=== 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: b234498075e47956c66564e93abbe409d5e88a7d ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, Cray-FFTW 3.3.8.10, # COSMA 2.6.2, ELPA 2022.11.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.1.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.1, # SIRIUS 7.4.3, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Author: Matthias Krack (03.02.2023) # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.1.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.1 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta_prefixed_scalapack.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b234498075e47956c66564e93abbe409d5e88a7d_performance_tests/01 job id: 45974019 --- 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/b234498075e47956c66564e93abbe409d5e88a7d_performance_tests/02 job id: 45974025 --- 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/b234498075e47956c66564e93abbe409d5e88a7d_performance_tests/03 job id: 45974029 --- 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/b234498075e47956c66564e93abbe409d5e88a7d_performance_tests/04 job id: 45974031 --- 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/b234498075e47956c66564e93abbe409d5e88a7d_performance_tests/05 job id: 45974035 --- 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/b234498075e47956c66564e93abbe409d5e88a7d_performance_tests/06 job id: 45974037 --- 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/b234498075e47956c66564e93abbe409d5e88a7d_performance_tests/07 job id: 45974043 --- 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/b234498075e47956c66564e93abbe409d5e88a7d_performance_tests/08 job id: 45974046 --- 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/b234498075e47956c66564e93abbe409d5e88a7d_performance_tests/09 job id: 45974050 --- 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/b234498075e47956c66564e93abbe409d5e88a7d_performance_tests/10 job id: 45974053 --- 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/b234498075e47956c66564e93abbe409d5e88a7d_performance_tests/11 job id: 45974056 --- 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/b234498075e47956c66564e93abbe409d5e88a7d_performance_tests/12 job id: 45974058 --- 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/b234498075e47956c66564e93abbe409d5e88a7d_performance_tests/13 job id: 45974061 --- 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/b234498075e47956c66564e93abbe409d5e88a7d_performance_tests/14 job id: 45974062 --- 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/b234498075e47956c66564e93abbe409d5e88a7d_performance_tests/15 job id: 45974064 --- 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/b234498075e47956c66564e93abbe409d5e88a7d_performance_tests/16 job id: 45974067 --- 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/b234498075e47956c66564e93abbe409d5e88a7d_performance_tests/17 job id: 45974068 --- 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/b234498075e47956c66564e93abbe409d5e88a7d_performance_tests/18 job id: 45974070 --- 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/b234498075e47956c66564e93abbe409d5e88a7d_performance_tests/19 job id: 45974072 --- 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/b234498075e47956c66564e93abbe409d5e88a7d_performance_tests/20 job id: 45974074 --- 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/b234498075e47956c66564e93abbe409d5e88a7d_performance_tests/21 job id: 45974075 --- 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/b234498075e47956c66564e93abbe409d5e88a7d_performance_tests/22 job id: 45974076 --- 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/b234498075e47956c66564e93abbe409d5e88a7d_performance_tests/23 job id: 45974078 --- 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/b234498075e47956c66564e93abbe409d5e88a7d_performance_tests/24 job id: 45974079 --- 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/b234498075e47956c66564e93abbe409d5e88a7d_performance_tests/25 job id: 45974080 --- 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/b234498075e47956c66564e93abbe409d5e88a7d_performance_tests/26 job id: 45974081 --- 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/b234498075e47956c66564e93abbe409d5e88a7d_performance_tests/27 job id: 45974082 --- 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/b234498075e47956c66564e93abbe409d5e88a7d_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 0.022 0.034 135.468 135.469 farming_run 1 2.0 134.697 134.699 135.440 135.443 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.463128E+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 0.006 0.025 115.894 115.895 qs_energies 1 2.0 0.000 0.000 115.704 115.706 mp2_main 1 3.0 0.000 0.000 113.651 113.653 mp2_gpw_main 1 4.0 0.020 0.026 112.646 112.649 mp2_ri_gpw_compute_in 1 5.0 0.171 0.174 93.779 93.864 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 55.407 55.492 mp2_eri_3c_integrate_gpw 272 7.0 0.152 0.167 41.709 46.619 get_2c_integrals 1 6.0 0.000 0.000 37.676 38.200 integrate_v_rspace 273 8.0 0.435 0.450 25.062 29.649 pw_transfer 6555 10.6 0.369 0.384 27.387 27.969 fft_wrap_pw1pw2 5465 11.4 0.045 0.049 26.024 26.407 grid_integrate_task_list 273 9.0 20.886 25.923 20.886 25.923 fft_wrap_pw1pw2_100 2178 12.4 1.219 1.441 23.550 23.952 compute_2c_integrals 1 7.0 0.004 0.004 19.771 19.772 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.922 19.467 mp2_eri_2c_integrate_gpw 1 9.0 2.372 2.431 18.919 19.464 rpa_ri_compute_en 1 5.0 0.001 0.001 18.757 18.851 cp_fm_cholesky_decompose 12 8.2 17.845 18.369 17.845 18.369 cholesky_decomp 1 7.0 0.000 0.000 16.741 17.259 fft3d_s 5443 13.4 16.129 16.328 16.151 16.350 ao_to_mo_and_store_B_mult_1 272 7.0 10.856 15.586 10.856 15.586 calculate_wavefunction 272 8.0 5.465 5.540 12.564 13.211 rpa_num_int 1 6.0 0.000 0.001 10.533 10.533 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.501 10.530 calc_mat_Q 8 8.0 0.000 0.000 9.354 9.472 contract_S_to_Q 8 9.0 0.000 0.000 8.774 8.894 calc_potential_gpw 544 9.5 0.005 0.006 8.256 8.687 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.231 8.576 parallel_gemm_fm 14 9.1 0.000 0.000 8.340 8.463 parallel_gemm_fm_cosma 14 10.1 8.340 8.462 8.340 8.462 potential_pw2rs 545 10.0 0.107 0.108 7.696 8.357 create_integ_mat 1 6.0 0.021 0.028 8.043 8.043 collocate_single_gaussian 272 10.0 0.039 0.041 7.498 7.787 array2fm 1 7.0 0.000 0.000 6.829 7.492 pw_scatter_s 2720 13.7 4.421 4.578 4.421 4.578 pw_gather_s 2722 13.2 3.849 4.287 3.849 4.287 array2fm_buffer_send 1 8.0 3.033 3.171 3.033 3.171 pw_poisson_solve 545 10.5 1.108 1.146 2.181 2.356 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.648517, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2736.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b234498075e47956c66564e93abbe409d5e88a7d_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 0.038 0.050 395.367 395.368 farming_run 1 2.0 394.556 394.565 395.319 395.331 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.229537E+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 8644517. 