 30.352 pw_transfer 6555 10.6 0.380 0.396 27.378 28.070 fft_wrap_pw1pw2 5465 11.4 0.044 0.046 26.039 26.730 grid_integrate_task_list 273 9.0 21.033 26.665 21.033 26.665 fft_wrap_pw1pw2_100 2178 12.4 1.139 1.231 23.569 24.260 rpa_ri_compute_en 1 5.0 0.019 0.020 18.962 19.410 compute_2c_integrals 1 7.0 0.002 0.002 19.160 19.161 compute_2c_integrals_loop_lm 1 8.0 0.003 0.003 18.789 18.897 mp2_eri_2c_integrate_gpw 1 9.0 2.395 2.434 18.787 18.893 cp_fm_cholesky_decompose 12 8.2 17.429 17.979 17.429 17.979 cholesky_decomp 1 7.0 0.000 0.000 16.348 16.889 fft3d_s 5443 13.4 16.201 16.765 16.224 16.788 ao_to_mo_and_store_B_mult_1 272 7.0 10.843 15.567 10.843 15.567 calculate_wavefunction 272 8.0 5.425 5.586 12.567 13.198 rpa_num_int 1 6.0 0.000 0.001 10.608 10.608 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.586 10.607 calc_mat_Q 8 8.0 0.000 0.000 9.460 9.558 contract_S_to_Q 8 9.0 0.000 0.000 8.882 8.981 parallel_gemm_fm 14 9.1 0.000 0.000 8.463 8.561 parallel_gemm_fm_cosma 14 10.1 8.463 8.561 8.463 8.561 calc_potential_gpw 544 9.5 0.005 0.005 8.236 8.552 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 8.155 8.387 potential_pw2rs 545 10.0 0.107 0.110 7.672 8.225 create_integ_mat 1 6.0 0.022 0.027 7.745 7.745 collocate_single_gaussian 272 10.0 0.039 0.042 7.420 7.629 array2fm 1 7.0 0.000 0.000 6.744 7.161 pw_scatter_s 2720 13.7 4.419 4.588 4.419 4.588 pw_gather_s 2722 13.2 3.879 4.080 3.879 4.080 array2fm_buffer_send 1 8.0 2.987 3.182 2.987 3.182 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.940227, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2794.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7d51e9aba56ae0e0080ac2083bd6f96ee184da78_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 0.027 0.037 395.955 395.958 farming_run 1 2.0 395.212 395.222 395.916 395.920 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827141120 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788822 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.210466E+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 743 386399. MP_Allreduce 1941 22272. MP_Sync 37 MP_Alltoall 77 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 0.010 0.029 209.891 209.892 qs_energies 1 2.0 0.000 0.002 209.678 209.685 scf_env_do_scf 1 3.0 0.000 0.000 106.685 106.685 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.811 105.819 rebuild_ks_matrix 4 6.0 0.000 0.000 105.810 105.818 qs_ks_build_kohn_sham_matrix 4 7.0 0.057 0.065 105.810 105.818 hfx_ks_matrix 4 8.0 0.001 0.001 105.427 105.432 integrate_four_center 4 9.0 0.144 0.461 105.427 105.431 mp2_main 1 3.0 0.000 0.003 102.703 102.709 mp2_gpw_main 1 4.0 0.037 0.088 101.847 101.856 integrate_four_center_main 4 10.0 0.096 0.637 96.419 100.034 integrate_four_center_bin 269 11.0 96.324 100.014 96.324 100.014 init_scf_loop 1 4.0 0.000 0.000 92.531 92.531 mp2_ri_gpw_compute_in 1 5.0 0.066 0.084 74.960 76.110 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.004 54.502 55.653 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.158 42.255 47.244 integrate_v_rspace 95 8.0 0.395 0.568 28.638 33.456 pw_transfer 2240 10.6 0.144 0.164 29.947 30.491 fft_wrap_pw1pw2 1868 11.4 0.017 0.020 28.946 29.464 grid_integrate_task_list 95 9.0 23.929 28.969 23.929 28.969 ao_to_mo_and_store_B_mult_1 91 7.0 10.561 28.887 10.561 28.887 mp2_ri_gpw_compute_en 1 5.0 0.055 0.063 26.692 28.417 fft_wrap_pw1pw2_100 730 12.4 1.294 1.461 26.646 27.078 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.833 1.899 24.894 24.905 get_2c_integrals 1 6.0 0.000 0.000 20.368 20.393 compute_2c_integrals 1 7.0 0.003 0.004 19.349 19.355 compute_2c_integrals_loop_lm 1 8.0 0.001 0.002 18.921 19.224 mp2_eri_2c_integrate_gpw 1 9.0 1.731 1.854 18.920 19.222 fft3d_s 1823 13.4 18.447 18.717 18.460 18.731 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.152 14.152 calculate_wavefunction 91 8.0 2.008 2.049 9.733 9.960 mp2_ri_gpw_compute_en_expansio 172 7.0 0.557 0.587 8.748 9.298 potential_pw2rs 186 10.0 0.033 0.036 8.653 9.179 local_gemm 172 8.0 8.190 8.730 8.190 8.730 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.000 0.001 8.287 8.603 mp2_ri_gpw_compute_en_comm 22 7.0 0.506 0.535 7.920 8.434 calc_potential_gpw 182 9.5 0.002 0.002 7.943 8.229 collocate_single_gaussian 91 10.0 0.017 0.021 7.897 8.168 mp_sync 37 10.5 3.652 7.454 3.652 7.454 mp_sendrecv_dm3 2068 8.0 5.942 6.475 5.942 6.475 mp2_ri_gpw_compute_en_ener 172 7.0 6.350 6.458 6.350 6.458 pw_gather_s 912 13.2 4.885 5.303 4.885 5.303 pw_scatter_s 910 13.7 3.930 4.255 3.930 4.255 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.845486, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1497.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7d51e9aba56ae0e0080ac2083bd6f96ee184da78_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 432.218112E+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 272. MP_Sync 530 MP_Alltoall 2083 291651. MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.029 52.740 52.741 qs_mol_dyn_low 1 2.0 0.003 0.004 52.534 52.542 qs_forces 11 3.9 0.002 0.003 52.466 52.467 qs_energies 11 4.9 0.001 0.002 51.006 51.022 scf_env_do_scf 11 5.9 0.001 0.001 45.057 45.058 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 43.063 43.064 dbcsr_multiply_generic 2286 12.5 0.093 0.100 33.039 33.460 qs_scf_new_mos 108 7.5 0.000 0.001 33.125 33.425 qs_scf_loop_do_ot 108 8.5 0.000 0.001 33.124 33.424 ot_scf_mini 108 9.5 0.002 0.002 31.458 31.656 multiply_cannon 2286 13.5 0.183 0.189 25.760 27.135 multiply_cannon_loop 2286 14.5 1.471 1.563 25.114 26.482 velocity_verlet 10 3.0 0.001 0.002 25.035 25.035 ot_mini 108 10.5 0.001 0.001 18.760 19.011 qs_ot_get_derivative 108 11.5 0.001 0.005 15.842 16.008 mp_waitall_1 245248 16.5 8.133 14.443 8.133 14.443 multiply_cannon_metrocomm3 54864 15.5 0.067 0.072 5.855 12.658 multiply_cannon_multrec 54864 15.5 4.277 6.708 7.721 11.191 qs_ot_get_p 119 10.4 0.001 0.001 8.097 8.400 rebuild_ks_matrix 119 8.3 0.000 0.000 7.777 7.930 qs_ks_build_kohn_sham_matrix 119 9.3 0.011 0.012 7.777 7.929 mp_sum_l 7207 12.9 5.436 7.127 5.436 7.127 multiply_cannon_sync_h2d 54864 15.5 5.954 7.095 5.954 7.095 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.857 6.993 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.198 5.650 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.409 5.516 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.374 5.435 init_scf_run 11 5.9 0.000 0.001 4.746 4.746 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.746 4.746 dbcsr_mm_accdrv_process 76910 16.1 1.142 1.815 3.365 4.731 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.551 4.551 sum_up_and_integrate 119 10.3 0.012 0.015 4.419 4.426 integrate_v_rspace 119 11.3 0.002 0.003 4.407 4.415 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.363 4.363 cp_fm_redistribute_end 50 14.0 2.225 4.339 2.230 4.342 cp_fm_diag_elpa_base 50 14.0 2.106 4.233 2.110 4.241 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.064 4.215 calculate_rho_elec 119 8.7 0.012 0.017 4.064 4.215 calculate_dm_sparse 119 9.5 0.000 0.001 2.936 3.109 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.891 3.082 apply_single 119 13.6 0.000 0.000 2.891 3.082 jit_kernel_multiply 13 15.8 2.163 2.829 2.163 2.829 rs_pw_transfer 974 11.9 0.012 0.013 2.616 2.715 multiply_cannon_metrocomm1 54864 15.5 0.052 0.056 1.670 2.682 calculate_first_density_matrix 1 7.0 0.000 0.001 2.646 2.650 ot_diis_step 108 11.5 0.006 0.006 2.637 2.637 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.440 2.504 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.343 2.345 density_rs2pw 119 9.7 0.004 0.004 2.110 2.231 acc_transpose_blocks 54864 15.5 0.223 0.243 1.733 2.194 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.128 2.173 grid_integrate_task_list 119 12.3 2.015 2.119 2.015 2.119 wfi_extrapolate 11 7.9 0.001 0.001 2.042 2.042 init_scf_loop 11 6.9 0.000 0.000 1.977 1.978 mp_sum_d 4135 12.0 1.265 1.956 1.265 1.956 potential_pw2rs 119 12.3 0.004 0.004 1.790 1.797 pw_transfer 1439 11.6 0.052 0.056 1.663 1.728 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.588 1.654 make_m2s 4572 13.5 0.054 0.056 1.528 1.573 make_images 4572 14.5 0.132 0.138 1.446 1.491 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.433 1.457 mp_waitany 12084 13.8 1.213 1.407 1.213 1.407 grid_collocate_task_list 119 9.7 1.291 1.385 1.291 1.385 mp_alltoall_d11v 2130 13.8 1.199 1.383 1.199 1.383 fft3d_ps 1201 14.6 0.371 0.479 1.232 1.295 fft_wrap_pw1pw2_140 487 13.2 0.183 0.199 1.224 1.290 dbcsr_dot_sd 1205 11.9 0.049 0.059 0.696 1.102 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=52.741000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=412.272727, yerr=0.962091 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7d51e9aba56ae0e0080ac2083bd6f96ee184da78_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 468.393984E+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 54 MP_Alltoall 2060 1029299. 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 0.013 0.034 38.166 38.166 qs_mol_dyn_low 1 2.0 0.003 0.005 37.983 37.992 qs_forces 11 3.9 0.004 0.010 37.906 37.907 qs_energies 11 4.9 0.002 0.007 36.224 36.227 scf_env_do_scf 11 5.9 0.001 0.001 31.068 31.068 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 28.679 28.680 dbcsr_multiply_generic 2286 12.5 0.101 0.104 21.196 21.547 qs_scf_new_mos 108 7.5 0.001 0.001 19.689 19.931 qs_scf_loop_do_ot 108 8.5 0.001 0.001 19.688 19.931 ot_scf_mini 108 9.5 0.003 0.003 18.802 18.975 multiply_cannon 2286 13.5 0.207 0.218 16.366 17.962 velocity_verlet 10 3.0 0.004 0.005 17.910 17.911 multiply_cannon_loop 2286 14.5 0.892 0.959 15.291 16.653 ot_mini 108 10.5 0.001 0.002 11.547 11.785 mp_waitall_1 200699 16.5 5.519 10.760 5.519 10.760 multiply_cannon_metrocomm3 27432 15.5 0.066 0.069 4.121 9.525 qs_ot_get_derivative 108 11.5 0.001 0.002 9.107 9.285 multiply_cannon_multrec 27432 15.5 1.971 4.341 5.996 8.737 rebuild_ks_matrix 119 8.3 0.000 0.000 7.084 7.220 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.018 7.083 7.220 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.256 6.380 dbcsr_mm_accdrv_process 47894 16.0 3.156 5.337 3.956 5.813 qs_ot_get_p 119 10.4 0.001 0.001 4.369 4.606 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.518 4.368 apply_preconditioner_dbcsr 119 12.6 0.000 0.001 3.037 4.112 apply_single 119 13.6 0.000 0.000 3.037 4.112 sum_up_and_integrate 119 10.3 0.025 0.030 4.100 4.106 integrate_v_rspace 119 11.3 0.002 0.003 4.075 4.083 mp_sum_l 7207 12.9 1.996 3.947 1.996 3.947 init_scf_run 11 5.9 0.000 0.001 3.945 3.946 scf_env_initial_rho_setup 11 6.9 0.001 0.002 3.945 3.945 qs_rho_update_rho_low 119 7.7 0.001 0.002 3.690 3.721 calculate_rho_elec 119 8.7 0.021 0.024 3.689 3.720 qs_ot_p2m_diag 50 11.0 0.009 0.013 2.980 2.998 rs_pw_transfer 974 11.9 0.010 0.011 2.457 2.881 multiply_cannon_sync_h2d 27432 15.5 2.209 2.837 2.209 2.837 make_m2s 4572 13.5 0.052 0.054 2.381 2.592 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.550 2.551 make_images 4572 14.5 0.200 0.237 2.294 2.503 calculate_first_density_matrix 1 7.0 0.001 0.001 2.466 2.469 density_rs2pw 119 9.7 0.004 0.004 2.021 2.456 ot_diis_step 108 11.5 0.011 0.013 2.391 2.391 init_scf_loop 11 6.9 0.000 0.000 2.369 2.369 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.235 2.235 cp_fm_redistribute_end 50 14.0 1.133 2.210 1.136 2.213 calculate_dm_sparse 119 9.5 0.001 0.001 2.125 2.200 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.074 2.166 cp_fm_diag_elpa_base 50 14.0 1.045 2.119 1.073 2.159 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.926 1.928 grid_integrate_task_list 119 12.3 1.836 1.916 1.836 1.916 pw_transfer 1439 11.6 0.065 0.069 1.883 1.913 jit_kernel_multiply 9 16.2 0.748 1.853 0.748 1.853 potential_pw2rs 119 12.3 0.006 0.006 1.835 1.845 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.791 1.822 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.746 1.784 make_images_data 4572 15.5 0.045 0.051 1.112 1.531 prepare_preconditioner 11 7.9 0.000 0.000 1.480 1.508 make_preconditioner 11 8.9 0.000 0.000 1.480 1.508 acc_transpose_blocks 27432 15.5 0.109 0.114 1.205 1.500 fft_wrap_pw1pw2_140 487 13.2 0.202 0.211 1.430 1.460 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.384 1.441 wfi_extrapolate 11 7.9 0.001 0.002 1.428 1.428 hybrid_alltoall_any 4725 16.4 0.050 0.111 0.959 1.399 grid_collocate_task_list 119 9.7 1.227 1.369 1.227 1.369 fft3d_ps 1201 14.6 0.523 0.582 1.341 1.367 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.343 1.351 mp_allgather_i34 2286 14.5 0.517 1.257 0.517 1.257 mp_alltoall_d11v 2130 13.8 1.124 1.226 1.124 1.226 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.171 1.219 mp_sum_d 4135 12.0 0.554 0.994 0.554 0.994 mp_waitany 5720 13.7 0.514 0.979 0.514 0.979 rs_pw_transfer_RS2PW_140 130 11.5 0.138 0.147 0.539 0.959 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.943 0.945 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.885 0.899 acc_transpose_blocks_kernels 27432 16.5 0.181 0.270 0.672 0.884 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=38.166000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=445.272727, yerr=1.863082 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7d51e9aba56ae0e0080ac2083bd6f96ee184da78_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 503.832576E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63497. MP_Allreduce 10075 307. MP_Sync 54 MP_Alltoall 1821 2533805. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.031 31.924 31.925 qs_mol_dyn_low 1 2.0 0.003 0.005 31.617 31.624 qs_forces 11 3.9 0.003 0.003 31.387 31.388 qs_energies 11 4.9 0.003 0.006 29.816 29.819 scf_env_do_scf 11 5.9 0.001 0.001 25.213 25.213 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 22.732 22.733 dbcsr_multiply_generic 2286 12.5 0.096 0.098 16.135 16.222 velocity_verlet 10 3.0 0.001 0.002 15.183 15.189 qs_scf_new_mos 108 7.5 0.001 0.001 14.627 14.657 qs_scf_loop_do_ot 108 8.5 0.001 0.001 14.626 14.657 ot_scf_mini 108 9.5 0.002 0.003 13.915 13.938 multiply_cannon 2286 13.5 0.196 0.207 12.947 13.793 multiply_cannon_loop 2286 14.5 0.630 0.657 12.176 13.054 ot_mini 108 10.5 0.001 0.001 8.598 8.623 multiply_cannon_multrec 18288 15.5 1.947 2.942 6.737 7.130 qs_ot_get_derivative 108 11.5 0.001 0.005 7.099 7.121 rebuild_ks_matrix 119 8.3 0.000 0.000 6.336 6.357 qs_ks_build_kohn_sham_matrix 119 9.3 0.014 0.018 6.335 6.356 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.597 5.615 dbcsr_mm_accdrv_process 38222 16.0 4.231 5.263 4.706 5.502 sum_up_and_integrate 119 10.3 0.030 0.031 3.923 3.929 integrate_v_rspace 119 11.3 0.003 0.003 3.893 3.902 mp_waitall_1 158411 16.6 2.533 3.554 2.533 3.554 init_scf_run 11 5.9 0.000 0.001 3.420 3.420 scf_env_initial_rho_setup 11 6.9 0.001 0.002 3.420 3.420 qs_rho_update_rho_low 119 7.7 0.001 0.002 3.407 3.415 calculate_rho_elec 119 8.7 0.031 0.032 3.407 3.415 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.620 3.251 qs_ot_get_p 119 10.4 0.001 0.001 3.165 3.187 init_scf_loop 11 6.9 0.001 0.003 2.463 2.465 rs_pw_transfer 974 11.9 0.009 0.010 2.206 2.452 multiply_cannon_metrocomm3 18288 15.5 0.044 0.045 1.449 2.450 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.057 2.356 apply_single 119 13.6 0.000 0.000 2.057 2.356 calculate_first_density_matrix 1 7.0 0.001 0.002 2.213 2.214 density_rs2pw 119 9.7 0.004 0.005 1.925 2.185 qs_ot_p2m_diag 50 11.0 0.013 0.017 2.151 2.156 pw_transfer 1439 11.6 0.065 0.069 1.888 1.898 grid_integrate_task_list 119 12.3 1.801 1.895 1.801 1.895 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.892 1.893 make_m2s 4572 13.5 0.044 0.046 1.731 1.889 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.796 1.808 calculate_dm_sparse 119 9.5 0.000 0.001 1.796 1.806 make_images 4572 14.5 0.190 0.204 1.647 1.803 potential_pw2rs 119 12.3 0.007 0.008 1.689 1.693 prepare_preconditioner 11 7.9 0.000 0.000 1.673 1.676 make_preconditioner 11 8.9 0.000 0.001 1.673 1.676 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.654 1.655 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.646 1.655 cp_fm_diag_elpa_base 50 14.0 1.623 1.633 1.644 1.653 multiply_cannon_sync_h2d 18288 15.5 1.399 1.615 1.399 1.615 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.526 1.610 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.520 1.527 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.516 1.526 ot_diis_step 108 11.5 0.011 0.011 1.480 1.480 fft_wrap_pw1pw2_140 487 13.2 0.254 0.261 1.448 1.458 mp_sum_l 7207 12.9 1.069 1.404 1.069 1.404 grid_collocate_task_list 119 9.7 1.215 1.366 1.215 1.366 fft3d_ps 1201 14.6 0.529 0.545 1.282 1.294 acc_transpose_blocks 18288 15.5 0.075 0.077 1.233 1.250 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.175 1.180 wfi_extrapolate 11 7.9 0.001 0.001 1.152 1.152 jit_kernel_multiply 6 16.3 0.424 1.093 0.424 1.093 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 0.954 0.956 make_images_data 4572 15.5 0.044 0.048 0.784 0.950 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.863 0.885 hybrid_alltoall_any 4725 16.4 0.055 0.113 0.675 0.877 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.808 0.811 acc_transpose_blocks_kernels 18288 16.5 0.210 0.217 0.794 0.805 mp_alltoall_d11v 2130 13.8 0.646 0.776 0.646 0.776 mp_waitany 9880 13.7 0.509 0.757 0.509 0.757 rs_pw_transfer_RS2PW_140 130 11.5 0.119 0.127 0.501 0.741 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.661 0.738 cp_fm_cholesky_invert 11 10.9 0.700 0.703 0.700 0.703 mp_alltoall_z22v 1201 16.6 0.609 0.681 0.609 0.681 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=31.925000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=478.181818, yerr=2.166614 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7d51e9aba56ae0e0080ac2083bd6f96ee184da78_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 536.616960E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63496. MP_Allreduce 10074 349. MP_Sync 54 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.026 34.333 34.333 qs_mol_dyn_low 1 2.0 0.003 0.003 34.151 34.172 qs_forces 11 3.9 0.002 0.003 34.087 34.089 qs_energies 11 4.9 0.002 0.002 32.373 32.380 scf_env_do_scf 11 5.9 0.001 0.002 27.483 27.484 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 24.148 24.149 dbcsr_multiply_generic 2286 12.5 0.100 0.104 17.720 17.841 velocity_verlet 10 3.0 0.001 0.002 17.510 17.511 qs_scf_new_mos 108 7.5 0.001 0.001 15.807 15.861 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.806 15.860 multiply_cannon 2286 13.5 0.226 0.259 14.377 15.185 ot_scf_mini 108 9.5 0.002 0.003 14.883 14.936 multiply_cannon_loop 2286 14.5 0.935 0.967 13.462 14.112 ot_mini 108 10.5 0.001 0.001 9.011 9.079 multiply_cannon_multrec 27432 15.5 2.320 2.992 8.562 8.966 dbcsr_mm_accdrv_process 47916 15.9 5.470 7.286 6.149 7.368 qs_ot_get_derivative 108 11.5 0.001 0.002 7.211 7.264 rebuild_ks_matrix 119 8.3 0.000 0.000 6.509 6.577 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 6.509 6.577 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.775 5.835 sum_up_and_integrate 119 10.3 0.035 0.038 3.758 3.766 integrate_v_rspace 119 11.3 0.003 0.003 3.722 3.731 init_scf_run 11 5.9 0.000 0.001 3.564 3.564 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.563 3.564 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.476 3.502 calculate_rho_elec 119 8.7 0.041 0.046 3.476 3.502 qs_ot_get_p 119 10.4 0.001 0.001 3.293 3.368 init_scf_loop 11 6.9 0.000 0.000 3.315 3.316 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.781 3.099 prepare_preconditioner 11 7.9 0.000 0.000 2.489 2.497 make_preconditioner 11 8.9 0.000 0.000 2.489 2.497 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.101 2.429 mp_waitall_1 137007 16.6 1.749 2.257 1.749 2.257 calculate_first_density_matrix 1 7.0 0.000 0.001 2.209 2.210 make_m2s 4572 13.5 0.054 0.056 2.070 2.189 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.951 2.145 apply_single 119 13.6 0.000 0.000 1.950 2.145 rs_pw_transfer 974 11.9 0.009 0.009 1.930 2.123 density_rs2pw 119 9.7 0.004 0.004 1.887 2.082 make_images 4572 14.5 0.271 0.332 1.963 2.080 qs_ot_p2m_diag 50 11.0 0.015 0.022 2.065 2.074 calculate_dm_sparse 119 9.5 0.000 0.000 2.013 2.068 pw_transfer 1439 11.6 0.065 0.070 1.975 2.012 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.882 1.922 grid_integrate_task_list 119 12.3 1.813 1.917 1.813 1.917 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.862 1.890 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.779 1.780 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.761 1.761 ot_diis_step 108 11.5 0.012 0.012 1.760 1.761 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.612 1.633 fft_wrap_pw1pw2_140 487 13.2 0.288 0.301 1.562 1.602 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.513 1.523 potential_pw2rs 119 12.3 0.008 0.009 1.517 1.521 cp_fm_diag_elpa_base 50 14.0 1.481 1.497 1.511 1.521 jit_kernel_multiply 8 16.0 0.619 1.486 0.619 1.486 acc_transpose_blocks 27432 15.5 0.109 0.111 1.441 1.475 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.858 1.421 fft3d_ps 1201 14.6 0.556 0.607 1.320 1.345 grid_collocate_task_list 119 9.7 1.221 1.343 1.221 1.343 wfi_extrapolate 11 7.9 0.001 0.001 1.307 1.307 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.207 1.216 mp_sum_l 7207 12.9 0.842 1.194 0.842 1.194 cp_fm_upper_to_full 72 14.2 0.812 1.156 0.812 1.156 multiply_cannon_sync_h2d 27432 15.5 1.003 1.097 1.003 1.097 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.079 1.096 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.079 1.081 dbcsr_complete_redistribute 329 12.2 0.119 0.142 0.735 1.008 make_images_data 4572 15.5 0.044 0.048 0.795 0.915 hybrid_alltoall_any 4725 16.4 0.062 0.151 0.706 0.894 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.793 0.873 mp_alltoall_d11v 2130 13.8 0.726 0.843 0.726 0.843 acc_transpose_blocks_kernels 27432 16.5 0.268 0.276 0.816 0.834 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.822 0.828 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.543 0.809 cp_fm_cholesky_invert 11 10.9 0.714 0.717 0.714 0.717 mp_alltoall_i22 627 13.8 0.426 0.717 0.426 0.717 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=34.333000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=509.727273, yerr=2.799055 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7d51e9aba56ae0e0080ac2083bd6f96ee184da78_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 594.022400E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63495. MP_Allreduce 10074 348. MP_Sync 54 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.032 27.963 27.963 qs_mol_dyn_low 1 2.0 0.003 0.004 27.521 27.801 qs_forces 11 3.9 0.002 0.003 27.437 27.438 qs_energies 11 4.9 0.001 0.002 25.710 25.713 scf_env_do_scf 11 5.9 0.001 0.002 21.056 21.056 scf_env_do_scf_inner_loop 108 6.5 0.004 0.007 18.604 18.605 velocity_verlet 10 3.0 0.001 0.002 14.297 14.300 dbcsr_multiply_generic 2286 12.5 0.094 0.099 11.884 11.955 qs_scf_new_mos 108 7.5 0.001 0.001 10.651 10.678 qs_scf_loop_do_ot 108 8.5 0.001 0.001 10.651 10.677 ot_scf_mini 108 9.5 0.002 0.002 9.982 10.009 multiply_cannon 2286 13.5 0.230 0.241 9.414 9.864 multiply_cannon_loop 2286 14.5 0.329 0.339 8.515 8.692 multiply_cannon_multrec 9144 15.5 1.574 1.824 5.726 5.984 rebuild_ks_matrix 119 8.3 0.000 0.000 5.962 5.980 qs_ks_build_kohn_sham_matrix 119 9.3 0.015 0.018 5.961 5.979 ot_mini 108 10.5 0.001 0.001 5.510 5.543 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.304 5.320 qs_ot_get_derivative 108 11.5 0.001 0.002 4.214 4.241 dbcsr_mm_accdrv_process 12550 15.8 3.144 4.100 4.052 4.142 sum_up_and_integrate 119 10.3 0.038 0.041 3.644 3.648 integrate_v_rspace 119 11.3 0.003 0.003 3.606 3.611 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.515 3.521 calculate_rho_elec 119 8.7 0.060 0.061 3.514 3.521 init_scf_run 11 5.9 0.000 0.001 3.221 3.221 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.221 3.221 qs_ot_get_p 119 10.4 0.001 0.001 2.700 2.751 init_scf_loop 11 6.9 0.000 0.000 2.433 2.434 calculate_first_density_matrix 1 7.0 0.000 0.001 2.094 2.095 pw_transfer 1439 11.6 0.066 0.069 2.044 2.055 density_rs2pw 119 9.7 0.004 0.004 1.827 1.964 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.951 1.962 grid_integrate_task_list 119 12.3 1.851 1.948 1.851 1.948 mp_waitall_1 115863 16.7 1.362 1.869 1.362 1.869 make_m2s 4572 13.5 0.035 0.036 1.713 1.867 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.794 1.797 rs_pw_transfer 974 11.9 0.008 0.008 1.659 1.793 make_images 4572 14.5 0.269 0.300 1.623 1.775 jit_kernel_multiply 10 15.9 0.870 1.739 0.870 1.739 calculate_dm_sparse 119 9.5 0.000 0.000 1.697 1.718 prepare_preconditioner 11 7.9 0.000 0.000 1.700 1.704 make_preconditioner 11 8.9 0.000 0.000 1.700 1.704 fft_wrap_pw1pw2_140 487 13.2 0.366 0.376 1.608 1.621 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.592 1.618 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.603 1.603 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.528 1.529 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.493 1.520 grid_collocate_task_list 119 9.7 1.271 1.390 1.271 1.390 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.374 1.382 potential_pw2rs 119 12.3 0.010 0.011 1.376 1.381 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.310 1.318 cp_fm_diag_elpa_base 50 14.0 1.284 1.299 1.308 1.317 fft3d_ps 1201 14.6 0.560 0.571 1.283 1.291 ot_diis_step 108 11.5 0.012 0.013 1.285 1.285 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.226 1.227 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.171 1.198 apply_single 119 13.6 0.000 0.000 1.170 1.198 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.120 1.124 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.093 1.104 wfi_extrapolate 11 7.9 0.001 0.001 1.078 1.078 hybrid_alltoall_any 4725 16.4 0.063 0.176 0.788 1.031 make_images_data 4572 15.5 0.038 0.042 0.802 0.990 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 0.869 0.925 acc_transpose_blocks 9144 15.5 0.038 0.039 0.898 0.907 mp_alltoall_d11v 2130 13.8 0.770 0.899 0.770 0.899 cp_fm_cholesky_invert 11 10.9 0.850 0.852 0.850 0.852 multiply_cannon_sync_h2d 9144 15.5 0.713 0.798 0.713 0.798 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.779 0.787 qs_env_update_s_mstruct 11 6.9 0.002 0.007 0.680 0.730 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.704 0.712 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.331 0.687 acc_transpose_blocks_kernels 9144 16.5 0.117 0.121 0.657 0.662 mp_allgather_i34 2286 14.5 0.230 0.647 0.230 0.647 mp_alltoall_z22v 1201 16.6 0.597 0.629 0.597 0.629 mp_waitany 5200 13.7 0.441 0.573 0.441 0.573 qs_create_task_list 11 7.9 0.000 0.001 0.541 0.568 generate_qs_task_list 11 8.9 0.187 0.209 0.541 0.568 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=27.963000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=562.000000, yerr=4.767313 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7d51e9aba56ae0e0080ac2083bd6f96ee184da78_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 732.737536E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63729. MP_Allreduce 10074 433. MP_Sync 54 MP_Alltoall 1582 7383731. MP_SendRecv 2499 189067. MP_ISendRecv 2499 189067. MP_Wait 6399 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.035 40.656 40.656 qs_mol_dyn_low 1 2.0 0.003 0.003 40.438 40.446 qs_forces 11 3.9 0.002 0.002 40.373 40.375 qs_energies 11 4.9 0.001 0.001 38.413 38.418 scf_env_do_scf 11 5.9 0.001 0.001 32.868 32.869 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 25.137 25.138 velocity_verlet 10 3.0 0.001 0.001 23.012 23.019 dbcsr_multiply_generic 2286 12.5 0.101 0.104 17.172 17.405 qs_scf_new_mos 108 7.5 0.001 0.001 15.538 15.636 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.537 15.635 ot_scf_mini 108 9.5 0.002 0.002 14.473 14.574 multiply_cannon 2286 13.5 0.303 0.310 13.369 14.322 multiply_cannon_loop 2286 14.5 0.344 0.349 12.137 13.091 ot_mini 108 10.5 0.001 0.001 8.702 8.817 multiply_cannon_multrec 9144 15.5 3.288 4.605 8.496 8.555 init_scf_loop 11 6.9 0.000 0.000 7.706 7.709 rebuild_ks_matrix 119 8.3 0.000 0.000 7.174 7.317 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.174 7.317 qs_ot_get_derivative 108 11.5 0.001 0.001 6.694 6.796 prepare_preconditioner 11 7.9 0.000 0.000 6.763 6.777 make_preconditioner 11 