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 0.009 0.029 209.256 209.257 qs_energies 1 2.0 0.000 0.000 209.051 209.057 scf_env_do_scf 1 3.0 0.000 0.000 106.240 106.240 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.373 105.380 rebuild_ks_matrix 4 6.0 0.000 0.000 105.371 105.379 qs_ks_build_kohn_sham_matrix 4 7.0 0.056 0.064 105.371 105.379 hfx_ks_matrix 4 8.0 0.001 0.001 104.988 104.993 integrate_four_center 4 9.0 0.144 0.459 104.987 104.992 mp2_main 1 3.0 0.000 0.000 102.519 102.525 mp2_gpw_main 1 4.0 0.033 0.044 101.661 101.670 integrate_four_center_main 4 10.0 0.090 0.524 96.766 98.846 integrate_four_center_bin 267 11.0 96.676 98.840 96.676 98.840 init_scf_loop 1 4.0 0.000 0.000 91.944 91.944 mp2_ri_gpw_compute_in 1 5.0 0.064 0.065 74.961 75.984 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.519 55.541 mp2_eri_3c_integrate_gpw 91 7.0 0.145 0.160 42.220 47.262 integrate_v_rspace 95 8.0 0.397 0.566 28.576 33.439 pw_transfer 2240 10.6 0.144 0.178 29.955 30.324 ao_to_mo_and_store_B_mult_1 91 7.0 10.609 29.620 10.609 29.620 fft_wrap_pw1pw2 1868 11.4 0.018 0.022 28.965 29.353 grid_integrate_task_list 95 9.0 23.864 28.932 23.864 28.932 mp2_ri_gpw_compute_en 1 5.0 0.055 0.062 26.552 28.361 fft_wrap_pw1pw2_100 730 12.4 1.281 1.458 26.676 27.102 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.833 1.881 24.867 24.877 get_2c_integrals 1 6.0 0.000 0.000 20.348 20.379 compute_2c_integrals 1 7.0 0.003 0.005 19.334 19.337 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 18.890 19.206 mp2_eri_2c_integrate_gpw 1 9.0 1.738 1.833 18.889 19.205 fft3d_s 1823 13.4 18.469 18.864 18.483 18.876 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.292 14.292 calculate_wavefunction 91 8.0 2.025 2.066 9.756 9.978 mp2_ri_gpw_compute_en_expansio 172 7.0 0.559 0.577 8.746 9.252 potential_pw2rs 186 10.0 0.033 0.035 8.647 9.160 local_gemm 172 8.0 8.186 8.687 8.186 8.687 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.256 8.645 mp2_ri_gpw_compute_en_comm 22 7.0 0.500 0.526 7.896 8.404 collocate_single_gaussian 91 10.0 0.016 0.022 7.889 8.149 calc_potential_gpw 182 9.5 0.002 0.002 7.923 8.122 mp_sync 37 10.5 2.868 6.677 2.868 6.677 mp_sendrecv_dm3 2068 8.0 5.925 6.431 5.925 6.431 mp2_ri_gpw_compute_en_ener 172 7.0 6.348 6.408 6.348 6.408 pw_gather_s 912 13.2 4.909 5.373 4.909 5.373 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.661291, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1516.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b234498075e47956c66564e93abbe409d5e88a7d_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 452.583424E+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 374354. 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 0.018 0.030 51.688 51.689 qs_mol_dyn_low 1 2.0 0.003 0.003 51.402 51.408 qs_forces 11 3.9 0.003 0.009 51.337 51.339 qs_energies 11 4.9 0.002 0.004 49.892 49.904 scf_env_do_scf 11 5.9 0.000 0.001 43.625 43.625 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 41.628 41.629 dbcsr_multiply_generic 2286 12.5 0.091 0.096 31.883 32.204 qs_scf_new_mos 108 7.5 0.000 0.001 31.620 31.918 qs_scf_loop_do_ot 108 8.5 0.000 0.001 31.619 31.918 ot_scf_mini 108 9.5 0.002 0.002 30.071 30.258 multiply_cannon 2286 13.5 0.186 0.200 25.429 26.814 multiply_cannon_loop 2286 14.5 1.477 1.565 24.753 26.188 velocity_verlet 10 3.0 0.001 0.002 24.230 24.231 ot_mini 108 10.5 0.001 0.001 17.890 18.144 qs_ot_get_derivative 108 11.5 0.001 0.001 15.109 15.314 mp_waitall_1 245248 16.5 7.496 13.802 7.496 13.802 multiply_cannon_metrocomm3 54864 15.5 0.069 0.077 5.452 12.358 multiply_cannon_multrec 54864 15.5 4.331 6.828 7.814 11.335 qs_ot_get_p 119 10.4 0.001 0.001 7.813 8.120 rebuild_ks_matrix 119 8.3 0.000 0.000 7.832 7.981 qs_ks_build_kohn_sham_matrix 119 9.3 0.011 0.012 7.831 7.981 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.908 7.045 multiply_cannon_sync_h2d 54864 15.5 6.084 6.895 6.084 6.895 mp_sum_l 7207 12.9 4.571 6.073 4.571 6.073 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.016 5.435 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.189 5.310 qs_ot_p2m_diag 50 11.0 0.004 0.007 5.257 5.292 init_scf_run 11 5.9 0.000 0.001 4.987 4.987 scf_env_initial_rho_setup 11 6.9 0.000 0.000 4.987 4.987 dbcsr_mm_accdrv_process 76910 16.1 1.136 1.818 3.405 4.804 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.502 4.503 sum_up_and_integrate 119 10.3 0.012 0.015 4.491 4.498 integrate_v_rspace 119 11.3 0.003 0.004 4.479 4.487 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.283 4.283 cp_fm_redistribute_end 50 14.0 2.185 4.257 2.192 4.261 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.110 4.217 calculate_rho_elec 119 8.7 0.012 0.017 4.109 4.216 cp_fm_diag_elpa_base 50 14.0 2.062 4.147 2.066 4.157 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.834 3.016 apply_single 119 13.6 0.000 0.000 2.834 3.016 calculate_dm_sparse 119 9.5 0.000 0.001 2.870 3.011 jit_kernel_multiply 13 15.8 2.209 2.906 2.209 2.906 calculate_first_density_matrix 1 7.0 0.000 0.001 2.860 2.864 rs_pw_transfer 974 11.9 0.012 0.013 2.722 2.810 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.557 2.558 ot_diis_step 108 11.5 0.006 0.006 2.542 2.542 multiply_cannon_metrocomm1 54864 15.5 0.053 0.058 1.406 2.432 acc_transpose_blocks 54864 15.5 0.226 0.255 1.803 2.307 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.255 2.304 density_rs2pw 119 9.7 0.004 0.005 2.160 2.284 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.203 2.263 grid_integrate_task_list 119 12.3 2.018 2.116 2.018 2.116 wfi_extrapolate 11 7.9 0.001 0.001 2.070 2.070 init_scf_loop 11 6.9 0.000 0.001 1.980 1.980 potential_pw2rs 119 12.3 0.004 0.004 1.850 1.865 pw_transfer 1439 11.6 0.052 0.058 1.673 1.741 mp_sum_d 4135 12.0 1.128 1.728 1.128 1.728 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.598 1.669 make_m2s 4572 13.5 0.053 0.056 1.573 1.619 make_images 4572 14.5 0.132 0.138 1.491 1.535 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.477 1.493 mp_waitany 12084 13.8 1.270 1.450 1.270 1.450 grid_collocate_task_list 119 9.7 1.295 1.365 1.295 1.365 mp_alltoall_d11v 2130 13.8 1.210 1.364 1.210 1.364 fft3d_ps 1201 14.6 0.365 0.470 1.241 1.308 fft_wrap_pw1pw2_140 487 13.2 0.184 0.199 1.233 1.305 acc_transpose_blocks_kernels 54864 16.5 0.242 0.386 0.837 1.142 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=51.689000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=431.181818, yerr=1.113404 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b234498075e47956c66564e93abbe409d5e88a7d_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 487.469056E+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 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 0.032 0.091 38.904 38.905 qs_mol_dyn_low 1 2.0 0.003 0.003 38.474 38.482 qs_forces 11 3.9 0.008 0.023 38.411 38.414 qs_energies 11 4.9 0.001 0.001 36.732 36.740 scf_env_do_scf 11 5.9 0.000 0.001 31.622 31.623 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 29.158 29.158 dbcsr_multiply_generic 2286 12.5 0.098 0.101 21.174 21.525 qs_scf_new_mos 108 7.5 0.001 0.001 20.012 20.252 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.011 20.252 ot_scf_mini 108 9.5 0.002 0.003 19.107 19.278 velocity_verlet 10 3.0 0.001 0.002 18.089 18.090 multiply_cannon 2286 13.5 0.208 0.217 16.221 17.932 multiply_cannon_loop 2286 14.5 0.897 0.973 15.114 16.559 ot_mini 108 10.5 0.001 0.001 11.664 11.902 mp_waitall_1 200699 16.5 5.597 10.725 5.597 10.725 qs_ot_get_derivative 108 11.5 0.001 0.001 9.210 9.386 multiply_cannon_metrocomm3 27432 15.5 0.067 0.071 4.115 9.361 multiply_cannon_multrec 27432 15.5 1.983 4.383 5.859 8.903 rebuild_ks_matrix 119 8.3 0.000 0.000 7.177 7.323 qs_ks_build_kohn_sham_matrix 119 9.3 0.015 0.023 7.176 7.322 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.347 6.480 dbcsr_mm_accdrv_process 47894 16.0 3.083 5.308 3.807 5.627 qs_ot_get_p 119 10.4 0.001 0.001 4.544 4.783 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.539 4.392 sum_up_and_integrate 119 10.3 0.025 0.028 4.176 4.183 integrate_v_rspace 119 11.3 0.002 0.003 4.151 4.159 