8.9 0.000 0.000 6.763 6.777 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.314 6.657 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.482 6.612 dbcsr_mm_accdrv_process 12550 15.8 4.602 6.282 5.086 6.320 cp_fm_upper_to_full 72 14.2 3.170 4.563 3.170 4.563 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.161 4.167 calculate_rho_elec 119 8.7 0.118 0.121 4.161 4.167 sum_up_and_integrate 119 10.3 0.065 0.066 3.930 3.937 integrate_v_rspace 119 11.3 0.003 0.004 3.865 3.871 init_scf_run 11 5.9 0.000 0.001 3.515 3.515 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.514 3.514 qs_ot_get_p 119 10.4 0.001 0.001 3.076 3.215 mp_waitall_1 94719 16.7 2.216 3.163 2.216 3.163 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.403 2.838 dbcsr_complete_redistribute 329 12.2 0.293 0.299 1.973 2.799 pw_transfer 1439 11.6 0.069 0.069 2.613 2.619 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.515 2.521 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.675 2.486 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.186 2.439 apply_single 119 13.6 0.000 0.000 2.186 2.439 make_m2s 4572 13.5 0.038 0.038 2.220 2.374 mp_alltoall_i22 627 13.8 1.419 2.262 1.419 2.262 make_images 4572 14.5 0.350 0.381 2.100 2.255 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.444 2.248 calculate_dm_sparse 119 9.5 0.000 0.000 2.133 2.152 fft_wrap_pw1pw2_140 487 13.2 0.618 0.621 2.128 2.135 density_rs2pw 119 9.7 0.004 0.004 2.110 2.131 multiply_cannon_metrocomm3 9144 15.5 0.019 0.020 1.207 2.102 calculate_first_density_matrix 1 7.0 0.000 0.000 2.087 2.088 grid_integrate_task_list 119 12.3 2.043 2.058 2.043 2.058 ot_diis_step 108 11.5 0.014 0.014 1.984 1.984 qs_ot_p2m_diag 50 11.0 0.042 0.043 1.863 1.864 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.785 1.786 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.767 1.769 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.638 1.691 mp_sum_l 7207 12.9 1.035 1.676 1.035 1.676 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.595 1.595 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.539 1.580 fft3d_ps 1201 14.6 0.594 0.604 1.530 1.535 grid_collocate_task_list 119 9.7 1.476 1.492 1.476 1.492 rs_pw_transfer 974 11.9 0.009 0.009 1.399 1.425 cp_fm_cholesky_invert 11 10.9 1.402 1.405 1.402 1.405 wfi_extrapolate 11 7.9 0.001 0.001 1.361 1.361 hybrid_alltoall_any 4725 16.4 0.088 0.147 1.128 1.351 potential_pw2rs 119 12.3 0.014 0.014 1.341 1.343 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.316 1.316 cp_fm_diag_elpa_base 50 14.0 1.171 1.223 1.315 1.315 make_images_data 4572 15.5 0.042 0.045 1.071 1.257 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.214 1.223 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.116 1.134 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.095 1.118 mp_alltoall_d11v 2130 13.8 1.037 1.066 1.037 1.066 multiply_cannon_sync_h2d 9144 15.5 1.041 1.045 1.041 1.045 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 0.984 1.019 qs_create_task_list 11 7.9 0.000 0.000 0.935 0.947 generate_qs_task_list 11 8.9 0.369 0.389 0.935 0.947 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.903 0.917 acc_transpose_blocks 9144 15.5 0.039 0.040 0.898 0.905 mp_alltoall_z22v 1201 16.6 0.801 0.816 0.801 0.816 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=40.656000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=696.818182, yerr=6.293116 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7d51e9aba56ae0e0080ac2083bd6f96ee184da78_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 482.680832E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66218. MP_Allreduce 9696 492. MP_Sync 52 MP_Alltoall 1938 1631035. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.029 80.550 80.551 qs_mol_dyn_low 1 2.0 0.003 0.004 80.293 80.302 qs_forces 11 3.9 0.003 0.004 80.201 80.201 qs_energies 11 4.9 0.002 0.002 77.334 77.357 scf_env_do_scf 11 5.9 0.001 0.004 68.616 68.619 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 63.249 63.250 dbcsr_multiply_generic 2055 12.4 0.107 0.109 49.893 50.112 qs_scf_new_mos 99 7.5 0.000 0.001 46.027 46.153 qs_scf_loop_do_ot 99 8.5 0.001 0.001 46.026 46.153 ot_scf_mini 99 9.5 0.002 0.002 43.707 43.783 multiply_cannon 2055 13.4 0.179 0.185 41.887 42.859 multiply_cannon_loop 2055 14.4 1.509 1.541 40.970 41.962 velocity_verlet 10 3.0 0.001 0.002 41.689 41.690 ot_mini 99 10.5 0.001 0.001 25.582 25.658 qs_ot_get_derivative 99 11.5 0.001 0.002 18.867 18.975 multiply_cannon_multrec 49320 15.4 12.535 13.171 17.382 18.165 rebuild_ks_matrix 110 8.3 0.000 0.001 14.508 14.608 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.013 14.508 14.607 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.715 12.805 mp_waitall_1 220248 16.4 10.305 11.031 10.305 11.031 multiply_cannon_sync_h2d 49320 15.4 10.388 10.956 10.388 10.956 qs_ot_get_p 110 10.4 0.001 0.001 9.718 9.810 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.185 7.722 apply_single 110 13.6 0.000 0.001 7.185 7.722 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 6.955 7.482 multiply_cannon_metrocomm3 49320 15.4 0.078 0.081 6.314 7.346 sum_up_and_integrate 110 10.3 0.036 0.043 7.056 7.069 integrate_v_rspace 110 11.3 0.003 0.004 7.020 7.042 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.683 6.723 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.509 6.679 calculate_rho_elec 110 8.6 0.021 0.026 6.509 6.678 init_scf_run 11 5.9 0.000 0.001 6.673 6.673 scf_env_initial_rho_setup 11 6.9 0.001 0.002 6.672 6.673 ot_diis_step 99 11.5 0.006 0.006 6.534 6.534 cp_dbcsr_syevd 48 12.0 0.003 0.003 5.699 5.699 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.498 5.542 init_scf_loop 11 6.9 0.000 0.000 5.340 5.341 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.138 5.164 cp_fm_diag_elpa_base 48 14.0 5.126 5.151 5.137 5.163 dbcsr_mm_accdrv_process 87628 16.1 1.915 1.979 4.723 4.980 mp_sum_l 6514 12.8 3.812 4.694 3.812 4.694 rs_pw_transfer 902 11.9 0.011 0.013 3.642 4.117 wfi_extrapolate 11 7.9 0.001 0.001 3.986 3.986 make_m2s 4110 13.4 0.061 0.066 3.732 3.849 density_rs2pw 110 9.6 0.004 0.005 3.312 3.802 calculate_dm_sparse 110 9.5 0.001 0.001 3.690 3.791 make_images 4110 14.4 0.176 0.190 3.637 3.757 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.513 3.518 multiply_cannon_metrocomm1 49320 15.4 0.060 0.062 2.211 3.441 grid_integrate_task_list 110 12.3 3.248 3.405 3.248 3.405 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.197 3.227 prepare_preconditioner 11 7.9 0.000 0.000 3.202 3.218 make_preconditioner 11 8.9 0.000 0.000 3.202 3.218 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.133 3.180 pw_transfer 1331 11.6 0.055 0.068 3.067 3.141 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.014 3.063 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 2.979 3.058 fft_wrap_pw1pw2_140 451 13.1 0.452 0.496 2.557 2.639 potential_pw2rs 110 12.3 0.006 0.007 2.588 2.607 calculate_first_density_matrix 1 7.0 0.000 0.001 2.597 2.606 jit_kernel_multiply 13 15.9 2.534 2.548 2.534 2.548 mp_alltoall_d11v 2046 13.8 2.018 2.457 2.018 2.457 mp_waitany 14300 13.8 1.834 2.384 1.834 2.384 grid_collocate_task_list 110 9.6 2.084 2.349 2.084 2.349 fft3d_ps 1111 14.6 0.791 0.876 2.175 2.224 acc_transpose_blocks 49320 15.4 0.207 0.213 2.070 2.124 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.922 1.940 make_images_data 4110 15.4 0.042 0.046 1.712 1.828 mp_sum_d 3889 11.9 1.297 1.814 1.297 1.814 hybrid_alltoall_any 4261 16.3 0.081 0.479 1.493 1.778 cp_fm_cholesky_invert 11 10.9 1.747 1.751 1.747 1.751 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.644 1.671 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=80.551000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=458.090909, yerr=2.712079 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7d51e9aba56ae0e0080ac2083bd6f96ee184da78_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 567.345152E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1972800 MPI messages size (bytes): total size 1.077520E+12 min size 0.000000E+00 max size 4.537280E+06 average size 546.188250E+03 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 222984 1826684928 8192 < size <= 32768 520356 13399818240 32768 < size <= 131072 372336 35386294272 131072 < size <= 4194304 787758 788321309808 4194304 < size <= 16777216 54450 238588003280 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66437. MP_Allreduce 9695 570. MP_Sync 52 MP_Alltoall 1717 2346002. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.028 68.396 68.397 qs_mol_dyn_low 1 2.0 0.003 0.004 68.193 68.203 qs_forces 11 3.9 0.003 0.004 68.113 68.117 qs_energies 11 4.9 0.001 0.002 64.776 64.782 scf_env_do_scf 11 5.9 0.001 0.002 56.320 56.324 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 48.871 48.871 dbcsr_multiply_generic 2055 12.4 0.115 0.121 37.475 37.676 velocity_verlet 10 3.0 0.001 0.002 35.918 35.920 qs_scf_new_mos 99 7.5 0.001 0.001 32.605 32.751 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.605 32.750 multiply_cannon 2055 13.4 0.221 0.243 31.000 32.052 ot_scf_mini 99 9.5 0.003 0.003 30.958 31.088 multiply_cannon_loop 2055 14.4 0.917 0.938 29.715 30.536 ot_mini 99 10.5 0.001 0.001 18.049 18.194 multiply_cannon_multrec 24660 15.4 7.637 9.354 13.836 15.514 rebuild_ks_matrix 110 8.3 0.000 0.001 13.773 13.929 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.019 13.773 13.929 qs_ot_get_derivative 99 11.5 0.001 0.004 12.223 12.351 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.112 12.243 mp_waitall_1 176588 16.5 7.557 10.205 7.557 10.205 multiply_cannon_sync_h2d 24660 15.4 7.080 7.938 7.080 7.938 multiply_cannon_metrocomm3 24660 15.4 0.067 0.072 5.132 7.740 init_scf_loop 11 6.9 0.000 0.000 7.415 7.417 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.556 7.287 apply_single 110 13.6 0.000 0.001 6.556 7.287 sum_up_and_integrate 110 10.3 0.053 0.059 6.609 6.619 integrate_v_rspace 110 11.3 0.003 0.003 6.556 6.568 qs_ot_get_p 110 10.4 0.001 0.001 6.287 6.437 dbcsr_mm_accdrv_process 52282 16.1 4.766 5.834 6.038 6.363 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.123 6.132 calculate_rho_elec 110 8.6 0.040 0.048 6.123 6.132 init_scf_run 11 5.9 0.000 0.001 6.109 6.110 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.109 6.109 ot_diis_step 99 11.5 0.010 0.010 5.780 5.780 prepare_preconditioner 11 7.9 0.000 0.000 5.402 5.415 make_preconditioner 11 8.9 0.000 0.000 5.402 5.415 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.614 5.346 make_full_inverse_cholesky 11 9.9 0.000 0.000 4.977 5.131 make_m2s 4110 13.4 0.057 0.059 4.106 4.580 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.462 4.483 make_images 4110 14.4 0.398 0.441 3.999 4.470 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.039 4.039 density_rs2pw 110 9.6 0.004 0.005 3.244 3.843 pw_transfer 1331 11.6 0.066 0.075 3.609 3.743 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.501 3.636 rs_pw_transfer 902 11.9 0.012 0.013 2.917 3.520 wfi_extrapolate 11 7.9 0.001 0.001 3.506 3.506 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.477 3.493 cp_fm_diag_elpa_base 48 14.0 3.432 3.453 3.474 3.490 grid_integrate_task_list 110 12.3 3.152 3.354 3.152 3.354 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.270 3.271 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.133 3.214 fft_wrap_pw1pw2_140 451 13.1 0.521 0.537 3.005 3.140 calculate_dm_sparse 110 9.5 0.001 0.001 2.954 2.981 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.937 2.978 hybrid_alltoall_any 4261 16.3 0.102 0.440 2.039 2.805 make_images_data 4110 15.4 0.046 0.050 2.330 2.805 fft3d_ps 1111 14.6 1.105 1.324 2.474 2.621 cp_fm_cholesky_invert 11 10.9 2.519 2.526 2.519 2.526 calculate_first_density_matrix 1 7.0 0.000 0.001 2.513 2.517 grid_collocate_task_list 110 9.6 2.056 2.468 2.056 2.468 mp_sum_l 6514 12.8 1.740 2.415 1.740 2.415 potential_pw2rs 110 12.3 0.008 0.008 2.396 2.410 mp_alltoall_d11v 2046 13.8 1.741 2.001 1.741 2.001 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.935 1.954 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.843 1.847 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.817 1.839 mp_waitany 10164 13.8 1.209 1.782 1.209 1.782 jit_kernel_multiply 9 16.4 0.922 1.671 0.922 1.671 mp_allgather_i34 2055 14.4 0.612 1.658 0.612 1.658 multiply_cannon_metrocomm4 22605 15.4 0.075 0.081 0.781 1.628 acc_transpose_blocks 24660 15.4 0.110 0.115 1.536 1.554 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.538 1.547 rs_pw_transfer_RS2PW_140 121 11.5 0.205 0.216 0.948 1.543 mp_irecv_dv 57340 16.2 0.656 1.475 0.656 1.475 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.368 1.473 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=68.397000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=537.181818, yerr=6.220467 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7d51e9aba56ae0e0080ac2083bd6f96ee184da78_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 641.998848E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66428. MP_Allreduce 9694 567. MP_Sync 52 MP_Alltoall 1496 4511006. MP_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.030 59.054 59.054 qs_mol_dyn_low 1 2.0 0.003 0.004 58.793 58.804 qs_forces 11 3.9 0.003 0.003 58.723 58.724 qs_energies 11 4.9 0.001 0.002 55.516 55.519 scf_env_do_scf 11 5.9 0.001 0.001 47.648 47.648 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 39.283 39.284 velocity_verlet 10 3.0 0.001 0.002 32.184 32.185 dbcsr_multiply_generic 2055 12.4 0.107 0.112 28.039 28.312 qs_scf_new_mos 99 7.5 0.001 0.001 24.347 24.456 qs_scf_loop_do_ot 99 8.5 0.001 0.001 24.346 24.456 ot_scf_mini 99 9.5 0.002 0.003 23.140 23.252 multiply_cannon 2055 13.4 0.212 0.219 21.973 23.098 multiply_cannon_loop 2055 14.4 0.614 0.627 20.826 21.852 ot_mini 99 10.5 0.001 0.001 13.232 13.342 rebuild_ks_matrix 110 8.3 0.000 0.000 12.302 12.486 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.015 12.302 12.485 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.829 10.997 multiply_cannon_multrec 16440 15.4 3.945 4.861 9.705 10.546 mp_waitall_1 139946 16.5 6.761 9.917 6.761 9.917 qs_ot_get_derivative 99 11.5 0.001 0.002 8.844 8.958 init_scf_loop 11 6.9 0.000 0.000 8.332 8.334 multiply_cannon_metrocomm3 16440 15.4 0.042 0.045 4.209 7.156 prepare_preconditioner 11 7.9 0.000 0.000 6.586 6.606 make_preconditioner 11 8.9 0.000 0.000 6.586 6.606 sum_up_and_integrate 110 10.3 0.061 0.061 6.493 6.508 integrate_v_rspace 110 11.3 0.003 0.003 6.433 6.447 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.933 6.289 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.914 5.924 calculate_rho_elec 110 8.6 0.059 0.060 5.913 5.924 dbcsr_mm_accdrv_process 34862 16.1 4.800 5.157 5.614 5.785 init_scf_run 11 5.9 0.000 0.001 5.491 5.491 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.491 5.491 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.904 5.329 apply_single 110 13.6 0.000 0.000 4.903 5.329 qs_ot_get_p 110 10.4 0.001 0.001 5.149 5.299 make_m2s 4110 13.4 0.049 0.051 4.041 4.414 ot_diis_step 99 11.5 0.010 0.011 4.362 4.362 make_images 4110 14.4 0.391 0.507 3.928 4.300 density_rs2pw 110 9.6 0.004 0.005 3.034 4.243 multiply_cannon_sync_h2d 16440 15.4 3.718 4.203 3.718 4.203 rs_pw_transfer 902 11.9 0.010 0.011 2.594 3.789 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.039 3.674 pw_transfer 1331 11.6 0.066 0.073 3.603 3.610 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.600 3.604 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.496 3.505 grid_integrate_task_list 110 12.3 3.197 3.425 3.197 3.425 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.272 3.272 fft_wrap_pw1pw2_140 451 13.1 0.638 0.644 3.036 3.047 wfi_extrapolate 11 7.9 0.001 0.001 2.939 2.939 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.837 2.839 make_images_data 4110 15.4 0.043 0.047 2.354 2.830 hybrid_alltoall_any 4261 16.3 0.105 0.375 2.122 2.783 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.756 2.765 cp_fm_diag_elpa_base 48 14.0 2.692 2.722 2.755 2.763 cp_fm_cholesky_invert 11 10.9 2.565 2.571 2.565 2.571 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.505 2.556 calculate_dm_sparse 110 9.5 0.001 0.001 2.481 2.510 grid_collocate_task_list 110 9.6 2.081 2.477 2.081 2.477 calculate_first_density_matrix 1 7.0 0.000 0.001 2.468 2.469 multiply_cannon_metrocomm4 14385 15.4 0.044 0.048 0.861 2.416 mp_waitany 17072 13.8 1.164 2.414 1.164 2.414 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.263 2.335 fft3d_ps 1111 14.6 1.090 1.104 2.318 2.328 mp_irecv_dv 48980 15.7 0.792 2.292 0.792 2.292 potential_pw2rs 110 12.3 0.011 0.011 2.232 2.240 rs_pw_transfer_RS2PW_140 121 11.5 0.174 0.178 0.885 2.069 mp_sum_l 6514 12.8 1.404 2.038 1.404 2.038 mp_alltoall_d11v 2046 13.8 1.685 2.022 1.685 2.022 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.952 1.953 dbcsr_complete_redistribute 325 12.2 0.342 0.371 1.380 1.818 cp_fm_upper_to_full 70 14.2 1.381 1.743 1.381 1.743 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.665 1.682 mp_allgather_i34 2055 14.4 0.472 1.553 0.472 1.553 cp_fm_cholesky_decompose 22 10.9 1.535 1.551 1.535 1.551 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.374 1.500 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.453 1.463 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.947 1.385 acc_transpose_blocks 16440 15.4 0.073 0.076 1.221 1.240 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.222 1.234 rs_gather_matrices 110 12.3 0.234 0.261 0.926 1.209 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=59.054000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=608.454545, yerr=9.069065 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7d51e9aba56ae0e0080ac2083bd6f96ee184da78_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 720.195584E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66426. MP_Allreduce 9694 608. MP_Sync 52 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.024 0.055 64.296 64.298 qs_mol_dyn_low 1 2.0 0.003 0.003 63.984 63.994 qs_forces 11 3.9 0.003 0.003 63.914 63.914 qs_energies 11 4.9 0.003 0.007 60.498 60.501 scf_env_do_scf 11 5.9 0.001 0.002 52.198 52.201 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 40.679 40.681 velocity_verlet 10 3.0 0.001 0.002 36.526 36.528 dbcsr_multiply_generic 2055 12.4 0.115 0.121 28.975 29.177 qs_scf_new_mos 99 7.5 0.001 0.001 25.765 25.881 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.765 25.880 ot_scf_mini 99 9.5 0.003 0.004 24.154 24.270 multiply_cannon 2055 13.4 0.243 0.261 22.289 23.258 multiply_cannon_loop 2055 14.4 0.880 0.903 20.928 21.526 ot_mini 99 10.5 0.001 0.001 13.776 13.910 multiply_cannon_multrec 24660 15.4 4.216 6.929 12.612 13.831 rebuild_ks_matrix 110 8.3 0.000 0.000 12.119 12.219 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.016 12.118 12.218 init_scf_loop 11 6.9 0.001 0.006 11.480 11.488 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.694 10.780 qs_ot_get_derivative 99 11.5 0.001 0.002 9.642 9.759 prepare_preconditioner 11 7.9 0.000 0.000 9.718 9.732 make_preconditioner 11 8.9 0.000 0.001 9.718 9.732 make_full_inverse_cholesky 11 9.9 0.000 0.000 7.933 9.395 dbcsr_mm_accdrv_process 52304 16.0 7.337 8.690 8.250 9.203 sum_up_and_integrate 110 10.3 0.067 0.070 6.478 6.494 integrate_v_rspace 110 11.3 0.003 0.003 6.411 6.427 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.046 6.059 calculate_rho_elec 110 8.6 0.078 0.082 6.046 6.059 mp_waitall_1 121746 16.5 4.073 5.931 4.073 5.931 qs_ot_get_p 110 10.4 0.002 0.006 5.469 5.600 init_scf_run 11 5.9 0.000 0.001 5.558 5.558 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.557 5.558 make_m2s 4110 13.4 0.059 0.062 5.157 5.475 make_images 4110 14.4 0.576 0.701 5.017 5.330 cp_fm_upper_to_full 70 14.2 3.314 4.818 3.314 4.818 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.014 4.108 apply_single 110 13.6 0.000 0.000 4.014 4.108 ot_diis_step 99 11.5 0.011 0.011 4.097 4.097 dbcsr_complete_redistribute 325 12.2 0.428 0.472 2.646 3.783 qs_ot_p2m_diag 48 11.0 0.054 0.064 3.742 3.755 pw_transfer 1331 11.6 0.066 0.075 3.706 3.739 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.600 3.635 density_rs2pw 110 9.6 0.004 0.004 2.978 3.490 grid_integrate_task_list 110 12.3 3.267 3.481 3.267 3.481 multiply_cannon_sync_h2d 24660 15.4 3.200 3.373 3.200 3.373 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.233 3.294 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.275 3.276 multiply_cannon_metrocomm3 24660 15.4 0.036 0.037 1.421 3.261 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.124 3.253 fft_wrap_pw1pw2_140 451 13.1 0.669 0.690 3.110 3.148 qs_ot_get_derivative_diag 47 12.0 0.001 0.003 2.927 2.989 calculate_dm_sparse 110 9.5 0.001 0.001 2.913 2.945 wfi_extrapolate 11 7.9 0.001 0.001 2.944 2.944 hybrid_alltoall_any 4261 16.3 0.120 0.460 2.196 2.892 make_images_data 4110 15.4 0.045 0.049 2.572 2.891 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.776 2.888 rs_pw_transfer 902 11.9 0.010 0.010 2.343 2.843 mp_alltoall_i22 605 13.7 1.634 2.785 1.634 2.785 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.770 2.772 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.751 2.761 cp_fm_diag_elpa_base 48 14.0 2.606 2.664 2.749 2.759 cp_fm_cholesky_invert 11 10.9 2.580 2.588 2.580 2.588 calculate_first_density_matrix 1 7.0 0.001 0.003 2.521 2.526 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.423 2.464 grid_collocate_task_list 110 9.6 2.179 2.439 2.179 2.439 fft3d_ps 1111 14.6 1.085 1.113 2.362 2.381 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.251 2.252 potential_pw2rs 110 12.3 0.012 0.013 2.140 2.149 mp_alltoall_d11v 2046 13.8 1.696 1.988 1.696 1.988 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.676 1.708 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.598 1.704 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.669 1.681 cp_fm_cholesky_decompose 22 10.9 1.629 1.678 1.629 1.678 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.536 1.555 acc_transpose_blocks 24660 15.4 0.102 0.104 1.502 1.527 mp_waitany 13376 13.8 1.062 1.524 1.062 1.524 multiply_cannon_metrocomm4 20550 15.4 0.058 0.062 0.846 1.524 mp_sum_l 6514 12.8 0.847 1.452 0.847 1.452 mp_irecv_dv 62702 16.1 0.746 1.446 0.746 1.446 mp_allgather_i34 2055 14.4 0.458 1.414 0.458 1.414 rs_pw_transfer_RS2PW_140 121 11.5 0.169 0.179 0.811 1.323 qs_env_update_s_mstruct 11 6.9 0.001 0.006 1.200 1.306 jit_kernel_multiply 7 16.2 0.583 1.289 0.583 1.289 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=64.298000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=680.272727, yerr=10.384349 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7d51e9aba56ae0e0080ac2083bd6f96ee184da78_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 820.101120E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66424. MP_Allreduce 9694 649. MP_Sync 52 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.030 55.134 55.135 qs_mol_dyn_low 1 2.0 0.003 0.003 54.910 54.918 qs_forces 11 3.9 0.003 0.003 54.838 54.839 qs_energies 11 4.9 0.001 0.002 51.147 51.153 scf_env_do_scf 11 5.9 0.001 0.002 42.921 42.922 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 35.223 35.223 velocity_verlet 10 3.0 0.001 0.002 31.114 31.118 dbcsr_multiply_generic 2055 12.4 0.106 0.112 22.712 22.837 qs_scf_new_mos 99 7.5 0.001 0.001 20.160 20.220 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.159 20.219 ot_scf_mini 99 9.5 0.002 0.002 18.924 18.981 multiply_cannon 2055 13.4 0.244 0.256 17.369 18.443 multiply_cannon_loop 2055 14.4 0.320 0.333 16.040 16.237 rebuild_ks_matrix 110 8.3 0.000 0.000 11.953 12.009 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.015 11.953 12.008 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.619 10.669 ot_mini 99 10.5 0.001 0.001 10.323 10.375 multiply_cannon_multrec 8220 15.4 3.209 4.426 7.444 8.536 mp_waitall_1 103326 16.6 6.079 7.809 6.079 7.809 init_scf_loop 11 6.9 0.000 0.000 7.652 7.653 qs_ot_get_derivative 99 11.5 0.002 0.006 6.502 6.556 sum_up_and_integrate 110 10.3 0.080 0.081 6.527 6.538 integrate_v_rspace 110 11.3 0.003 0.003 6.447 6.458 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.271 6.291 calculate_rho_elec 110 8.6 0.114 0.115 6.270 6.290 prepare_preconditioner 11 7.9 0.000 0.000 5.990 5.998 make_preconditioner 11 8.9 0.000 0.000 5.990 5.998 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.567 5.642 init_scf_run 11 5.9 0.000 0.001 5.088 5.088 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.087 5.087 dbcsr_mm_accdrv_process 17442 15.9 2.788 3.933 4.104 4.998 qs_ot_get_p 110 10.4 0.001 0.001 4.668 4.728 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 3.163 4.496 make_m2s 4110 13.4 0.039 0.040 4.115 4.387 make_images 4110 14.4 0.638 0.690 3.986 4.260 pw_transfer 1331 11.6 0.066 0.071 3.933 3.941 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.825 3.837 ot_diis_step 99 11.5 0.012 0.012 3.802 3.802 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.722 3.769 apply_single 110 13.6 0.000 0.000 3.722 3.768 grid_integrate_task_list 110 12.3 3.364 3.530 3.364 3.530 density_rs2pw 110 9.6 0.004 0.004 3.019 3.392 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.334 3.338 fft_wrap_pw1pw2_140 451 13.1 0.835 0.846 3.301 3.320 multiply_cannon_sync_h2d 8220 15.4 2.911 3.054 2.911 3.054 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.039 3.039 cp_fm_cholesky_invert 11 10.9 2.900 2.904 2.900 2.904 make_images_data 4110 15.4 0.038 0.043 2.338 2.769 wfi_extrapolate 11 7.9 0.001 0.001 2.724 2.724 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.687 2.688 hybrid_alltoall_any 4261 16.3 0.200 0.864 2.284 2.659 rs_pw_transfer 902 11.9 0.010 0.010 2.181 2.589 grid_collocate_task_list 110 9.6 2.277 2.538 2.277 2.538 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.522 2.530 cp_fm_diag_elpa_base 48 14.0 2.468 2.489 2.520 2.528 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.505 2.507 calculate_dm_sparse 110 9.5 0.001 0.001 2.465 2.495 fft3d_ps 1111 14.6 1.139 1.164 2.370 2.382 calculate_first_density_matrix 1 7.0 0.001 0.001 2.265 2.268 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.159 2.176 potential_pw2rs 110 12.3 0.015 0.015 2.068 2.071 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.775 1.999 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.870 1.895 mp_alltoall_d11v 2046 13.8 1.635 1.879 1.635 1.879 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.747 1.781 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.663 1.668 cp_fm_cholesky_decompose 22 10.9 1.628 1.644 1.628 1.644 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.516 1.643 mp_allgather_i34 2055 14.4 0.515 1.581 0.515 1.581 dbcsr_complete_redistribute 325 12.2 0.560 0.579 1.436 1.521 mp_waitany 9240 13.8 1.059 1.497 1.059 1.497 jit_kernel_multiply 8 15.9 1.010 1.446 1.010 1.446 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.428 1.441 qs_create_task_list 11 7.9 0.000 0.001 1.222 1.326 generate_qs_task_list 11 8.9 0.374 0.441 1.222 1.326 multiply_cannon_metrocomm4 6165 15.4 0.017 0.019 0.481 1.298 mp_irecv_dv 24056 15.7 0.456 1.253 0.456 1.253 rs_pw_transfer_RS2PW_140 121 11.5 0.172 0.184 0.801 1.215 rs_gather_matrices 110 12.3 0.325 0.366 0.925 1.167 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.137 1.164 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=55.135000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=772.545455, yerr=11.949619 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7d51e9aba56ae0e0080ac2083bd6f96ee184da78_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 1.322529E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67104. MP_Allreduce 9672 819. MP_Sync 52 MP_Alltoall 1474 16505187. MP_SendRecv 2310 360267. MP_ISendRecv 2310 360267. MP_Wait 5214 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.033 86.286 86.286 qs_mol_dyn_low 1 2.0 0.003 0.003 85.988 85.998 qs_forces 11 3.9 0.003 0.003 85.915 85.916 qs_energies 11 4.9 0.001 0.001 81.751 81.753 scf_env_do_scf 11 5.9 0.001 0.001 71.824 71.825 velocity_verlet 10 3.0 0.001 0.001 55.571 55.578 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 43.368 43.370 init_scf_loop 11 6.9 0.000 0.000 28.386 28.390 dbcsr_multiply_generic 2055 12.4 0.120 0.126 27.965 28.045 prepare_preconditioner 11 7.9 0.000 0.000 26.385 26.392 make_preconditioner 