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.049 4.128 apply_single 119 13.6 0.000 0.000 3.049 4.127 mp_sum_l 7207 12.9 2.048 3.977 2.048 3.977 init_scf_run 11 5.9 0.000 0.001 3.842 3.842 scf_env_initial_rho_setup 11 6.9 0.000 0.001 3.842 3.842 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.760 3.792 calculate_rho_elec 119 8.7 0.021 0.024 3.760 3.791 qs_ot_p2m_diag 50 11.0 0.009 0.013 3.101 3.120 rs_pw_transfer 974 11.9 0.010 0.011 2.585 3.065 multiply_cannon_sync_h2d 27432 15.5 2.185 2.821 2.185 2.821 make_m2s 4572 13.5 0.052 0.054 2.455 2.673 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.666 2.666 make_images 4572 14.5 0.200 0.237 2.370 2.586 density_rs2pw 119 9.7 0.004 0.004 2.084 2.574 init_scf_loop 11 6.9 0.006 0.017 2.439 2.439 ot_diis_step 108 11.5 0.010 0.011 2.402 2.402 calculate_first_density_matrix 1 7.0 0.003 0.007 2.357 2.360 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.340 2.340 cp_fm_redistribute_end 50 14.0 1.185 2.311 1.190 2.315 cp_fm_diag_elpa_base 50 14.0 1.090 2.220 1.120 2.255 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.122 2.210 calculate_dm_sparse 119 9.5 0.000 0.001 2.044 2.116 pw_transfer 1439 11.6 0.065 0.070 1.905 1.939 grid_integrate_task_list 119 12.3 1.835 1.933 1.835 1.933 potential_pw2rs 119 12.3 0.006 0.006 1.896 1.906 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.882 1.883 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.813 1.849 jit_kernel_multiply 9 16.2 0.672 1.744 0.672 1.744 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.656 1.701 prepare_preconditioner 11 7.9 0.000 0.000 1.540 1.573 make_preconditioner 11 8.9 0.000 0.000 1.540 1.573 make_images_data 4572 15.5 0.045 0.051 1.139 1.568 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.444 1.508 acc_transpose_blocks 27432 15.5 0.109 0.115 1.187 1.487 fft_wrap_pw1pw2_140 487 13.2 0.202 0.211 1.439 1.474 hybrid_alltoall_any 4725 16.4 0.051 0.112 0.988 1.448 wfi_extrapolate 11 7.9 0.001 0.001 1.430 1.430 fft3d_ps 1201 14.6 0.517 0.576 1.364 1.394 grid_collocate_task_list 119 9.7 1.227 1.368 1.227 1.368 mp_alltoall_d11v 2130 13.8 1.172 1.364 1.172 1.364 mp_allgather_i34 2286 14.5 0.548 1.311 0.548 1.311 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.252 1.261 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.180 1.228 mp_waitany 5720 13.7 0.543 1.054 0.543 1.054 rs_pw_transfer_RS2PW_140 130 11.5 0.141 0.149 0.571 1.052 mp_sum_d 4135 12.0 0.571 1.024 0.571 1.024 qs_energies_init_hamiltonians 11 5.9 0.002 0.006 0.998 1.005 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.887 0.901 acc_transpose_blocks_kernels 27432 16.5 0.180 0.269 0.653 0.858 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=38.905000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=464.000000, yerr=1.705606 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b234498075e47956c66564e93abbe409d5e88a7d_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 522.342400E+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 1153191. 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 0.056 0.112 32.446 32.447 qs_mol_dyn_low 1 2.0 0.003 0.003 31.979 31.986 qs_forces 11 3.9 0.016 0.057 31.918 31.921 qs_energies 11 4.9 0.007 0.046 30.351 30.368 scf_env_do_scf 11 5.9 0.003 0.018 25.459 25.462 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 22.887 22.889 dbcsr_multiply_generic 2286 12.5 0.092 0.094 16.231 16.307 velocity_verlet 10 3.0 0.002 0.002 15.029 15.031 qs_scf_new_mos 108 7.5 0.001 0.001 14.745 14.765 qs_scf_loop_do_ot 108 8.5 0.001 0.001 14.745 14.765 ot_scf_mini 108 9.5 0.002 0.004 14.030 14.041 multiply_cannon 2286 13.5 0.196 0.204 12.976 13.693 multiply_cannon_loop 2286 14.5 0.630 0.660 12.215 12.956 ot_mini 108 10.5 0.001 0.001 8.611 8.626 qs_ot_get_derivative 108 11.5 0.001 0.001 7.120 7.131 multiply_cannon_multrec 18288 15.5 1.936 2.877 6.742 7.043 rebuild_ks_matrix 119 8.3 0.000 0.000 6.335 6.354 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.335 6.354 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.595 5.610 dbcsr_mm_accdrv_process 38222 16.0 3.886 5.166 4.721 5.577 sum_up_and_integrate 119 10.3 0.031 0.031 3.936 3.942 integrate_v_rspace 119 11.3 0.003 0.003 3.905 3.915 init_scf_run 11 5.9 0.000 0.001 3.634 3.634 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.634 3.634 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.442 3.451 calculate_rho_elec 119 8.7 0.031 0.031 3.442 3.450 mp_waitall_1 158411 16.6 2.552 3.416 2.552 3.416 qs_ot_get_p 119 10.4 0.001 0.001 3.290 3.320 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.641 3.289 rs_pw_transfer 974 11.9 0.009 0.010 2.304 2.576 init_scf_loop 11 6.9 0.002 0.014 2.554 2.559 calculate_first_density_matrix 1 7.0 0.001 0.003 2.431 2.432 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.055 2.363 apply_single 119 13.6 0.000 0.000 2.054 2.363 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.240 2.248 density_rs2pw 119 9.7 0.004 0.004 1.965 2.242 multiply_cannon_metrocomm3 18288 15.5 0.044 0.046 1.373 2.067 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.979 1.980 grid_integrate_task_list 119 12.3 1.797 1.889 1.797 1.889 pw_transfer 1439 11.6 0.065 0.070 1.881 1.888 jit_kernel_multiply 10 16.0 0.784 1.885 0.784 1.885 make_m2s 4572 13.5 0.044 0.045 1.743 1.883 calculate_dm_sparse 119 9.5 0.000 0.001 1.817 1.831 make_images 4572 14.5 0.190 0.201 1.659 1.798 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.789 1.798 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.763 1.765 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.731 1.741 cp_fm_diag_elpa_base 50 14.0 1.707 1.717 1.729 1.739 potential_pw2rs 119 12.3 0.007 0.008 1.703 1.709 prepare_preconditioner 11 7.9 0.000 0.000 1.700 1.702 make_preconditioner 11 8.9 0.000 0.001 1.700 1.702 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.551 1.637 multiply_cannon_sync_h2d 18288 15.5 1.397 1.620 1.397 1.620 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.591 1.600 qs_ot_get_derivative_diag 49 12.0 0.001 0.002 1.540 1.547 mp_sum_l 7207 12.9 1.165 1.487 1.165 1.487 ot_diis_step 108 11.5 0.011 0.011 1.470 1.471 fft_wrap_pw1pw2_140 487 13.2 0.255 0.262 1.446 1.455 grid_collocate_task_list 119 9.7 1.214 1.355 1.214 1.355 fft3d_ps 1201 14.6 0.528 0.545 1.275 1.285 acc_transpose_blocks 18288 15.5 0.076 0.079 1.242 1.274 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.193 1.199 wfi_extrapolate 11 7.9 0.001 0.001 1.141 1.141 multiply_cannon_metrocomm1 18288 15.5 0.028 0.029 0.436 1.062 qs_energies_init_hamiltonians 11 5.9 0.000 0.002 1.014 1.030 make_images_data 4572 15.5 0.044 0.048 0.767 0.920 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.867 0.892 mp_waitany 9880 13.7 0.564 0.840 0.564 0.840 rs_pw_transfer_RS2PW_140 130 11.5 0.121 0.132 0.567 0.839 hybrid_alltoall_any 4725 16.4 0.055 0.113 0.650 0.834 acc_transpose_blocks_kernels 18288 16.5 0.209 0.217 0.800 0.823 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.811 0.813 mp_alltoall_d11v 2130 13.8 0.652 0.741 0.652 0.741 cp_fm_cholesky_invert 11 10.9 0.711 0.715 0.711 0.715 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.641 0.706 mp_alltoall_z22v 1201 16.6 0.606 0.675 0.606 0.675 qs_env_update_s_mstruct 11 6.9 0.023 0.033 0.599 0.657 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=32.447000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=497.636364, yerr=1.822722 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b234498075e47956c66564e93abbe409d5e88a7d_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 556.576768E+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 0.013 0.031 34.583 34.584 qs_mol_dyn_low 1 2.0 0.003 0.003 34.395 34.400 qs_forces 11 3.9 0.003 0.004 34.326 34.330 qs_energies 11 4.9 0.001 0.002 32.608 32.616 scf_env_do_scf 11 5.9 0.000 0.001 27.638 27.640 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 24.255 24.255 dbcsr_multiply_generic 2286 12.5 0.096 0.099 17.756 17.867 velocity_verlet 10 3.0 0.002 0.002 17.550 17.552 qs_scf_new_mos 108 7.5 0.001 0.001 15.851 15.905 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.851 15.904 ot_scf_mini 108 9.5 0.002 0.003 14.922 14.977 multiply_cannon 2286 13.5 0.225 0.273 14.314 14.758 multiply_cannon_loop 2286 14.5 0.934 0.965 13.376 13.721 ot_mini 108 10.5 0.001 0.001 8.983 9.054 multiply_cannon_multrec 27432 15.5 2.332 3.010 8.556 8.936 dbcsr_mm_accdrv_process 47916 15.9 5.446 6.969 6.129 7.330 qs_ot_get_derivative 108 11.5 0.001 0.001 7.191 7.248 rebuild_ks_matrix 119 8.3 0.000 0.000 6.567 6.628 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.015 6.567 6.627 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.837 5.890 sum_up_and_integrate 119 10.3 0.035 0.037 3.777 3.793 integrate_v_rspace 119 11.3 0.003 0.003 3.742 3.758 init_scf_run 11 5.9 0.000 0.001 3.635 3.635 scf_env_initial_rho_setup 11 6.9 0.000 0.001 3.635 3.635 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.498 3.528 calculate_rho_elec 119 8.7 0.040 0.046 3.498 3.527 qs_ot_get_p 119 10.4 0.001 0.001 3.354 3.434 init_scf_loop 11 6.9 0.001 0.002 3.365 3.366 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.704 3.095 prepare_preconditioner 11 7.9 0.000 0.000 2.500 2.509 make_preconditioner 11 8.9 0.000 0.000 2.500 2.509 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.108 2.437 calculate_first_density_matrix 1 7.0 0.001 0.001 2.286 2.288 mp_waitall_1 137007 16.6 1.687 2.254 1.687 2.254 make_m2s 4572 13.5 0.054 0.056 2.104 2.239 rs_pw_transfer 974 11.9 0.009 0.010 1.954 2.141 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.010 2.134 apply_single 119 13.6 0.000 0.000 2.010 2.134 make_images 4572 14.5 0.272 0.333 1.997 2.131 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.113 2.122 density_rs2pw 119 9.7 0.004 0.004 1.909 2.104 calculate_dm_sparse 119 9.5 0.000 0.000 2.018 2.067 pw_transfer 1439 11.6 0.065 0.070 1.990 2.027 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.898 1.939 grid_integrate_task_list 119 12.3 1.815 1.908 1.815 1.908 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.864 1.892 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.824 1.825 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.802 1.802 ot_diis_step 108 11.5 0.012 0.012 1.752 1.752 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.621 1.633 fft_wrap_pw1pw2_140 487 13.2 0.290 0.302 1.568 1.609 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.567 1.576 cp_fm_diag_elpa_base 50 14.0 1.534 1.549 1.565 1.574 potential_pw2rs 119 12.3 0.008 0.009 1.535 1.537 jit_kernel_multiply 8 16.2 0.623 1.487 0.623 1.487 acc_transpose_blocks 27432 15.5 0.113 0.116 1.454 1.476 fft3d_ps 1201 14.6 0.555 0.601 1.335 1.361 grid_collocate_task_list 119 9.7 1.221 1.315 1.221 1.315 wfi_extrapolate 11 7.9 0.001 0.001 1.301 1.301 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.749 1.297 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.206 1.216 mp_sum_l 7207 12.9 0.917 1.190 0.917 1.190 cp_fm_upper_to_full 72 14.2 0.821 1.168 0.821 1.168 multiply_cannon_sync_h2d 27432 15.5 0.996 1.129 0.996 1.129 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.079 1.101 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.089 1.092 dbcsr_complete_redistribute 329 12.2 0.139 0.184 0.778 1.051 make_images_data 4572 15.5 0.045 0.048 0.818 0.951 hybrid_alltoall_any 4725 16.4 0.061 0.150 0.697 0.899 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.793 0.870 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.593 0.859 mp_alltoall_d11v 2130 13.8 0.705 0.852 0.705 0.852 acc_transpose_blocks_kernels 27432 16.5 0.266 0.276 0.825 0.838 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.828 0.835 cp_fm_cholesky_invert 11 10.9 0.719 0.722 0.719 0.722 mp_alltoall_i22 627 13.8 0.424 0.709 0.424 0.709 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=34.584000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=529.090909, yerr=2.429162 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b234498075e47956c66564e93abbe409d5e88a7d_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 609.726464E+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 0.088 0.097 27.428 27.429 qs_mol_dyn_low 1 2.0 0.003 0.003 27.124 27.131 qs_forces 11 3.9 0.015 0.023 27.063 27.066 qs_energies 11 4.9 0.001 0.002 25.352 25.358 scf_env_do_scf 11 5.9 0.000 0.001 20.708 20.708 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 18.294 18.295 velocity_verlet 10 3.0 0.002 0.002 14.057 14.060 dbcsr_multiply_generic 2286 12.5 0.089 0.094 11.705 11.776 qs_scf_new_mos 108 7.5 0.001 0.001 10.426 10.449 qs_scf_loop_do_ot 108 8.5 0.001 0.001 10.425 10.449 ot_scf_mini 108 9.5 0.002 0.002 9.765 9.791 multiply_cannon 2286 13.5 0.230 0.239 9.332 9.781 multiply_cannon_loop 2286 14.5 0.327 0.337 8.438 8.625 rebuild_ks_matrix 119 8.3 0.000 0.000 5.910 5.932 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 5.910 5.932 multiply_cannon_multrec 9144 15.5 1.566 1.817 5.704 5.919 ot_mini 108 10.5 0.001 0.001 5.415 5.446 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.268 5.287 qs_ot_get_derivative 108 11.5 0.001 0.001 4.145 4.171 dbcsr_mm_accdrv_process 12550 15.8 3.055 4.066 4.038 4.128 sum_up_and_integrate 119 10.3 0.038 0.042 3.614 3.621 integrate_v_rspace 119 11.3 0.003 0.003 3.576 3.583 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.449 3.455 calculate_rho_elec 119 8.7 0.060 0.061 3.449 3.455 init_scf_run 11 5.9 0.000 0.001 3.201 3.201 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.201 3.201 qs_ot_get_p 119 10.4 0.001 0.001 2.606 2.647 init_scf_loop 11 6.9 0.001 0.001 2.395 2.397 calculate_first_density_matrix 1 7.0 0.014 0.015 2.095 2.097 pw_transfer 1439 11.6 0.066 0.070 2.000 2.011 grid_integrate_task_list 119 12.3 1.854 1.927 1.854 1.927 density_rs2pw 119 9.7 0.004 0.004 1.786 1.925 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.906 1.918 make_m2s 4572 13.5 0.034 0.035 1.647 1.814 rs_pw_transfer 974 11.9 0.008 0.008 1.639 1.765 mp_waitall_1 115863 16.7 1.248 1.764 1.248 1.764 jit_kernel_multiply 10 15.6 0.945 1.737 0.945 1.737 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.730 1.732 make_images 4572 14.5 0.268 0.306 1.561 1.725 calculate_dm_sparse 119 9.5 0.000 0.000 1.691 1.706 prepare_preconditioner 11 7.9 0.000 0.000 1.664 1.668 make_preconditioner 11 8.9 0.000 0.000 1.664 1.668 fft_wrap_pw1pw2_140 487 13.2 0.366 0.373 1.578 1.591 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.553 1.579 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.540 1.541 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.515 1.516 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.457 1.476 grid_collocate_task_list 119 9.7 1.270 1.388 1.270 1.388 potential_pw2rs 119 12.3 0.010 0.011 1.353 1.356 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.345 1.355 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.273 1.282 cp_fm_diag_elpa_base 50 14.0 1.246 1.261 1.272 1.280 ot_diis_step 108 11.5 0.012 0.013 1.259 1.259 fft3d_ps 1201 14.6 0.557 0.570 1.239 1.247 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.240 1.243 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.140 1.166 apply_single 119 13.6 0.000 0.000 1.139 1.166 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.121 