11 8.9 0.000 0.000 26.385 26.392 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.538 25.860 qs_scf_new_mos 99 7.5 0.001 0.001 25.356 25.403 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.356 25.403 ot_scf_mini 99 9.5 0.002 0.002 23.611 23.649 multiply_cannon 2055 13.4 0.346 0.376 21.253 21.920 multiply_cannon_loop 2055 14.4 0.341 0.344 19.517 19.861 cp_fm_upper_to_full 70 14.2 12.989 18.553 12.989 18.553 rebuild_ks_matrix 110 8.3 0.000 0.001 13.938 13.971 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 13.937 13.971 ot_mini 99 10.5 0.001 0.001 13.209 13.249 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.621 12.651 dbcsr_complete_redistribute 325 12.2 1.028 1.041 7.380 10.854 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.384 9.849 multiply_cannon_multrec 8220 15.4 4.350 4.548 9.505 9.618 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.831 9.254 mp_alltoall_i22 605 13.7 5.457 8.943 5.457 8.943 qs_ot_get_derivative 99 11.5 0.001 0.001 8.718 8.757 mp_waitall_1 84994 16.7 7.221 8.020 7.221 8.020 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.625 7.660 calculate_rho_elec 110 8.6 0.225 0.226 7.624 7.659 sum_up_and_integrate 110 10.3 0.150 0.151 7.219 7.232 integrate_v_rspace 110 11.3 0.004 0.004 7.068 7.080 init_scf_run 11 5.9 0.000 0.001 5.654 5.654 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.654 5.654 make_m2s 4110 13.4 0.043 0.044 5.032 5.512 qs_ot_get_p 110 10.4 0.001 0.001 5.279 5.326 make_images 4110 14.4 0.884 0.934 4.845 5.325 dbcsr_mm_accdrv_process 11614 15.7 3.326 3.655 5.014 5.275 cp_fm_cholesky_invert 11 10.9 5.092 5.097 5.092 5.097 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.616 5.090 apply_single 110 13.6 0.000 0.000 4.616 5.090 multiply_cannon_metrocomm3 8220 15.4 0.018 0.019 4.669 5.007 pw_transfer 1331 11.6 0.075 0.076 4.977 4.980 fft_wrap_pw1pw2 1111 12.6 0.009 0.010 4.859 4.862 ot_diis_step 99 11.5 0.015 0.015 4.470 4.470 fft_wrap_pw1pw2_140 451 13.1 1.338 1.343 4.247 4.251 multiply_cannon_sync_h2d 8220 15.4 3.949 3.956 3.949 3.956 qs_ot_p2m_diag 48 11.0 0.151 0.155 3.819 3.826 grid_integrate_task_list 110 12.3 3.656 3.713 3.656 3.713 density_rs2pw 110 9.6 0.004 0.004 3.663 3.700 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.677 3.678 hybrid_alltoall_any 4261 16.3 0.257 0.552 2.802 3.425 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.386 3.386 make_images_data 4110 15.4 0.041 0.044 2.698 3.296 wfi_extrapolate 11 7.9 0.001 0.001 3.288 3.288 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.785 3.235 calculate_dm_sparse 110 9.5 0.001 0.001 3.105 3.132 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.823 2.828 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.815 2.815 cp_fm_diag_elpa_base 48 14.0 2.275 2.477 2.814 2.814 fft3d_ps 1111 14.6 1.306 1.315 2.787 2.794 grid_collocate_task_list 110 9.6 2.631 2.653 2.631 2.653 potential_pw2rs 110 12.3 0.021 0.021 2.363 2.365 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.333 2.347 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.310 2.334 qs_env_update_s_mstruct 11 6.9 0.001 0.001 2.196 2.260 calculate_first_density_matrix 1 7.0 0.000 0.000 2.246 2.247 rs_pw_transfer 902 11.9 0.010 0.011 2.221 2.240 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.089 2.200 mp_alltoall_d11v 2046 13.8 1.865 1.973 1.865 1.973 cp_fm_cholesky_decompose 22 10.9 1.948 1.968 1.948 1.968 qs_create_task_list 11 7.9 0.000 0.001 1.895 1.947 generate_qs_task_list 11 8.9 0.731 0.786 1.894 1.947 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.916 1.920 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.723 1.752 jit_kernel_multiply 10 15.2 1.487 1.730 1.487 1.730 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=86.286000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1192.363636, yerr=53.129811 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7d51e9aba56ae0e0080ac2083bd6f96ee184da78_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 618.446848E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175954870160 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 57910. MP_Allreduce 10979 803. MP_Sync 87 MP_Alltoall 2226 2510114. MP_SendRecv 24320 18752. MP_ISendRecv 24320 18752. MP_Wait 42476 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.037 206.133 206.145 qs_mol_dyn_low 1 2.0 0.003 0.006 205.741 205.753 qs_forces 11 3.9 0.005 0.005 205.650 205.660 qs_energies 11 4.9 0.002 0.003 200.053 200.077 scf_env_do_scf 11 5.9 0.001 0.002 183.462 183.475 scf_env_do_scf_inner_loop 117 6.6 0.003 0.010 162.970 162.980 dbcsr_multiply_generic 2507 12.6 0.178 0.184 123.751 124.255 velocity_verlet 10 3.0 0.001 0.002 123.529 123.536 qs_scf_new_mos 117 7.6 0.001 0.001 122.975 123.283 qs_scf_loop_do_ot 117 8.6 0.001 0.001 122.975 123.282 ot_scf_mini 117 9.6 0.003 0.003 116.363 116.647 multiply_cannon 2507 13.6 0.240 0.250 100.806 102.184 multiply_cannon_loop 2507 14.6 2.051 2.111 98.657 100.031 ot_mini 117 10.6 0.001 0.001 65.050 65.410 multiply_cannon_multrec 60168 15.6 33.636 35.773 41.702 43.828 qs_ot_get_derivative 117 11.6 0.002 0.006 40.240 40.545 rebuild_ks_matrix 128 8.3 0.001 0.001 33.735 34.063 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.018 33.734 34.062 mp_waitall_1 267128 16.5 27.724 31.248 27.724 31.248 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.288 30.578 multiply_cannon_sync_h2d 60168 15.6 27.792 29.983 27.792 29.983 qs_ot_get_p 128 10.4 0.001 0.001 29.407 29.792 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.415 25.104 apply_single 128 13.6 0.001 0.001 24.415 25.104 ot_diis_step 117 11.6 0.007 0.008 24.576 24.578 qs_ot_p2m_diag 83 11.4 0.079 0.092 22.946 23.031 init_scf_loop 11 6.9 0.000 0.001 20.417 20.419 cp_dbcsr_syevd 83 12.4 0.005 0.005 20.402 20.404 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.063 19.261 multiply_cannon_metrocomm3 60168 15.6 0.114 0.119 15.640 17.918 cp_fm_diag_elpa 83 13.4 0.000 0.000 17.431 17.463 cp_fm_diag_elpa_base 83 14.4 17.363 17.409 17.427 17.460 prepare_preconditioner 11 7.9 0.000 0.000 15.825 15.875 make_preconditioner 11 8.9 0.000 0.000 15.825 15.875 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.062 15.238 sum_up_and_integrate 128 10.3 0.090 0.106 14.219 14.238 integrate_v_rspace 128 11.3 0.004 0.005 14.129 14.151 make_m2s 5014 13.6 0.104 0.113 13.623 13.920 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.675 13.843 calculate_rho_elec 128 8.7 0.046 0.065 13.675 13.842 make_images 5014 14.6 0.401 0.420 13.442 13.745 init_scf_run 11 5.9 0.000 0.001 12.417 12.418 scf_env_initial_rho_setup 11 6.9 0.001 0.002 12.417 12.418 density_rs2pw 128 9.7 0.006 0.007 7.051 10.225 mp_sum_l 7870 13.0 8.053 9.385 8.053 9.385 wfi_extrapolate 11 7.9 0.001 0.001 9.163 9.163 cp_fm_cholesky_invert 11 10.9 9.061 9.069 9.061 9.069 rs_pw_transfer 1046 11.9 0.016 0.018 5.746 8.933 calculate_dm_sparse 128 9.5 0.001 0.001 8.425 8.535 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.995 8.190 dbcsr_mm_accdrv_process 124484 16.2 3.059 3.241 7.624 8.113 pw_transfer 1547 11.6 0.074 0.090 7.770 7.947 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.838 7.941 fft_wrap_pw1pw2 1291 12.7 0.011 0.013 7.567 7.744 grid_integrate_task_list 128 12.3 7.044 7.519 7.044 7.519 make_images_data 5014 15.6 0.067 0.072 6.656 7.494 multiply_cannon_metrocomm1 60168 15.6 0.088 0.092 5.673 7.351 hybrid_alltoall_any 5200 16.5 0.290 2.258 5.857 7.126 fft_wrap_pw1pw2_140 523 13.2 1.272 1.315 6.630 6.811 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.646 6.657 mp_waitany 16020 13.9 2.683 5.948 2.683 5.948 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.803 5.905 mp_alltoall_d11v 2415 14.1 4.206 5.845 4.206 5.845 grid_collocate_task_list 128 9.7 4.558 5.795 4.558 5.795 fft3d_ps 1291 14.7 2.144 2.784 5.338 5.652 rs_pw_transfer_RS2PW_140 139 11.5 0.278 0.298 2.113 5.319 potential_pw2rs 128 12.3 0.009 0.010 4.617 4.632 cp_fm_cholesky_decompose 22 10.9 4.598 4.612 4.598 4.612 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=206.145000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=579.545455, yerr=6.880323 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7d51e9aba56ae0e0080ac2083bd6f96ee184da78_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 812.191744E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100943E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834845264 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 58220. MP_Allreduce 11003 967. MP_Sync 87 MP_Alltoall 1969 5618855. MP_SendRecv 12032 47072. MP_ISendRecv 12032 47072. MP_Wait 25916 MP_ISend 11748 212467. MP_IRecv 11748 212467. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.033 188.591 188.593 qs_mol_dyn_low 1 2.0 0.003 0.004 188.213 188.228 qs_forces 11 3.9 0.005 0.005 188.125 188.127 qs_energies 11 4.9 0.001 0.002 181.332 181.342 scf_env_do_scf 11 5.9 0.001 0.001 164.989 164.999 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 132.339 132.342 velocity_verlet 10 3.0 0.001 0.001 118.911 118.913 dbcsr_multiply_generic 2507 12.6 0.189 0.194 96.497 97.661 qs_scf_new_mos 117 7.6 0.001 0.001 92.623 93.067 qs_scf_loop_do_ot 117 8.6 0.001 0.001 92.622 93.066 ot_scf_mini 117 9.6 0.004 0.004 87.834 88.319 multiply_cannon 2507 13.6 0.476 0.532 76.864 80.826 multiply_cannon_loop 2507 14.6 1.240 1.277 74.009 76.507 ot_mini 117 10.6 0.001 0.001 49.271 49.743 mp_waitall_1 214728 16.6 24.058 37.314 24.058 37.314 multiply_cannon_multrec 30084 15.6 22.084 26.713 31.687 36.740 rebuild_ks_matrix 128 8.3 0.001 0.001 33.055 33.633 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.019 33.054 33.633 init_scf_loop 11 6.9 0.000 0.000 32.562 32.563 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.719 30.252 multiply_cannon_metrocomm3 30084 15.6 0.090 0.095 15.524 28.552 prepare_preconditioner 11 7.9 0.000 0.000 28.200 28.283 make_preconditioner 11 8.9 0.000 0.000 28.200 28.283 qs_ot_get_derivative 117 11.6 0.001 0.002 27.415 27.890 make_full_inverse_cholesky 11 9.9 0.000 0.000 26.891 27.460 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 21.950 23.010 apply_single 128 13.6 0.001 0.001 21.950 23.009 ot_diis_step 117 11.6 0.014 0.015 21.693 21.695 multiply_cannon_sync_h2d 30084 15.6 19.368 21.621 19.368 21.621 qs_ot_get_p 128 10.4 0.001 0.001 20.502 21.038 cp_fm_cholesky_invert 11 10.9 16.514 16.526 16.514 16.526 qs_ot_p2m_diag 83 11.4 0.187 0.216 15.801 15.835 make_m2s 5014 13.6 0.088 0.094 14.008 15.172 make_images 5014 14.6 1.157 1.350 13.802 14.967 cp_dbcsr_syevd 83 12.4 0.005 0.006 14.668 14.669 sum_up_and_integrate 128 10.3 0.116 0.132 14.618 14.647 integrate_v_rspace 128 11.3 0.004 0.004 14.502 14.535 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.942 13.967 calculate_rho_elec 128 8.7 0.088 0.104 13.941 13.966 cp_fm_diag_elpa 83 13.4 0.000 0.000 11.551 11.581 cp_fm_diag_elpa_base 83 14.4 11.300 11.412 11.546 11.576 init_scf_run 11 5.9 0.000 0.001 11.569 11.570 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.569 11.570 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 10.798 11.136 multiply_cannon_metrocomm4 27577 15.6 0.102 0.117 3.745 10.476 mp_irecv_dv 69486 16.3 3.546 10.085 3.546 10.085 make_images_data 5014 15.6 0.065 0.074 8.531 9.967 density_rs2pw 128 9.7 0.006 0.007 7.340 9.967 dbcsr_mm_accdrv_process 62242 16.2 4.483 5.236 9.058 9.644 hybrid_alltoall_any 5200 16.5 0.343 1.510 7.236 9.320 pw_transfer 1547 11.6 0.086 0.096 8.716 8.767 fft_wrap_pw1pw2 1291 12.7 0.011 0.011 8.491 8.547 wfi_extrapolate 11 7.9 0.001 0.001 8.375 8.375 rs_pw_transfer 1046 11.9 0.014 0.017 5.593 8.241 fft_wrap_pw1pw2_140 523 13.2 1.337 1.361 7.529 7.605 grid_integrate_task_list 128 12.3 7.173 7.495 7.173 7.495 cp_fm_cholesky_decompose 22 10.9 6.800 6.895 6.800 6.895 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.125 6.864 calculate_dm_sparse 128 9.5 0.001 0.001 6.440 6.591 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.104 6.113 grid_collocate_task_list 128 9.7 4.705 5.846 4.705 5.846 fft3d_ps 1291 14.7 2.813 2.987 5.798 5.846 mp_sum_l 7870 13.0 3.841 5.655 3.841 5.655 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.276 5.432 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.296 5.355 mp_waitany 11748 13.9 2.458 5.166 2.458 5.166 potential_pw2rs 128 12.3 0.015 0.017 4.877 4.894 rs_pw_transfer_RS2PW_140 139 11.5 0.350 0.381 2.072 4.705 mp_alltoall_d11v 2415 14.1 4.063 4.646 4.063 4.646 mp_allgather_i34 2507 14.6 1.427 4.484 1.427 4.484 dbcsr_complete_redistribute 395 12.7 0.770 0.850 3.083 3.935 mp_sum_d 4470 12.1 2.562 3.926 2.562 3.926 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=188.593000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=774.454545, yerr=1.724879 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7d51e9aba56ae0e0080ac2083bd6f96ee184da78_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410023282688 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444707676160 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796579E+12 0.0% 0.0% 100.0% flops max/rank 2.906045E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500928 0.0% 0.0% 100.0% number of processed stacks 3951168 0.0% 0.0% 100.0% average stack size 0.0 0.0 1697.1 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 935.665664E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1033760 MPI messages size (bytes): total size 2.695213E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.607194E+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 279168 36591108096 131072 < size <= 4194304 654272 987691483136 4194304 < size <= 16777216 65184 925172905552 16777216 < size 28448 745747251200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3997 58284. MP_Allreduce 10990 1007. MP_Sync 86 MP_Alltoall 1700 9383497. MP_SendRecv 7874 75008. MP_ISendRecv 7874 75008. MP_Wait 21654 MP_ISend 11660 275234. MP_IRecv 11660 275234. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.031 174.380 174.381 qs_mol_dyn_low 1 2.0 0.003 0.003 173.989 174.003 qs_forces 11 3.9 0.004 0.005 173.881 173.887 qs_energies 11 4.9 0.001 0.002 167.212 167.222 scf_env_do_scf 11 5.9 0.001 0.003 151.611 151.612 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 116.821 116.822 velocity_verlet 10 3.0 0.001 0.002 111.838 111.840 dbcsr_multiply_generic 2485 12.5 0.180 0.186 80.992 82.106 qs_scf_new_mos 116 7.6 0.001 0.001 78.889 79.244 qs_scf_loop_do_ot 116 8.6 0.001 0.001 78.888 79.243 ot_scf_mini 116 9.6 0.003 0.004 74.767 75.139 multiply_cannon 2485 13.5 0.497 0.518 61.118 65.435 multiply_cannon_loop 2485 14.5 0.846 0.877 58.099 60.468 ot_mini 116 10.6 0.001 0.001 42.102 42.476 init_scf_loop 11 6.9 0.000 0.001 34.691 34.693 mp_waitall_1 169034 16.6 25.110 34.557 25.110 34.557 rebuild_ks_matrix 127 8.3 0.001 0.001 30.622 31.135 qs_ks_build_kohn_sham_matrix 127 9.3 0.018 0.022 30.621 31.135 prepare_preconditioner 11 7.9 0.000 0.000 30.632 30.702 make_preconditioner 11 8.9 0.000 0.000 30.632 30.702 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.238 29.704 qs_ks_update_qs_env 127 7.6 0.001 0.001 27.544 28.011 multiply_cannon_multrec 19880 15.5 13.366 16.170 21.999 24.868 multiply_cannon_metrocomm3 19880 15.5 0.057 0.062 15.100 24.214 qs_ot_get_derivative 116 11.6 0.002 0.002 22.368 22.736 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.875 20.988 apply_single 127 13.6 0.001 0.001 19.875 20.988 ot_diis_step 116 11.6 0.017 0.018 19.632 19.633 qs_ot_get_p 127 10.4 0.001 0.001 18.100 18.565 make_m2s 4970 13.5 0.080 0.087 14.973 15.820 multiply_cannon_sync_h2d 19880 15.5 14.135 15.727 14.135 15.727 make_images 4970 14.5 1.169 1.257 14.744 15.584 sum_up_and_integrate 127 10.3 0.132 0.143 14.517 14.546 integrate_v_rspace 127 11.3 0.004 0.005 14.385 14.411 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.197 14.229 calculate_rho_elec 127 8.7 0.129 0.144 14.196 14.229 cp_fm_cholesky_invert 11 10.9 14.201 14.210 14.201 14.210 qs_ot_p2m_diag 82 11.4 0.262 0.269 13.955 13.963 cp_dbcsr_syevd 82 12.4 0.005 0.006 12.965 12.966 make_images_data 4970 15.5 0.059 0.066 9.398 10.690 init_scf_run 11 5.9 0.000 0.001 10.607 10.608 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.607 10.607 hybrid_alltoall_any 5155 16.4 0.429 1.956 8.127 9.931 cp_fm_diag_elpa 82 13.4 0.000 0.001 9.863 9.883 cp_fm_diag_elpa_base 82 14.4 9.466 9.617 9.860 9.879 multiply_cannon_metrocomm4 17395 15.5 0.061 0.071 3.427 9.165 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 8.705 8.958 density_rs2pw 127 9.7 0.006 0.006 7.090 8.956 mp_irecv_dv 49801 16.2 3.307 8.918 3.307 8.918 pw_transfer 1535 11.6 0.085 0.102 8.804 8.909 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.582 8.691 dbcsr_mm_accdrv_process 41158 16.2 4.584 5.249 8.087 8.176 fft_wrap_pw1pw2_140 519 13.2 1.404 1.428 7.611 7.736 grid_integrate_task_list 127 12.3 7.230 7.716 7.230 7.716 wfi_extrapolate 11 7.9 0.001 0.001 7.480 7.480 cp_fm_upper_to_full 104 14.8 5.746 7.275 5.746 7.275 cp_fm_cholesky_decompose 22 10.9 7.217 7.245 7.217 7.245 rs_pw_transfer 1038 11.9 0.013 0.014 5.100 6.994 dbcsr_complete_redistribute 393 12.7 1.165 1.192 4.608 6.388 calculate_dm_sparse 127 9.5 0.001 0.001 5.787 5.891 fft3d_ps 1281 14.7 2.718 2.941 5.723 5.807 grid_collocate_task_list 127 9.7 4.873 5.679 4.873 5.679 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.403 5.412 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.423 5.197 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.483 5.183 mp_alltoall_d11v 2401 14.1 4.357 4.925 4.357 4.925 mp_allgather_i34 2485 14.5 1.553 4.797 1.553 4.797 potential_pw2rs 127 12.3 0.020 0.022 4.689 4.707 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.522 4.649 mp_sum_l 7804 13.0 3.093 4.436 3.093 4.436 mp_waitany 11660 13.9 2.334 4.302 2.334 4.302 transfer_fm_to_dbcsr 11 9.9 0.019 0.026 2.375 4.097 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.995 4.014 mp_alltoall_i22 712 14.1 1.939 3.801 1.939 3.801 rs_pw_transfer_RS2PW_140 138 11.5 0.328 0.353 1.887 3.773 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 3.736 3.737 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.574 3.622 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=174.381000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=883.181818, yerr=8.589827 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7d51e9aba56ae0e0080ac2083bd6f96ee184da78_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410024443904 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444712984576 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796586E+12 0.0% 0.0% 100.0% flops max/rank 4.320340E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705502176 0.0% 0.0% 100.0% number of processed stacks 5927808 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.2 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.136771E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1133160 MPI messages size (bytes): total size 2.008142E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.772161E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 315952 35695099904 131072 < size <= 4194304 709496 778939400192 4194304 < size <= 16777216 69840 660837893360 16777216 < size 30480 532676608000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4003 58195. MP_Allreduce 11005 1091. MP_Sync 86 MP_Alltoall 1700 12496381. MP_SendRecv 5842 75008. MP_ISendRecv 5842 75008. MP_Wait 22272 MP_ISend 14840 244848. MP_IRecv 14840 244848. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.026 0.045 186.241 186.242 qs_mol_dyn_low 1 2.0 0.003 0.006 185.791 185.804 qs_forces 11 3.9 0.006 0.023 185.673 185.677 qs_energies 11 4.9 0.002 0.006 178.490 178.504 scf_env_do_scf 11 5.9 0.001 0.003 161.462 161.473 velocity_verlet 10 3.0 0.001 0.003 123.367 123.371 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 115.369 115.370 dbcsr_multiply_generic 2485 12.5 0.185 0.191 78.508 79.155 qs_scf_new_mos 116 7.6 0.001 0.001 78.196 78.495 qs_scf_loop_do_ot 116 8.6 0.001 0.001 78.195 78.494 ot_scf_mini 116 9.6 0.003 0.005 73.744 74.032 multiply_cannon 2485 13.5 0.557 0.593 54.322 57.198 multiply_cannon_loop 2485 14.5 1.167 1.193 50.603 52.061 init_scf_loop 11 6.9 0.001 0.004 45.972 45.973 ot_mini 116 10.6 0.001 0.003 41.655 41.928 prepare_preconditioner 11 7.9 0.000 0.000 41.851 41.885 make_preconditioner 11 8.9 0.000 0.001 41.851 41.885 make_full_inverse_cholesky 11 9.9 0.011 0.023 35.539 40.496 multiply_cannon_multrec 29820 15.5 13.934 19.275 25.850 30.813 rebuild_ks_matrix 127 8.3 0.001 0.001 29.743 29.975 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.022 29.742 29.975 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.784 26.990 mp_waitall_1 146592 16.7 17.130 26.957 17.130 26.957 qs_ot_get_derivative 116 11.6 0.002 0.003 22.156 22.442 make_m2s 4970 13.5 0.095 0.100 19.960 21.200 make_images 4970 14.5 1.924 2.279 19.656 20.896 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 18.961 19.501 apply_single 127 13.6 0.001 0.001 18.961 19.501 ot_diis_step 116 11.6 0.017 0.020 19.365 19.367 qs_ot_get_p 127 10.4 0.001 0.003 18.452 18.760 cp_fm_upper_to_full 104 14.8 11.331 16.681 11.331 16.681 cp_fm_cholesky_invert 11 10.9 16.000 16.009 16.000 16.009 multiply_cannon_metrocomm3 29820 15.5 0.047 0.050 6.490 15.207 sum_up_and_integrate 127 10.3 0.139 0.151 14.661 14.690 integrate_v_rspace 127 11.3 0.004 0.004 14.522 14.555 qs_ot_p2m_diag 82 11.4 0.339 0.385 14.502 14.553 qs_rho_update_rho_low 127 7.7 0.001 0.002 14.436 14.470 calculate_rho_elec 127 8.7 0.172 0.188 14.435 14.470 cp_dbcsr_syevd 82 12.4 0.006 0.008 13.251 13.253 make_images_data 4970 15.5 0.062 0.066 10.805 12.819 dbcsr_complete_redistribute 393 12.7 1.497 1.621 8.898 12.579 multiply_cannon_sync_h2d 29820 15.5 11.682 12.420 11.682 12.420 dbcsr_mm_accdrv_process 61748 16.2 7.439 8.321 11.491 11.973 hybrid_alltoall_any 5155 16.4 0.520 2.202 9.691 11.910 copy_fm_to_dbcsr 208 11.6 0.002 0.002 7.524 11.234 init_scf_run 11 5.9 0.000 0.001 11.035 11.036 scf_env_initial_rho_setup 11 6.9 0.001 0.002 11.035 11.036 cp_fm_diag_elpa 82 13.4 0.000 0.001 10.235 10.247 cp_fm_diag_elpa_base 82 14.4 9.297 9.595 10.228 10.238 transfer_fm_to_dbcsr 11 9.9 0.001 0.005 6.289 9.930 qs_ot_get_derivative_diag 76 12.4 0.002 0.003 9.074 9.276 mp_alltoall_i22 712 14.1 5.523 9.206 5.523 9.206 pw_transfer 1535 11.6 0.086 0.102 9.068 9.147 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.843 8.927 density_rs2pw 127 9.7 0.006 0.008 6.904 8.162 fft_wrap_pw1pw2_140 519 13.2 1.551 1.591 7.873 7.970 grid_integrate_task_list 127 12.3 7.408 7.842 7.408 7.842 wfi_extrapolate 11 7.9 0.001 0.004 7.837 7.837 cp_fm_cholesky_decompose 22 10.9 7.472 7.568 7.472 7.568 multiply_cannon_metrocomm4 24850 15.5 0.078 0.090 2.794 7.456 mp_irecv_dv 75445 16.2 2.647 7.187 2.647 7.187 calculate_dm_sparse 127 9.5 0.001 0.001 6.184 6.258 mp_alltoall_d11v 2401 14.1 4.879 6.061 4.879 6.061 rs_pw_transfer 1038 11.9 0.013 0.014 4.693 6.018 fft3d_ps 1281 14.7 2.819 2.896 5.788 5.848 grid_collocate_task_list 127 9.7 5.043 5.729 5.043 5.729 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.542 5.654 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 4.542 4.544 potential_pw2rs 127 12.3 0.023 0.024 4.515 4.528 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.462 4.517 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.297 4.378 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.148 4.206 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=186.242000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1070.181818, yerr=18.575733 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7d51e9aba56ae0e0080ac2083bd6f96ee184da78_performance_tests/19/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 5.909858E+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 1976192 0.0% 0.0% 100.0% average stack size 0.0 0.0 3444.2 marketing flops 145.646636E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 1.480122E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 242592 MPI messages size (bytes): total size 1.341705E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.530704E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 114816 60196651008 4194304 < size <= 16777216 105840 554906419200 16777216 < size 20352 726591885696 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 9003 51. MP_Alltoall 9708 795382. MP_ISend 40388 2101966. MP_IRecv 40388 2101070. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4024 57898. MP_Allreduce 11057 1172. MP_Sync 87 MP_Alltoall 1724 18848449. MP_SendRecv 3870 122880. MP_ISendRecv 3870 122880. MP_Wait 16244 MP_ISend 10760 423501. MP_IRecv 10760 423501. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.030 172.538 172.540 qs_mol_dyn_low 1 2.0 0.003 0.003 172.156 172.168 qs_forces 11 3.9 0.014 0.015 172.036 172.044 qs_energies 11 4.9 0.001 0.002 164.495 164.502 scf_env_do_scf 11 5.9 0.002 0.004 147.109 147.113 velocity_verlet 10 3.0 0.002 0.002 113.672 113.678 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 111.870 111.872 dbcsr_multiply_generic 2527 12.6 0.183 0.196 74.417 74.837 qs_scf_new_mos 118 7.6 0.001 0.001 73.213 73.310 qs_scf_loop_do_ot 118 8.6 0.001 0.001 73.212 73.309 ot_scf_mini 118 9.6 0.004 0.004 68.789 68.861 multiply_cannon 2527 13.6 0.589 0.633 54.606 58.931 multiply_cannon_loop 2527 14.6 0.447 0.463 50.024 50.837 ot_mini 118 10.6 0.001 0.001 39.665 39.734 init_scf_loop 11 6.9 0.000 0.000 35.094 35.097 mp_waitall_1 126780 16.7 26.030 33.246 26.030 33.246 prepare_preconditioner 11 7.9 0.000 0.000 31.196 31.215 make_preconditioner 11 8.9 0.000 0.000 31.196 31.215 rebuild_ks_matrix 129 8.3 0.001 0.001 29.939 30.029 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.019 29.938 30.028 make_full_inverse_cholesky 11 9.9 0.017 0.027 29.166 29.440 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.237 27.329 multiply_cannon_multrec 10108 15.6 10.461 14.378 17.860 21.082 ot_diis_step 118 11.6 0.020 0.020 20.093 20.093 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.624 19.873 apply_single 129 13.6 0.001 0.001 19.624 19.873 qs_ot_get_derivative 118 11.6 0.002 0.003 19.504 19.578 multiply_cannon_metrocomm3 10108 15.6 0.024 0.025 12.417 19.222 make_m2s 5054 13.6 0.067 0.071 16.333 18.784 make_images 5054 14.6 2.328 2.795 16.025 18.472 cp_fm_cholesky_invert 11 10.9 17.820 17.826 17.820 17.826 qs_ot_get_p 129 10.4 0.001 0.001 16.134 16.221 qs_rho_update_rho_low 129 7.7 0.001 0.001 15.271 15.302 calculate_rho_elec 129 8.7 0.258 0.269 15.271 15.301 sum_up_and_integrate 129 10.3 0.183 0.192 14.974 15.022 integrate_v_rspace 129 11.3 0.004 0.004 14.790 14.838 make_images_data 5054 15.6 0.052 0.060 10.003 12.537 qs_ot_p2m_diag 83 11.4 0.495 0.503 12.500 12.517 multiply_cannon_sync_h2d 10108 15.6 11.723 12.462 11.723 12.462 hybrid_alltoall_any 5240 16.5 0.825 3.850 9.858 12.458 cp_dbcsr_syevd 83 12.4 0.005 0.006 11.387 11.388 init_scf_run 11 5.9 0.000 0.001 10.570 10.570 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.569 10.570 pw_transfer 1559 11.6 0.086 0.096 9.422 9.442 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 9.195 9.221 cp_fm_diag_elpa 83 13.4 0.000 0.000 8.453 8.465 cp_fm_diag_elpa_base 83 14.4 8.222 8.296 8.449 8.461 fft_wrap_pw1pw2_140 527 13.2 1.930 1.960 8.170 8.194 grid_integrate_task_list 129 12.3 7.843 8.122 7.843 8.122 cp_fm_cholesky_decompose 22 10.9 7.803 7.926 7.803 7.926 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 7.755 7.807 density_rs2pw 129 9.7 0.006 0.006 6.811 7.696 dbcsr_mm_accdrv_process 20926 16.1 2.744 3.530 7.022 7.684 multiply_cannon_metrocomm1 10108 15.6 0.028 0.029 4.403 7.501 wfi_extrapolate 11 7.9 0.001 0.001 7.473 7.473 mp_allgather_i34 2527 14.6 2.772 7.231 2.772 7.231 calculate_dm_sparse 129 9.5 0.001 0.001 6.143 6.248 grid_collocate_task_list 129 9.7 5.440 6.114 5.440 6.114 fft3d_ps 1301 14.7 2.789 2.858 5.646 5.681 dbcsr_complete_redistribute 395 12.7 2.116 2.151 4.952 5.353 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.283 5.283 mp_alltoall_d11v 2423 14.1 4.756 5.184 4.756 5.184 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.157 5.167 rs_pw_transfer 1054 12.0 0.012 0.013 4.055 4.951 potential_pw2rs 129 12.3 0.026 0.026 4.362 4.376 multiply_cannon_metrocomm4 7581 15.6 0.024 0.027 1.863 4.254 mp_irecv_dv 29086 15.9 1.827 4.181 1.827 4.181 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.111 4.134 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.564 3.879 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.698 3.734 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.616 3.622 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.242 3.547 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.441 3.538 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 3.484 3.509 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=172.540000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1383.545455, yerr=36.532336 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7d51e9aba56ae0e0080ac2083bd6f96ee184da78_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456039424 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613072052224 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176077312 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176077312 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.233020E+12 0.0% 0.0% 100.0% flops max/rank 11.786061E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383904 0.0% 0.0% 100.0% number of processed stacks 1980288 0.0% 0.0% 100.0% average stack size 0.0 0.0 3437.1 marketing flops 145.650931E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 3.101835E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101160 MPI messages size (bytes): total size 1.144970E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.318403E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 45648 35433480192 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592466352 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 58564. MP_Allreduce 11104 1511. MP_Sync 88 MP_Alltoall 1724 36993632. MP_SendRecv 1806 218624. MP_ISendRecv 1806 218624. MP_Wait 9876 MP_ISend 6456 1080169. MP_IRecv 6456 1080169. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.024 0.048 296.171 296.171 qs_mol_dyn_low 1 2.0 0.003 0.003 295.650 295.663 qs_forces 11 3.9 0.005 0.005 295.527 295.531 qs_energies 11 4.9 0.005 0.005 286.263 286.274 scf_env_do_scf 11 5.9 0.001 0.001 263.790 263.804 velocity_verlet 10 3.0 0.001 0.001 213.630 213.640 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 140.722 140.724 init_scf_loop 11 6.9 0.000 0.000 122.814 122.817 prepare_preconditioner 11 7.9 0.000 0.000 117.780 117.807 make_preconditioner 11 8.9 0.000 0.000 117.780 117.807 make_full_inverse_cholesky 11 9.9 0.036 0.039 94.460 114.975 qs_scf_new_mos 118 7.6 0.001 0.001 89.662 89.784 qs_scf_loop_do_ot 118 8.6 0.001 0.001 89.661 89.783 ot_scf_mini 118 9.6 0.004 0.004 84.867 84.909 dbcsr_multiply_generic 2529 12.6 0.216 0.229 83.468 84.041 cp_fm_upper_to_full 106 14.8 52.638 74.976 52.638 74.976 multiply_cannon 2529 13.6 0.708 0.737 58.755 59.264 multiply_cannon_loop 2529 14.6 0.477 0.489 55.139 56.439 ot_mini 118 10.6 0.001 0.001 44.188 44.225 dbcsr_complete_redistribute 397 12.7 4.015 4.088 28.899 41.885 rebuild_ks_matrix 129 8.3 0.001 0.001 38.540 38.587 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.018 38.539 38.587 copy_fm_to_dbcsr 210 11.7 0.002 0.002 25.555 38.549 transfer_fm_to_dbcsr 11 9.9 0.031 0.033 23.273 36.084 qs_ks_update_qs_env 129 7.6 0.001 0.001 35.593 35.639 mp_alltoall_i22 720 14.1 21.051 34.165 21.051 34.165 cp_fm_cholesky_invert 11 10.9 32.962 32.968 32.962 32.968 mp_waitall_1 104580 16.8 28.200 32.082 28.200 32.082 qs_ot_get_p 129 10.4 0.001 0.001 25.159 25.174 qs_ot_get_derivative 118 11.6 0.002 0.002 24.082 24.123 qs_ot_p2m_diag 84 11.4 0.889 0.895 21.141 21.170 make_m2s 5058 13.6 0.076 0.079 20.151 20.997 qs_rho_update_rho_low 129 7.7 0.001 0.001 20.517 20.540 calculate_rho_elec 129 8.7 0.485 0.486 20.516 20.540 make_images 5058 14.6 3.826 3.914 19.669 20.517 ot_diis_step 118 11.6 0.022 0.023 20.080 20.081 multiply_cannon_metrocomm3 10116 15.6 0.023 0.025 18.775 19.975 sum_up_and_integrate 129 10.3 0.324 0.326 19.753 19.837 integrate_v_rspace 129 11.3 0.004 0.004 19.428 19.512 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.294 19.498 apply_single 129 13.6 0.001 0.001 19.294 19.498 cp_dbcsr_syevd 84 12.4 0.006 0.006 19.419 19.420 multiply_cannon_multrec 10116 15.6 10.543 12.304 18.007 18.090 cp_fm_diag_elpa 84 13.4 0.000 0.000 16.250 16.251 cp_fm_diag_elpa_base 84 14.4 11.826 13.446 16.246 16.247 multiply_cannon_sync_h2d 10116 15.6 15.783 15.795 15.783 15.795 hybrid_alltoall_any 5245 16.5 1.311 3.064 10.885 12.764 init_scf_run 11 5.9 0.000 0.001 12.434 12.434 scf_env_initial_rho_setup 11 6.9 0.036 0.039 12.434 12.434 make_images_data 5058 15.6 0.059 0.065 10.707 12.401 pw_transfer 1559 11.6 0.094 0.096 11.320 11.324 fft_wrap_pw1pw2 1301 12.7 0.011 0.012 11.082 11.087 fft_wrap_pw1pw2_140 527 13.2 3.091 3.108 9.859 9.870 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.441 9.476 mp_alltoall_d11v 2429 14.1 8.348 9.373 8.348 9.373 wfi_extrapolate 11 7.9 0.001 0.001 9.106 9.106 dbcsr_mm_accdrv_process 20934 16.1 3.875 5.650 7.223 9.042 grid_integrate_task_list 129 12.3 8.650 8.814 8.650 8.814 cp_fm_cholesky_decompose 22 10.9 8.690 8.719 8.690 8.719 density_rs2pw 129 9.7 0.005 0.006 7.931 8.061 qs_energies_init_hamiltonians 11 5.9 0.014 0.023 7.985 7.986 calculate_dm_sparse 129 9.5 0.001 0.001 6.655 6.756 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.456 6.506 grid_collocate_task_list 129 9.7 6.411 6.442 6.411 6.442 rs_scatter_matrices 140 9.7 3.751 4.816 6.143 6.368 fft3d_ps 1301 14.7 2.865 2.874 6.127 6.146 copy_dbcsr_to_fm 187 11.8 0.004 0.004 5.983 6.048 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=296.171000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2777.181818, yerr=163.668078 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7d51e9aba56ae0e0080ac2083bd6f96ee184da78_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 1.239204E+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 6799628. 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 0.023 0.040 84.764 84.765 qs_energies 1 2.0 0.000 0.000 84.274 84.280 ls_scf 1 3.0 0.000 0.002 83.371 83.378 dbcsr_multiply_generic 111 6.7 0.016 0.022 72.428 72.572 multiply_cannon 111 7.7 0.018 0.021 55.786 57.049 multiply_cannon_loop 111 8.7 0.207 0.218 52.369 53.709 ls_scf_main 1 4.0 0.000 0.002 52.118 52.119 density_matrix_trs4 2 5.0 0.002 0.003 46.618 46.708 ls_scf_init_scf 1 4.0 0.000 0.000 28.224 28.226 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.137 27.190 mp_waitall_1 11031 10.9 22.336 25.248 22.336 25.248 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 24.949 24.967 multiply_cannon_multrec 2664 9.7 8.177 8.794 15.484 17.059 multiply_cannon_sync_h2d 2664 9.7 13.582 15.257 13.582 15.257 make_m2s 222 7.7 0.008 0.011 12.962 13.526 make_images 222 8.7 0.099 0.110 12.940 13.506 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 9.620 12.324 make_images_data 222 9.7 0.004 0.005 7.565 8.112 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.488 7.925 dbcsr_mm_accdrv_process 4760 10.4 0.511 0.606 6.925 7.841 hybrid_alltoall_any 227 10.6 0.216 1.828 6.572 7.751 dbcsr_mm_accdrv_process_sort 4760 11.4 6.214 7.078 6.214 7.078 calculate_norms 4752 9.8 5.513 6.107 5.513 6.107 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.048 5.190 mp_sum_l 807 5.4 3.144 4.749 3.144 4.749 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.079 3.659 mp_irecv_dv 6231 10.9 2.063 3.636 2.063 3.636 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.374 3.598 make_images_sizes 222 9.7 0.000 0.000 0.707 3.349 mp_alltoall_i44 222 10.7 0.707 3.349 0.707 3.349 arnoldi_extremal 4 6.8 0.000 0.000 3.209 3.235 arnoldi_normal_ev 4 7.8 0.026 0.038 3.209 3.235 build_subspace 16 8.4 0.009 0.012 3.088 3.091 ls_scf_post 1 4.0 0.000 0.002 3.028 3.035 ls_scf_store_result 1 5.0 0.000 0.000 2.851 2.889 dbcsr_special_finalize 555 9.7 0.005 0.006 2.337 2.734 dbcsr_merge_single_wm 555 10.7 0.453 0.591 2.329 2.725 make_images_pack 222 9.7 2.207 2.626 2.209 2.627 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.304 2.563 dbcsr_matrix_vector_mult_local 304 10.0 2.064 2.474 2.066 2.476 dbcsr_sort_data 658 11.4 2.133 2.461 2.133 2.461 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.246 2.320 buffer_matrices_ensure_size 222 8.7 1.744 2.039 1.744 2.039 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.762 1.764 rebuild_ks_matrix 3 7.3 0.000 0.000 1.752 1.755 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.005 1.752 1.755 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=84.765000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1098.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7d51e9aba56ae0e0080ac2083bd6f96ee184da78_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 2.081841E+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 0.026 0.058 90.960 90.961 qs_energies 1 2.0 0.000 0.000 90.446 90.449 ls_scf 1 3.0 0.000 0.000 89.025 89.029 dbcsr_multiply_generic 111 6.7 0.016 0.017 75.164 75.501 multiply_cannon 111 7.7 0.028 0.037 53.345 56.433 ls_scf_main 1 4.0 0.000 0.000 54.725 54.731 multiply_cannon_loop 111 8.7 0.116 0.124 50.077 52.689 density_matrix_trs4 2 5.0 0.002 0.003 49.065 49.245 ls_scf_init_scf 1 4.0 0.000 0.000 30.786 30.787 mp_waitall_1 9105 10.9 21.146 30.581 21.146 30.581 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.602 29.701 multiply_cannon_multrec 1332 9.7 13.144 17.198 22.358 27.582 matrix_sqrt_Newton_Schulz 2 6.5 0.002 0.003 27.189 27.210 multiply_cannon_metrocomm3 1332 9.7 0.006 0.008 11.816 21.473 make_m2s 222 7.7 0.006 0.007 15.302 16.017 make_images 222 8.7 1.573 1.922 15.272 15.986 dbcsr_mm_accdrv_process 4041 10.4 0.298 0.449 8.809 10.361 dbcsr_mm_accdrv_process_sort 4041 11.4 8.399 9.913 8.399 9.913 make_images_data 222 9.7 0.004 0.004 8.776 9.701 hybrid_alltoall_any 227 10.6 0.523 2.428 8.196 9.255 multiply_cannon_metrocomm4 1221 9.7 0.006 0.009 3.249 8.076 mp_irecv_dv 3311 11.0 3.229 8.021 3.229 8.021 mp_sum_l 807 5.4 5.202 7.609 5.202 7.609 calculate_norms 2376 9.8 6.003 6.802 6.003 6.802 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.002 6.271 multiply_cannon_sync_h2d 1332 9.7 4.837 5.976 4.837 5.976 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.999 5.208 arnoldi_extremal 4 6.8 0.000 0.000 4.580 4.606 arnoldi_normal_ev 4 7.8 0.001 0.005 4.580 4.606 build_subspace 16 8.4 0.014 0.021 4.322 4.325 ls_scf_post 1 4.0 0.000 0.000 3.513 3.517 ls_scf_store_result 1 5.0 0.000 0.000 3.193 3.337 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.102 3.336 dbcsr_matrix_vector_mult_local 304 10.0 2.724 3.205 2.726 3.207 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.572 2.663 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 1.191 2.485 make_images_pack 222 9.7 2.025 2.418 2.027 2.420 mp_allgather_i34 111 8.7 0.970 2.419 0.970 2.419 dbcsr_sort_data 436 11.2 1.854 2.105 1.854 2.105 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.868 1.871 rebuild_ks_matrix 3 7.3 0.000 0.000 1.855 1.857 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.855 1.857 dbcsr_data_new 4174 10.1 1.613 1.852 1.613 1.852 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=90.961000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1729.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7d51e9aba56ae0e0080ac2083bd6f96ee184da78_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 2.764124E+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 0.031 0.048 94.274 94.276 qs_energies 1 2.0 0.000 0.000 93.701 93.706 ls_scf 1 3.0 0.000 0.001 92.188 92.193 dbcsr_multiply_generic 111 6.7 0.016 0.017 77.033 77.321 ls_scf_main 1 4.0 0.000 0.002 57.788 57.794 multiply_cannon 111 7.7 0.042 0.095 53.112 56.787 multiply_cannon_loop 111 8.7 0.100 0.105 49.574 53.602 density_matrix_trs4 2 5.0 0.002 0.006 51.866 52.070 mp_waitall_1 7281 11.0 24.231 34.343 24.231 34.343 ls_scf_init_scf 1 4.0 0.000 0.001 30.820 30.822 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.653 29.712 matrix_sqrt_Newton_Schulz 2 6.5 0.004 0.006 27.231 27.243 multiply_cannon_multrec 888 9.7 12.564 15.197 21.094 24.329 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.271 23.596 make_m2s 222 7.7 0.006 0.007 17.181 18.467 make_images 222 8.7 1.971 2.281 17.143 18.427 hybrid_alltoall_any 227 10.6 0.622 2.861 9.487 10.988 make_images_data 222 9.7 0.003 0.004 9.807 10.842 dbcsr_mm_accdrv_process 3754 10.4 0.255 0.453 8.052 9.260 mp_sum_l 807 5.4 5.167 9.004 5.167 9.004 dbcsr_mm_accdrv_process_sort 3754 11.4 7.671 8.807 7.671 8.807 multiply_cannon_sync_h2d 888 9.7 6.073 7.321 6.073 7.321 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.020 7.185 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.460 7.156 mp_irecv_dv 2335 11.1 2.444 7.114 2.444 7.114 multiply_cannon_metrocomm1 888 9.7 0.002 0.003 3.776 6.659 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.889 5.104 arnoldi_extremal 4 6.8 0.000 0.000 4.997 5.015 arnoldi_normal_ev 4 7.8 0.001 0.005 4.997 5.015 build_subspace 16 8.4 0.014 0.020 4.700 4.706 calculate_norms 1584 9.8 4.250 4.595 4.250 4.595 mp_allgather_i34 111 8.7 1.386 3.843 1.386 3.843 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.389 3.721 ls_scf_post 1 4.0 0.000 0.001 3.580 3.585 dbcsr_matrix_vector_mult_local 304 10.0 2.993 3.564 2.995 3.566 ls_scf_store_result 1 5.0 0.000 0.000 3.325 3.404 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.827 2.937 make_images_sizes 222 9.7 0.000 0.000 1.063 2.279 mp_alltoall_i44 222 10.7 1.063 2.278 1.063 2.278 dbcsr_sort_data 325 11.1 1.867 2.188 1.867 2.188 make_images_pack 222 9.7 1.819 2.109 1.822 2.111 dbcsr_data_release 9322 10.9 1.336 2.005 1.336 2.005 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.903 1.905 rebuild_ks_matrix 3 7.3 0.000 0.000 1.885 1.887 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.885 1.887 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=94.276000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2144.