1.127 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.081 1.094 wfi_extrapolate 11 7.9 0.001 0.001 1.059 1.059 hybrid_alltoall_any 4725 16.4 0.062 0.173 0.725 0.979 make_images_data 4572 15.5 0.039 0.042 0.767 0.963 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.868 0.918 acc_transpose_blocks 9144 15.5 0.038 0.039 0.892 0.898 cp_fm_cholesky_invert 11 10.9 0.828 0.831 0.828 0.831 multiply_cannon_sync_h2d 9144 15.5 0.706 0.790 0.706 0.790 mp_alltoall_d11v 2130 13.8 0.709 0.772 0.709 0.772 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.765 0.767 qs_env_update_s_mstruct 11 6.9 0.005 0.012 0.695 0.747 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.695 0.705 multiply_cannon_metrocomm3 9144 15.5 0.018 0.019 0.316 0.667 acc_transpose_blocks_kernels 9144 16.5 0.116 0.119 0.652 0.655 mp_allgather_i34 2286 14.5 0.226 0.627 0.226 0.627 mp_waitany 5200 13.7 0.470 0.586 0.470 0.586 mp_alltoall_z22v 1201 16.6 0.556 0.582 0.556 0.582 qs_create_task_list 11 7.9 0.000 0.001 0.542 0.568 generate_qs_task_list 11 8.9 0.187 0.210 0.541 0.568 rs_pw_transfer_RS2PW_140 130 11.5 0.101 0.104 0.431 0.556 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=27.429000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=578.181818, yerr=4.130115 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b234498075e47956c66564e93abbe409d5e88a7d_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 787.406848E+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 0.028 0.047 40.678 40.679 qs_mol_dyn_low 1 2.0 0.003 0.003 40.429 40.436 qs_forces 11 3.9 0.002 0.002 40.366 40.367 qs_energies 11 4.9 0.001 0.001 38.408 38.411 scf_env_do_scf 11 5.9 0.001 0.001 32.921 32.921 scf_env_do_scf_inner_loop 108 6.5 0.007 0.011 25.069 25.069 velocity_verlet 10 3.0 0.002 0.002 22.980 22.986 dbcsr_multiply_generic 2286 12.5 0.097 0.098 17.069 17.342 qs_scf_new_mos 108 7.5 0.001 0.001 15.492 15.586 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.492 15.585 ot_scf_mini 108 9.5 0.002 0.002 14.429 14.525 multiply_cannon 2286 13.5 0.301 0.304 13.474 14.368 multiply_cannon_loop 2286 14.5 0.343 0.351 12.251 13.191 ot_mini 108 10.5 0.001 0.001 8.655 8.767 multiply_cannon_multrec 9144 15.5 3.424 4.782 8.560 8.693 init_scf_loop 11 6.9 0.001 0.001 7.827 7.828 rebuild_ks_matrix 119 8.3 0.000 0.001 7.153 7.289 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 7.152 7.288 prepare_preconditioner 11 7.9 0.000 0.000 6.798 6.811 make_preconditioner 11 8.9 0.000 0.000 6.798 6.811 qs_ot_get_derivative 108 11.5 0.001 0.001 6.672 6.768 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.403 6.689 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.463 6.587 dbcsr_mm_accdrv_process 12550 15.8 4.052 5.292 5.012 6.334 cp_fm_upper_to_full 72 14.2 3.257 4.592 3.257 4.592 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.154 4.158 calculate_rho_elec 119 8.7 0.119 0.122 4.153 4.157 sum_up_and_integrate 119 10.3 0.065 0.066 3.910 3.916 integrate_v_rspace 119 11.3 0.004 0.004 3.845 3.853 init_scf_run 11 5.9 0.000 0.001 3.464 3.464 scf_env_initial_rho_setup 11 6.9 0.000 0.001 3.464 3.464 qs_ot_get_p 119 10.4 0.001 0.001 3.100 3.231 mp_waitall_1 94719 16.7 2.158 3.180 2.158 3.180 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.436 2.817 dbcsr_complete_redistribute 329 12.2 0.289 0.293 1.907 2.680 pw_transfer 1439 11.6 0.069 0.070 2.599 2.604 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.501 2.505 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.106 2.404 apply_single 119 13.6 0.000 0.000 2.105 2.404 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.611 2.363 make_m2s 4572 13.5 0.038 0.038 2.127 2.267 multiply_cannon_metrocomm3 9144 15.5 0.019 0.020 1.266 2.198 make_images 4572 14.5 0.359 0.391 2.007 2.147 mp_alltoall_i22 627 13.8 1.365 2.139 1.365 2.139 calculate_dm_sparse 119 9.5 0.000 0.000 2.123 2.138 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.390 2.135 density_rs2pw 119 9.7 0.004 0.004 2.112 2.130 fft_wrap_pw1pw2_140 487 13.2 0.618 0.622 2.119 2.125 grid_integrate_task_list 119 12.3 2.033 2.050 2.033 2.050 calculate_first_density_matrix 1 7.0 0.001 0.001 2.046 2.046 ot_diis_step 108 11.5 0.014 0.014 1.960 1.961 qs_ot_p2m_diag 50 11.0 0.043 0.044 1.874 1.875 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.764 1.765 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.764 1.764 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.640 1.689 mp_sum_l 7207 12.9 0.928 1.640 0.928 1.640 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.608 1.609 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.520 1.554 fft3d_ps 1201 14.6 0.594 0.605 1.517 1.521 grid_collocate_task_list 119 9.7 1.468 1.476 1.468 1.476 rs_pw_transfer 974 11.9 0.009 0.009 1.402 1.426 cp_fm_cholesky_invert 11 10.9 1.408 1.412 1.408 1.412 jit_kernel_multiply 8 15.6 0.934 1.411 0.934 1.411 wfi_extrapolate 11 7.9 0.001 0.001 1.354 1.354 potential_pw2rs 119 12.3 0.014 0.014 1.335 1.337 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.334 1.335 cp_fm_diag_elpa_base 50 14.0 1.186 1.236 1.333 1.333 hybrid_alltoall_any 4725 16.4 0.087 0.147 1.034 1.223 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.204 1.210 make_images_data 4572 15.5 0.043 0.046 0.959 1.131 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.108 1.126 qs_env_update_s_mstruct 11 6.9 0.002 0.003 1.098 1.110 mp_alltoall_d11v 2130 13.8 1.030 1.059 1.030 1.059 multiply_cannon_sync_h2d 9144 15.5 1.039 1.046 1.039 1.046 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.982 1.019 qs_create_task_list 11 7.9 0.007 0.008 0.940 0.952 generate_qs_task_list 11 8.9 0.368 0.387 0.932 0.944 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.901 0.914 acc_transpose_blocks 9144 15.5 0.038 0.039 0.894 0.898 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=40.679000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=730.545455, yerr=18.127647 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b234498075e47956c66564e93abbe409d5e88a7d_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 503.189504E+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 1961582. 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 0.017 0.041 81.297 81.297 qs_mol_dyn_low 1 2.0 0.003 0.005 80.986 80.995 qs_forces 11 3.9 0.004 0.010 80.912 80.913 qs_energies 11 4.9 0.002 0.007 78.063 78.075 scf_env_do_scf 11 5.9 0.001 0.002 69.234 69.236 scf_env_do_scf_inner_loop 99 6.5 0.002 0.008 63.825 63.826 dbcsr_multiply_generic 2055 12.4 0.105 0.128 50.413 50.759 qs_scf_new_mos 99 7.5 0.001 0.001 46.567 46.699 qs_scf_loop_do_ot 99 8.5 0.001 0.001 46.566 46.699 ot_scf_mini 99 9.5 0.002 0.003 44.195 44.307 multiply_cannon 2055 13.4 0.182 0.189 42.062 42.912 multiply_cannon_loop 2055 14.4 1.510 1.547 41.159 41.983 velocity_verlet 10 3.0 0.001 0.002 41.777 41.778 ot_mini 99 10.5 0.006 0.042 25.919 26.031 qs_ot_get_derivative 99 11.5 0.001 0.002 19.049 19.169 multiply_cannon_multrec 49320 15.4 12.514 13.261 17.384 18.018 rebuild_ks_matrix 110 8.3 0.000 0.001 14.568 14.672 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.013 14.567 14.672 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.778 12.863 mp_waitall_1 220248 16.4 10.480 11.417 10.480 11.417 multiply_cannon_sync_h2d 49320 15.4 10.391 11.080 10.391 11.080 qs_ot_get_p 110 10.4 0.001 0.001 9.776 9.925 apply_preconditioner_dbcsr 110 12.6 0.000 0.001 7.308 7.981 apply_single 110 13.6 0.000 0.001 7.308 7.981 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 6.996 7.515 multiply_cannon_metrocomm3 49320 15.4 0.079 0.083 6.412 7.299 sum_up_and_integrate 110 10.3 0.037 0.043 7.064 7.079 integrate_v_rspace 110 11.3 0.003 0.004 7.027 7.051 init_scf_run 11 5.9 0.000 0.001 6.758 6.758 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.758 6.758 qs_ot_p2m_diag 48 11.0 0.013 0.019 6.670 6.695 ot_diis_step 99 11.5 0.015 0.086 6.683 6.688 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.538 6.669 calculate_rho_elec 110 8.6 0.022 0.026 6.538 6.668 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.689 5.689 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.500 5.567 init_scf_loop 11 6.9 0.000 0.003 5.382 5.382 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.136 5.155 cp_fm_diag_elpa_base 48 14.0 5.124 5.145 5.135 5.153 dbcsr_mm_accdrv_process 87628 16.1 1.883 1.955 4.745 5.035 mp_sum_l 6514 12.8 4.118 4.956 4.118 4.956 rs_pw_transfer 902 11.9 0.012 0.013 3.678 4.365 density_rs2pw 110 9.6 0.005 0.006 3.335 4.072 wfi_extrapolate 11 7.9 0.001 0.001 3.986 3.986 make_m2s 4110 13.4 0.060 0.065 3.779 3.875 calculate_dm_sparse 110 9.5 0.001 0.001 3.741 3.844 make_images 4110 14.4 0.177 0.190 3.684 3.785 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.568 3.573 grid_integrate_task_list 110 12.3 3.238 3.410 3.238 3.410 multiply_cannon_metrocomm1 49320 15.4 0.060 0.063 2.297 3.339 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.202 3.276 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.217 3.273 prepare_preconditioner 11 7.9 0.000 0.000 3.181 3.205 make_preconditioner 11 8.9 0.000 0.001 3.181 3.205 pw_transfer 1331 11.6 0.056 0.071 3.077 3.158 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 2.988 3.065 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.971 3.030 calculate_first_density_matrix 1 7.0 0.001 0.003 2.678 2.683 fft_wrap_pw1pw2_140 451 13.1 0.454 0.497 2.558 2.638 jit_kernel_multiply 13 15.9 2.589 2.614 2.589 2.614 potential_pw2rs 110 12.3 0.006 0.007 2.592 2.609 mp_waitany 14300 13.8 1.866 2.563 1.866 2.563 mp_alltoall_d11v 2046 13.8 2.057 2.554 2.057 2.554 grid_collocate_task_list 110 9.6 2.086 2.282 2.086 2.282 fft3d_ps 1111 14.6 0.779 0.863 2.179 2.235 acc_transpose_blocks 49320 15.4 0.210 0.218 2.063 2.101 mp_sum_d 3889 11.9 1.398 1.947 1.398 1.947 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.920 1.938 make_images_data 4110 15.4 0.042 0.046 1.735 1.883 hybrid_alltoall_any 4261 16.3 0.083 0.482 1.527 1.793 cp_fm_cholesky_invert 11 10.9 1.729 1.733 1.729 1.733 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.637 1.664 rs_gather_matrices 110 12.3 0.160 0.177 1.099 1.643 rs_pw_transfer_RS2PW_140 121 11.5 0.171 0.182 0.948 1.630 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=81.297000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=477.090909, yerr=2.609582 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b234498075e47956c66564e93abbe409d5e88a7d_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 589.492224E+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 2335502. 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 0.023 0.049 69.321 69.323 qs_mol_dyn_low 1 2.0 0.003 0.003 68.774 68.985 qs_forces 11 3.9 0.004 0.007 68.493 68.494 qs_energies 11 4.9 0.007 0.048 65.148 65.151 scf_env_do_scf 11 5.9 0.002 0.010 56.492 56.497 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 48.882 48.884 dbcsr_multiply_generic 2055 12.4 0.113 0.116 37.290 37.460 velocity_verlet 10 3.0 0.001 0.002 36.043 36.048 qs_scf_new_mos 99 7.5 0.001 0.001 32.581 32.703 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.581 32.703 multiply_cannon 2055 13.4 0.221 0.241 30.876 31.989 ot_scf_mini 99 9.5 0.003 0.004 30.930 31.041 multiply_cannon_loop 2055 14.4 0.923 0.949 29.633 30.484 ot_mini 99 10.5 0.001 0.001 17.958 18.084 multiply_cannon_multrec 24660 15.4 7.649 9.387 13.838 15.594 rebuild_ks_matrix 110 8.3 0.000 0.001 13.858 13.924 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.015 13.857 13.923 qs_ot_get_derivative 99 11.5 0.001 0.001 12.195 12.308 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.198 12.262 mp_waitall_1 176588 16.5 7.511 9.794 7.511 9.794 multiply_cannon_sync_h2d 24660 15.4 6.999 8.013 6.999 8.013 multiply_cannon_metrocomm3 24660 15.4 0.067 0.070 5.104 7.925 init_scf_loop 11 6.9 0.003 0.019 7.574 7.577 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.463 7.129 apply_single 110 13.6 0.000 0.001 6.463 7.129 sum_up_and_integrate 110 10.3 0.054 0.060 6.657 6.671 integrate_v_rspace 110 11.3 0.003 0.010 6.603 6.621 qs_ot_get_p 110 10.4 0.001 0.001 6.396 6.551 dbcsr_mm_accdrv_process 52282 16.1 4.666 5.407 6.028 6.356 init_scf_run 11 5.9 0.000 0.001 6.214 6.215 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.214 6.215 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.191 6.201 calculate_rho_elec 110 8.6 0.041 0.055 6.190 6.200 ot_diis_step 99 11.5 0.010 0.010 5.716 5.716 prepare_preconditioner 11 7.9 0.000 0.000 5.447 5.465 make_preconditioner 11 8.9 0.000 0.002 5.447 5.465 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.600 5.328 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.033 5.188 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.558 4.577 make_m2s 4110 13.4 0.056 0.059 4.076 4.506 make_images 4110 14.4 0.402 0.447 3.973 4.401 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.135 4.136 pw_transfer 1331 11.6 0.067 0.075 3.633 3.772 density_rs2pw 110 9.6 0.004 0.005 3.306 3.761 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.526 3.667 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.574 3.586 cp_fm_diag_elpa_base 48 14.0 3.526 3.545 3.571 3.584 rs_pw_transfer 902 11.9 0.012 0.014 3.042 3.554 wfi_extrapolate 11 7.9 0.001 0.001 3.475 3.475 grid_integrate_task_list 110 12.3 3.150 3.350 3.150 3.350 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.300 3.302 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.124 3.203 fft_wrap_pw1pw2_140 451 13.1 0.524 0.537 3.026 3.167 calculate_dm_sparse 110 9.5 0.001 0.001 3.001 3.034 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.909 2.944 make_images_data 4110 15.4 0.047 0.050 2.294 2.731 hybrid_alltoall_any 4261 16.3 0.103 0.446 1.998 2.711 fft3d_ps 1111 14.6 1.105 1.328 2.496 2.647 calculate_first_density_matrix 1 7.0 0.001 0.002 2.618 2.621 grid_collocate_task_list 110 9.6 2.054 2.515 2.054 2.515 cp_fm_cholesky_invert 11 10.9 2.500 2.506 2.500 2.506 potential_pw2rs 110 12.3 0.008 0.008 2.448 2.464 mp_sum_l 6514 12.8 1.717 2.373 1.717 2.373 mp_alltoall_d11v 2046 13.8 1.737 2.020 1.737 2.020 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.929 1.952 jit_kernel_multiply 10 16.3 1.013 1.910 1.013 1.910 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.884 1.885 mp_waitany 10164 13.8 1.302 1.863 1.302 1.863 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.822 1.834 multiply_cannon_metrocomm4 22605 15.4 0.076 0.079 0.803 1.779 mp_irecv_dv 57340 16.2 0.678 1.660 0.678 1.660 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.582 1.597 mp_allgather_i34 2055 14.4 0.570 1.554 0.570 1.554 acc_transpose_blocks 24660 15.4 0.112 0.114 1.532 1.550 rs_pw_transfer_RS2PW_140 121 11.5 0.209 0.221 1.032 1.530 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.367 1.477 dbcsr_complete_redistribute 325 12.2 0.261 0.354 1.176 1.440 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=69.323000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=556.727273, yerr=6.593260 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b234498075e47956c66564e93abbe409d5e88a7d_performance_tests/11/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 404.681598E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 3346752 0.0% 0.0% 100.0% average