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7d51e9aba56ae0e0080ac2083bd6f96ee184da78_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 3.303657E+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 0.038 0.055 96.978 96.979 qs_energies 1 2.0 0.000 0.000 96.330 96.335 ls_scf 1 3.0 0.000 0.001 94.492 94.498 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.406 78.663 ls_scf_main 1 4.0 0.000 0.001 58.567 58.569 multiply_cannon 111 7.7 0.057 0.127 51.723 55.948 density_matrix_trs4 2 5.0 0.002 0.005 52.478 52.608 multiply_cannon_loop 111 8.7 0.114 0.124 46.726 49.539 ls_scf_init_scf 1 4.0 0.000 0.001 32.688 32.690 ls_scf_init_matrix_S 1 5.0 0.000 0.001 31.472 31.551 mp_waitall_1 6369 11.0 22.927 30.186 22.927 30.186 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 28.964 28.976 multiply_cannon_multrec 1332 9.7 14.122 17.311 21.966 25.419 make_m2s 222 7.7 0.007 0.008 21.133 22.493 make_images 222 8.7 3.138 3.601 21.083 22.445 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.491 17.112 make_images_data 222 9.7 0.004 0.004 11.775 13.401 hybrid_alltoall_any 227 10.6 0.796 3.735 11.206 12.970 dbcsr_mm_accdrv_process 3641 10.4 0.223 0.412 7.486 8.993 dbcsr_mm_accdrv_process_sort 3641 11.4 7.104 8.572 7.104 8.572 mp_sum_l 807 5.4 3.955 7.073 3.955 7.073 multiply_cannon_sync_h2d 1332 9.7 5.506 6.066 5.506 6.066 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.059 6.009 mp_irecv_dv 3229 10.9 2.037 5.928 2.037 5.928 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.005 5.534 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.491 5.181 arnoldi_extremal 4 6.8 0.000 0.000 5.130 5.144 arnoldi_normal_ev 4 7.8 0.001 0.005 5.130 5.144 build_subspace 16 8.4 0.014 0.021 4.791 4.799 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.569 4.748 calculate_norms 2376 9.8 4.217 4.564 4.217 4.564 mp_allgather_i34 111 8.7 2.116 4.315 2.116 4.315 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.524 3.819 dbcsr_matrix_vector_mult_local 304 10.0 3.148 3.640 3.150 3.642 dbcsr_sort_data 658 11.4 3.091 3.424 3.091 3.424 ls_scf_post 1 4.0 0.000 0.001 3.236 3.240 dbcsr_special_finalize 555 9.7 0.006 0.007 2.829 3.157 dbcsr_merge_single_wm 555 10.7 0.537 0.653 2.820 3.149 ls_scf_store_result 1 5.0 0.000 0.000 2.978 3.038 ls_scf_dm_to_ks 2 5.0 0.000 0.003 2.977 3.012 dbcsr_data_release 10477 10.7 1.582 2.390 1.582 2.390 dbcsr_finalize 304 7.8 0.050 0.061 1.808 1.992 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=96.979000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2706.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7d51e9aba56ae0e0080ac2083bd6f96ee184da78_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 4.651459E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265558. MP_Allreduce 3049 15663. MP_Sync 4 MP_Alltoall 47 46208988. MP_SendRecv 45 115200. MP_ISendRecv 45 115200. MP_Wait 528 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.040 0.057 99.512 99.513 qs_energies 1 2.0 0.000 0.000 98.732 98.737 ls_scf 1 3.0 0.000 0.000 96.783 96.794 dbcsr_multiply_generic 111 6.7 0.018 0.019 78.415 78.666 ls_scf_main 1 4.0 0.000 0.000 62.276 62.276 multiply_cannon 111 7.7 0.098 0.202 55.582 60.337 density_matrix_trs4 2 5.0 0.002 0.003 55.138 55.225 multiply_cannon_loop 111 8.7 0.070 0.078 51.092 53.287 mp_waitall_1 5436 11.0 26.970 31.875 26.970 31.875 ls_scf_init_scf 1 4.0 0.000 0.000 30.874 30.879 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.619 29.657 matrix_sqrt_Newton_Schulz 2 6.5 0.002 0.003 27.397 27.409 multiply_cannon_multrec 444 9.7 14.119 16.224 21.181 23.869 make_m2s 222 7.7 0.005 0.005 17.886 20.307 make_images 222 8.7 3.731 4.419 17.824 20.248 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.622 16.472 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 5.940 14.348 make_images_data 222 9.7 0.003 0.004 10.210 12.490 hybrid_alltoall_any 227 10.6 0.790 3.772 9.959 12.371 multiply_cannon_sync_h2d 444 9.7 6.570 8.267 6.570 8.267 dbcsr_mm_accdrv_process 3003 10.4 0.184 0.377 6.754 7.893 dbcsr_mm_accdrv_process_sort 3003 11.4 6.409 7.507 6.409 7.507 mp_allgather_i34 111 8.7 2.679 7.019 2.679 7.019 arnoldi_extremal 4 6.8 0.000 0.000 5.716 5.730 arnoldi_normal_ev 4 7.8 0.002 0.005 5.716 5.730 build_subspace 16 8.4 0.015 0.020 5.330 5.341 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.600 4.756 mp_sum_l 807 5.4 2.991 4.469 2.991 4.469 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.092 4.291 dbcsr_matrix_vector_mult_local 304 10.0 3.622 4.099 3.624 4.101 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.603 3.860 mp_irecv_dv 1241 11.2 1.585 3.821 1.585 3.821 calculate_norms 792 9.8 3.532 3.682 3.532 3.682 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.548 3.663 ls_scf_post 1 4.0 0.000 0.000 3.634 3.639 make_images_sizes 222 9.7 0.000 0.000 1.013 3.557 mp_alltoall_i44 222 10.7 1.013 3.556 1.013 3.556 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.101 3.495 ls_scf_store_result 1 5.0 0.000 0.000 3.421 3.458 dbcsr_finalize 304 7.8 0.062 0.078 2.200 2.319 dbcsr_merge_all 275 8.9 0.473 0.530 2.053 2.155 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.065 2.067 rebuild_ks_matrix 3 7.3 0.000 0.000 2.033 2.034 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.033 2.034 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=99.513000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3652.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7d51e9aba56ae0e0080ac2083bd6f96ee184da78_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 8.693866E+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 0.078 0.100 107.609 107.610 qs_energies 1 2.0 0.000 0.000 106.210 106.232 ls_scf 1 3.0 0.000 0.000 103.276 103.298 dbcsr_multiply_generic 111 6.7 0.024 0.028 77.117 77.241 ls_scf_main 1 4.0 0.000 0.000 65.601 65.602 density_matrix_trs4 2 5.0 0.002 0.003 56.439 56.492 multiply_cannon 111 7.7 0.154 0.273 49.627 51.552 multiply_cannon_loop 111 8.7 0.068 0.070 46.148 47.055 ls_scf_init_scf 1 4.0 0.000 0.000 34.021 34.021 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.486 32.498 matrix_sqrt_Newton_Schulz 2 6.5 0.002 0.004 29.655 29.667 mp_waitall_1 4527 11.1 21.983 25.795 21.983 25.795 make_m2s 222 7.7 0.005 0.005 23.911 24.913 make_images 222 8.7 4.599 4.988 23.804 24.805 multiply_cannon_multrec 444 9.7 17.859 18.492 22.447 23.066 make_images_data 222 9.7 0.003 0.004 13.154 15.591 hybrid_alltoall_any 227 10.6 1.663 3.655 12.943 15.588 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.255 10.776 multiply_cannon_sync_h2d 444 9.7 8.850 8.890 8.850 8.890 arnoldi_extremal 4 6.8 0.000 0.000 7.198 7.210 arnoldi_normal_ev 4 7.8 0.012 0.017 7.198 7.210 build_subspace 16 8.4 0.026 0.036 6.634 6.646 dbcsr_matrix_vector_mult 304 9.0 0.017 0.033 5.308 5.441 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.992 5.241 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.142 5.232 dbcsr_matrix_vector_mult_local 304 10.0 4.874 5.170 4.877 5.172 dbcsr_mm_accdrv_process 1814 10.4 0.231 0.315 4.428 4.558 dbcsr_mm_accdrv_process_sort 1814 11.4 4.128 4.257 4.128 4.257 ls_scf_post 1 4.0 0.000 0.000 3.655 3.676 mp_allgather_i34 111 8.7 1.139 3.564 1.139 3.564 make_images_sizes 222 9.7 0.000 0.000 1.413 3.441 mp_alltoall_i44 222 10.7 1.412 3.441 1.412 3.441 ls_scf_store_result 1 5.0 0.000 0.000 3.405 3.414 calculate_norms 792 9.8 3.238 3.277 3.238 3.277 dbcsr_finalize 304 7.8 0.082 0.089 3.087 3.174 dbcsr_merge_all 275 8.9 0.892 0.920 2.872 2.953 qs_energies_init_hamiltonians 1 3.0 0.001 0.002 2.904 2.904 dbcsr_complete_redistribute 5 7.6 1.430 1.456 2.752 2.871 dbcsr_data_release 12724 10.6 2.332 2.857 2.332 2.857 matrix_ls_to_qs 2 6.0 0.000 0.000 2.416 2.547 dbcsr_sort_data 325 11.1 2.447 2.504 2.447 2.504 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.459 2.461 rebuild_ks_matrix 3 7.3 0.000 0.000 2.390 2.392 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.390 2.392 dbcsr_new_transposed 4 7.5 0.243 0.250 2.291 2.305 dbcsr_frobenius_norm 74 6.6 2.057 2.129 2.189 2.230 dbcsr_add_d 103 6.2 0.000 0.000 2.147 2.218 dbcsr_add_anytype 103 7.2 0.860 0.891 2.146 2.218 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=107.610000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6785.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7d51e9aba56ae0e0080ac2083bd6f96ee184da78_performance_tests/27/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 7009386627072 0.0% 0.0% 100.0% flops 9 x 9 x 32 7335108845568 0.0% 0.0% 100.0% flops 9 x 22 x 32 9866241589248 0.0% 0.0% 100.0% flops 22 x 9 x 32 9884108906496 0.0% 0.0% 100.0% flops 22 x 22 x 32 13354440523776 0.0% 0.0% 100.0% flops 32 x 32 x 9 20607185977344 0.0% 0.0% 100.0% flops 32 x 32 x 22 25186560638976 0.0% 0.0% 100.0% flops 9 x 32 x 32 28458319085568 0.0% 0.0% 100.0% flops 22 x 32 x 32 34782389993472 0.0% 0.0% 100.0% flops 9 x 32 x 9 42881542373376 0.0% 0.0% 100.0% flops 22 x 32 x 9 55680402235392 0.0% 0.0% 100.0% flops 9 x 32 x 22 55680402235392 0.0% 0.0% 100.0% flops 22 x 32 x 22 72328573419520 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 383.054662E+12 0.0% 0.0% 100.0% flops max/rank 733.641090E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 26899403712 0.0% 0.0% 100.0% number of processed stacks 118860288 0.0% 0.0% 100.0% average stack size 0.0 0.0 226.3 marketing flops 780.439111E+12 ------------------------------------------------------------------------------- # multiplications 1445 max memory usage/rank 567.328768E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 102097920 MPI messages size (bytes): total size 37.227590E+12 min size 0.000000E+00 max size 4.551360E+06 average size 364.626312E+03 MPI breakdown and total messages size (bytes): size <= 128 731472 0 128 < size <= 8192 11922720 97670922240 8192 < size <= 32768 24718992 614677610496 32768 < size <= 131072 20000256 1970081366016 131072 < size <= 4194304 42515668 24886801223040 4194304 < size <= 16777216 2208812 9656099886720 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4640 76559. MP_Allreduce 13152 2094. MP_Sync 1064 MP_Alltoall 2588 639444672. MP_SendRecv 168740 11136. MP_ISendRecv 92040 11136. MP_Wait 102830 MP_comm_split 40 MP_ISend 26090 85106. MP_IRecv 37890 59644. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.030 0.041 228.891 228.893 qs_mol_dyn_low 1 2.0 0.003 0.008 228.043 228.066 qs_forces 5 3.8 0.005 0.006 227.968 227.970 qs_energies 5 4.8 0.002 0.005 224.982 225.000 scf_env_do_scf 5 5.8 0.001 0.002 208.068 208.071 scf_env_do_scf_inner_loop 105 6.6 0.002 0.006 181.550 181.552 qs_scf_new_mos 105 7.6 0.000 0.001 140.020 140.200 qs_scf_loop_do_ot 105 8.6 0.001 0.001 140.019 140.199 dbcsr_multiply_generic 1445 12.2 0.125 0.136 132.582 133.123 ot_scf_mini 105 9.6 0.003 0.004 130.273 130.438 multiply_cannon 1445 13.2 0.276 0.294 114.296 116.379 multiply_cannon_loop 1445 14.2 2.377 2.511 112.485 113.680 velocity_verlet 4 3.0 0.001 0.002 105.725 105.726 ot_mini 105 10.6 0.001 0.003 58.195 58.348 multiply_cannon_multrec 69360 15.2 31.859 36.854 39.476 44.908 qs_ot_get_p 112 10.4 0.001 0.001 42.555 42.815 mp_waitall_1 488190 16.1 32.704 40.065 32.704 40.065 qs_ot_get_derivative 55 11.6 0.001 0.002 36.616 36.791 multiply_cannon_sync_h2d 69360 15.2 30.879 35.191 30.879 35.191 qs_ot_p2m_diag 40 11.0 0.020 0.030 31.831 31.903 multiply_cannon_metrocomm3 69360 15.2 0.194 0.202 23.924 31.430 rebuild_ks_matrix 110 8.4 0.000 0.001 29.454 29.657 qs_ks_build_kohn_sham_matrix 110 9.4 0.012 0.016 29.454 29.657 cp_dbcsr_syevd 40 12.0 0.002 0.002 28.807 28.808 qs_ks_update_qs_env 112 7.6 0.001 0.001 27.044 27.226 init_scf_loop 7 6.6 0.000 0.004 26.486 26.487 cp_fm_syevd 40 13.0 0.000 0.001 23.583 23.724 apply_preconditioner_dbcsr 62 12.6 0.000 0.001 23.001 23.294 apply_single 62 13.6 0.000 0.000 23.001 23.294 prepare_preconditioner 7 7.6 0.000 0.000 21.567 21.601 make_preconditioner 7 8.6 0.000 0.001 21.567 21.601 ot_new_cg_direction 55 11.6 0.001 0.002 20.953 20.953 cp_fm_redistribute_end 40 14.0 9.436 18.819 9.441 18.821 cp_fm_syevd_base 40 14.0 9.372 18.761 9.372 18.761 qs_rho_update_rho_low 110 7.6 0.001 0.001 18.039 18.371 calculate_rho_elec 110 8.6 0.030 0.034 18.038 18.370 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.354 14.464 make_full_inverse_cholesky 7 9.6 0.000 0.000 14.351 14.416 init_scf_run 5 5.8 0.000 0.001 14.233 14.234 scf_env_initial_rho_setup 5 6.8 0.002 0.003 14.233 14.234 rs_pw_transfer 690 11.5 0.011 0.013 12.382 13.570 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 13.257 13.372 density_rs2pw 110 9.6 0.006 0.007 12.077 13.305 pw_transfer 1645 12.4 0.083 0.107 12.197 12.433 fft_wrap_pw1pw2 1425 13.5 0.013 0.016 12.054 12.295 mp_sum_l 4684 12.4 11.279 11.985 11.279 11.985 calculate_dm_sparse 110 9.5 0.000 0.001 11.270 11.467 qs_vxc_create 110 10.4 0.002 0.004 10.503 10.548 fft_wrap_pw1pw2_240 915 15.0 1.185 1.282 10.251 10.496 cp_fm_cholesky_invert 7 10.6 9.988 9.997 9.988 9.997 check_diag 80 13.5 8.587 8.858 9.630 9.771 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.001 9.692 9.708 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 9.611 9.688 acc_transpose_blocks 69360 15.2 0.356 0.373 9.041 9.453 calculate_first_density_matrix 1 7.0 0.000 0.006 9.291 9.310 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 9.240 9.305 sum_up_and_integrate 60 10.3 0.028 0.038 8.962 8.978 integrate_v_rspace 60 11.3 0.002 0.002 8.934 8.946 fft3d_pb 915 16.0 2.383 2.661 8.331 8.631 dbcsr_mm_accdrv_process 154766 15.8 4.025 4.191 7.485 8.240 xc_rho_set_and_dset_create 110 12.4 0.075 0.096 7.323 7.579 acc_transpose_blocks_kernels 69360 16.2 0.855 0.903 7.151 7.496 make_m2s 2890 13.2 0.076 0.086 6.509 7.110 xc_vxc_pw_create 60 11.3 0.039 0.049 7.057 7.102 make_images 2890 14.2 0.241 0.259 6.406 7.005 make_full_single_inverse 7 9.6 0.001 0.002 6.941 6.977 multiply_cannon_metrocomm1 69360 15.2 0.090 0.096 4.004 6.798 jit_kernel_transpose 5 15.0 6.296 6.613 6.296 6.613 xc_pw_derive 510 13.4 0.005 0.007 6.259 6.344 mp_waitany 7680 13.5 4.538 5.738 4.538 5.738 mp_alltoall_z22v 2340 17.7 5.431 5.722 5.431 5.722 potential_pw2rs 60 12.3 0.003 0.003 5.103 5.140 multiply_cannon_metrocomm4 67915 15.2 0.179 0.195 2.049 4.932 wfi_extrapolate 5 7.8 0.000 0.002 4.653 4.653 grid_collocate_task_list 110 9.6 4.093 4.603 4.093 4.603 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=228.893000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=540.400000, yerr=2.059126 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 7d51e9aba56ae0e0080ac2083bd6f96ee184da78 Summary: empty Status: OK