stack size 0.0 0.0 294.1 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 659.320832E+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 0.016 0.031 59.295 59.296 qs_mol_dyn_low 1 2.0 0.003 0.003 59.019 59.040 qs_forces 11 3.9 0.003 0.003 58.945 58.945 qs_energies 11 4.9 0.001 0.002 55.753 55.756 scf_env_do_scf 11 5.9 0.000 0.001 47.958 47.958 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 39.577 39.577 velocity_verlet 10 3.0 0.001 0.002 32.285 32.286 dbcsr_multiply_generic 2055 12.4 0.104 0.110 28.010 28.306 qs_scf_new_mos 99 7.5 0.001 0.001 24.540 24.642 qs_scf_loop_do_ot 99 8.5 0.001 0.001 24.540 24.641 ot_scf_mini 99 9.5 0.002 0.003 23.329 23.447 multiply_cannon 2055 13.4 0.212 0.223 21.861 23.077 multiply_cannon_loop 2055 14.4 0.616 0.627 20.697 21.629 ot_mini 99 10.5 0.001 0.001 13.259 13.372 rebuild_ks_matrix 110 8.3 0.000 0.000 12.356 12.517 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.356 12.517 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.877 11.018 multiply_cannon_multrec 16440 15.4 4.019 4.979 9.731 10.651 mp_waitall_1 139946 16.5 6.692 9.355 6.692 9.355 qs_ot_get_derivative 99 11.5 0.001 0.001 8.872 8.991 init_scf_loop 11 6.9 0.001 0.002 8.345 8.345 multiply_cannon_metrocomm3 16440 15.4 0.042 0.044 4.104 6.848 prepare_preconditioner 11 7.9 0.000 0.000 6.582 6.604 make_preconditioner 11 8.9 0.000 0.000 6.582 6.604 sum_up_and_integrate 110 10.3 0.062 0.063 6.527 6.543 integrate_v_rspace 110 11.3 0.003 0.003 6.465 6.480 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.916 6.273 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.984 5.995 calculate_rho_elec 110 8.6 0.059 0.059 5.984 5.995 dbcsr_mm_accdrv_process 34862 16.1 4.565 5.295 5.565 5.689 init_scf_run 11 5.9 0.000 0.001 5.431 5.431 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.431 5.431 qs_ot_get_p 110 10.4 0.001 0.001 5.281 5.416 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.922 5.316 apply_single 110 13.6 0.000 0.000 4.922 5.316 make_m2s 4110 13.4 0.050 0.051 4.095 4.453 ot_diis_step 99 11.5 0.010 0.010 4.359 4.360 make_images 4110 14.4 0.392 0.505 3.985 4.341 multiply_cannon_sync_h2d 16440 15.4 3.710 4.315 3.710 4.315 density_rs2pw 110 9.6 0.004 0.005 3.086 4.271 rs_pw_transfer 902 11.9 0.011 0.012 2.633 3.813 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.712 3.715 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.024 3.665 pw_transfer 1331 11.6 0.066 0.075 3.642 3.650 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.535 3.546 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.383 3.383 grid_integrate_task_list 110 12.3 3.165 3.377 3.165 3.377 fft_wrap_pw1pw2_140 451 13.1 0.642 0.650 3.057 3.069 wfi_extrapolate 11 7.9 0.001 0.001 2.962 2.962 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.856 2.865 cp_fm_diag_elpa_base 48 14.0 2.789 2.819 2.855 2.863 hybrid_alltoall_any 4261 16.3 0.105 0.376 2.157 2.857 make_images_data 4110 15.4 0.043 0.048 2.397 2.844 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.719 2.721 cp_fm_cholesky_invert 11 10.9 2.570 2.576 2.570 2.576 calculate_dm_sparse 110 9.5 0.001 0.001 2.481 2.511 grid_collocate_task_list 110 9.6 2.083 2.474 2.083 2.474 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.408 2.457 mp_waitany 17072 13.8 1.159 2.403 1.159 2.403 calculate_first_density_matrix 1 7.0 0.001 0.001 2.384 2.385 fft3d_ps 1111 14.6 1.089 1.100 2.353 2.364 multiply_cannon_metrocomm4 14385 15.4 0.046 0.049 0.862 2.357 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.267 2.331 potential_pw2rs 110 12.3 0.011 0.011 2.260 2.268 mp_irecv_dv 48980 15.7 0.791 2.234 0.791 2.234 mp_alltoall_d11v 2046 13.8 1.717 2.101 1.717 2.101 rs_pw_transfer_RS2PW_140 121 11.5 0.180 0.184 0.883 2.053 mp_sum_l 6514 12.8 1.438 1.997 1.438 1.997 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.936 1.937 dbcsr_complete_redistribute 325 12.2 0.327 0.358 1.368 1.825 cp_fm_upper_to_full 70 14.2 1.390 1.769 1.390 1.769 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.674 1.694 cp_fm_cholesky_decompose 22 10.9 1.553 1.569 1.553 1.569 mp_allgather_i34 2055 14.4 0.484 1.543 0.484 1.543 jit_kernel_multiply 8 16.5 0.614 1.491 0.614 1.491 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.450 1.463 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.356 1.453 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.948 1.394 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.233 1.241 rs_gather_matrices 110 12.3 0.235 0.263 0.962 1.234 acc_transpose_blocks 16440 15.4 0.074 0.076 1.195 1.204 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=59.296000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=624.545455, yerr=6.959149 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b234498075e47956c66564e93abbe409d5e88a7d_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 724.471808E+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 0.026 0.059 65.300 65.311 qs_mol_dyn_low 1 2.0 0.003 0.005 64.937 64.945 qs_forces 11 3.9 0.004 0.008 64.422 64.423 qs_energies 11 4.9 0.003 0.006 60.987 60.991 scf_env_do_scf 11 5.9 0.001 0.002 52.638 52.640 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 40.961 40.963 velocity_verlet 10 3.0 0.003 0.015 37.244 37.264 dbcsr_multiply_generic 2055 12.4 0.111 0.120 29.141 29.333 qs_scf_new_mos 99 7.5 0.001 0.001 25.964 26.072 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.964 26.071 ot_scf_mini 99 9.5 0.003 0.004 24.358 24.452 multiply_cannon 2055 13.4 0.243 0.274 22.283 23.375 multiply_cannon_loop 2055 14.4 0.883 0.902 20.947 21.566 ot_mini 99 10.5 0.001 0.001 13.839 13.956 multiply_cannon_multrec 24660 15.4 4.230 6.865 12.629 13.934 rebuild_ks_matrix 110 8.3 0.000 0.000 12.259 12.369 qs_ks_build_kohn_sham_matrix 110 9.3 0.015 0.026 12.259 12.369 init_scf_loop 11 6.9 0.001 0.004 11.637 11.639 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.816 10.913 prepare_preconditioner 11 7.9 0.000 0.001 9.829 9.845 make_preconditioner 11 8.9 0.001 0.001 9.829 9.845 qs_ot_get_derivative 99 11.5 0.001 0.002 9.700 9.802 make_full_inverse_cholesky 11 9.9 0.000 0.001 8.027 9.531 dbcsr_mm_accdrv_process 52304 16.0 6.837 8.545 8.253 9.192 sum_up_and_integrate 110 10.3 0.068 0.072 6.533 6.551 integrate_v_rspace 110 11.3 0.003 0.005 6.464 6.484 qs_rho_update_rho_low 110 7.6 0.001 0.004 6.085 6.139 calculate_rho_elec 110 8.6 0.078 0.082 6.084 6.139 mp_waitall_1 121746 16.5 4.193 6.052 4.193 6.052 qs_ot_get_p 110 10.4 0.001 0.001 5.572 5.711 make_m2s 4110 13.4 0.059 0.060 5.297 5.589 init_scf_run 11 5.9 0.000 0.001 5.528 5.528 scf_env_initial_rho_setup 11 6.9 0.002 0.018 5.527 5.528 make_images 4110 14.4 0.581 0.705 5.159 5.447 cp_fm_upper_to_full 70 14.2 3.389 4.936 3.389 4.936 ot_diis_step 99 11.5 0.012 0.016 4.102 4.102 apply_preconditioner_dbcsr 110 12.6 0.000 0.001 4.011 4.082 apply_single 110 13.6 0.000 0.000 4.011 4.082 dbcsr_complete_redistribute 325 12.2 0.418 0.462 2.666 3.843 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.822 3.837 pw_transfer 1331 11.6 0.066 0.077 3.737 3.775 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.631 3.672 density_rs2pw 110 9.6 0.004 0.005 3.005 3.512 grid_integrate_task_list 110 12.3 3.262 3.445 3.262 3.445 cp_dbcsr_syevd 48 12.0 0.003 0.005 3.351 3.352 multiply_cannon_sync_h2d 24660 15.4 3.189 3.346 3.189 3.346 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.139 3.303 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.248 3.296 fft_wrap_pw1pw2_140 451 13.1 0.672 0.693 3.135 3.179 multiply_cannon_metrocomm3 24660 15.4 0.036 0.037 1.371 3.141 qs_ot_get_derivative_diag 47 12.0 0.001 0.002 2.949 3.003 make_images_data 4110 15.4 0.046 0.050 2.685 2.998 calculate_dm_sparse 110 9.5 0.001 0.001 2.945 2.980 hybrid_alltoall_any 4261 16.3 0.121 0.459 2.256 2.977 wfi_extrapolate 11 7.9 0.001 0.001 2.968 2.968 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.793 2.945 rs_pw_transfer 902 11.9 0.010 0.011 2.390 2.910 mp_alltoall_i22 605 13.7 1.681 2.862 1.681 2.862 cp_fm_diag_elpa 48 13.0 0.000 0.001 2.826 2.838 cp_fm_diag_elpa_base 48 14.0 2.675 2.736 2.822 2.835 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.758 2.760 cp_fm_cholesky_invert 11 10.9 2.622 2.632 2.622 2.632 calculate_first_density_matrix 1 7.0 0.001 0.005 2.442 2.448 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.410 2.445 fft3d_ps 1111 14.6 1.081 1.112 2.389 2.419 grid_collocate_task_list 110 9.6 2.179 2.418 2.179 2.418 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.332 2.333 potential_pw2rs 110 12.3 0.012 0.013 2.173 2.184 mp_alltoall_d11v 2046 13.8 1.710 2.009 1.710 2.009 jit_kernel_multiply 11 15.8 1.085 1.977 1.085 1.977 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.672 1.706 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.600 1.701 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.688 1.700 cp_fm_cholesky_decompose 22 10.9 1.615 1.660 1.615 1.660 multiply_cannon_metrocomm4 20550 15.4 0.062 0.065 0.843 1.604 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.574 1.595 acc_transpose_blocks 24660 15.4 0.108 0.112 1.523 1.556 mp_waitany 13376 13.8 1.066 1.555 1.066 1.555 mp_irecv_dv 62702 16.1 0.740 1.526 0.740 1.526 mp_sum_l 6514 12.8 0.885 1.483 0.885 1.483 mp_allgather_i34 2055 14.4 0.433 1.463 0.433 1.463 qs_env_update_s_mstruct 11 6.9 0.026 0.052 1.251 1.387 rs_pw_transfer_RS2PW_140 121 11.5 0.173 0.180 0.825 1.355 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=65.311000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=688.454545, yerr=7.426778 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b234498075e47956c66564e93abbe409d5e88a7d_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 826.834944E+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 0.015 0.034 54.713 54.715 qs_mol_dyn_low 1 2.0 0.003 0.003 54.482 54.488 qs_forces 11 3.9 0.003 0.005 54.408 54.411 qs_energies 11 4.9 0.003 0.013 50.729 50.733 scf_env_do_scf 11 5.9 0.000 0.001 42.552 42.553 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 34.909 34.910 velocity_verlet 10 3.0 0.001 0.002 30.888 30.891 dbcsr_multiply_generic 2055 12.4 0.102 0.107 22.450 22.587 qs_scf_new_mos 99 7.5 0.001 0.001 19.895 19.940 qs_scf_loop_do_ot 99 8.5 0.001 0.001 19.894 19.939 ot_scf_mini 99 9.5 0.002 0.002 18.672 18.701 multiply_cannon 2055 13.4 0.244 0.260 17.190 18.348 multiply_cannon_loop 2055 14.4 0.319 0.330 15.901 16.226 rebuild_ks_matrix 110 8.3 0.000 0.000 11.889 11.921 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 11.888 11.920 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.565 10.593 ot_mini 99 10.5 0.001 0.001 10.185 10.209 multiply_cannon_multrec 8220 15.4 3.216 4.918 7.407 8.673 init_scf_loop 11 6.9 0.001 0.001 7.597 7.598 mp_waitall_1 103326 16.6 5.934 7.516 5.934 7.516 sum_up_and_integrate 110 10.3 0.080 0.082 6.511 6.523 integrate_v_rspace 110 11.3 0.003 0.004 6.430 6.441 qs_ot_get_derivative 99 11.5 0.001 0.001 6.355 6.383 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.261 6.276 calculate_rho_elec 110 8.6 0.115 0.116 6.260 6.275 prepare_preconditioner 11 7.9 0.000 0.000 5.952 5.956 make_preconditioner 11 8.9 0.000 0.000 5.952 5.956 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.531 5.604 init_scf_run 11 5.9 0.000 0.001 5.002 5.002 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.002 5.002 dbcsr_mm_accdrv_process 17442 15.9 2.876 3.894 4.061 4.970 qs_ot_get_p 110 10.4 0.001 0.001 4.604 4.651 multiply_cannon_metrocomm3 8220 15.4 0.017 0.018 3.073 4.420 make_m2s 4110 13.4 0.039 0.040 4.055 4.282 make_images 4110 14.4 0.642 0.698 3.929 4.156 pw_transfer 1331 11.6 0.066 0.071 3.918 3.928 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.811 3.824 ot_diis_step 99 11.5 0.012 0.012 3.810 3.811 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.721 3.772 apply_single 110 13.6 0.000 0.000 3.720 3.771 grid_integrate_task_list 110 12.3 3.358 3.584 3.358 3.584 density_rs2pw 110 9.6 0.004 0.004 3.007 3.384 fft_wrap_pw1pw2_140 451 13.1 0.837 0.849 3.299 3.314 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.293 3.297 multiply_cannon_sync_h2d 8220 15.4 2.900 3.071 2.900 3.071 cp_dbcsr_syevd 48 12.0 0.003 0.003 2.999 2.999 cp_fm_cholesky_invert 11 10.9 2.872 2.876 2.872 2.876 wfi_extrapolate 11 7.9 0.001 0.001 2.704 2.704 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.677 2.680 make_images_data 4110 15.4 0.038 0.043 2.293 2.672 hybrid_alltoall_any 4261 16.3 0.201 0.863 2.234 2.593 rs_pw_transfer 902 11.9 0.010 0.011 2.157 2.559 grid_collocate_task_list 110 9.6 2.277 2.536 2.277 2.536 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.490 2.498 cp_fm_diag_elpa_base 48 14.0 2.433 2.457 2.489 2.496 calculate_dm_sparse 110 9.5 0.001 0.001 2.446 2.482 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.459 2.461 fft3d_ps 1111 14.6 1.130 1.144 2.350 2.367 calculate_first_density_matrix 1 7.0 0.001 0.001 2.204 2.206 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.117 2.132 potential_pw2rs 110 12.3 0.015 0.015 2.048 2.051 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.774 1.993 mp_alltoall_d11v 2046 13.8 1.611 1.963 1.611 1.963 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.843 1.858 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.734 1.749 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.654 1.659 cp_fm_cholesky_decompose 22 10.9 1.635 1.657 1.635 1.657 qs_env_update_s_mstruct 11 6.9 0.001 0.002 1.513 1.636 dbcsr_complete_redistribute 325 12.2 0.564 0.583 1.421 1.519 mp_allgather_i34 2055 14.4 0.476 1.508 0.476 1.508 mp_waitany 9240 13.8 1.060 1.485 1.060 1.485 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.433 1.443 jit_kernel_multiply 8 15.6 0.875 1.371 0.875 1.371 multiply_cannon_metrocomm4 6165 15.4 0.018 0.019 0.484 1.363 qs_create_task_list 11 7.9 0.003 0.003 1.222 1.326 generate_qs_task_list 11 8.9 0.373 0.441 1.219 1.323 mp_irecv_dv 24056 15.7 0.460 1.322 0.460 1.322 rs_gather_matrices 110 12.3 0.327 0.369 0.935 1.269 rs_pw_transfer_RS2PW_140 121 11.5 0.162 0.167 0.800 1.201 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.113 1.132 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=54.715000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=778.454545, yerr=10.603290 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b234498075e47956c66564e93abbe409d5e88a7d_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 1.332806E+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 0.019 0.039 87.056 87.058 qs_mol_dyn_low 1 2.0 0.003 0.003 86.756 86.765 qs_forces 11 3.9 0.003 0.004 86.376 86.376 qs_energies 11 4.9 0.002 0.002 82.196 82.197 scf_env_do_scf 11 5.9 0.001 0.001 72.296 72.296 velocity_verlet 10 3.0 0.002 0.002 55.955 56.000 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 43.903 43.905 init_scf_loop 11 6.9 0.001 0.001 28.323 28.326 dbcsr_multiply_generic 2055 12.4 0.117 0.122 28.273 28.316 prepare_preconditioner 11 7.9 0.000 0.000 26.286 26.293 make_preconditioner 11 8.9 0.000 0.000 26.286 26.293 qs_scf_new_mos 99 7.5 0.001 0.001 25.763 25.792 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.763 25.791 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.436 25.753 ot_scf_mini 99 9.5 0.002 0.002 24.009 24.032 multiply_cannon 2055 13.4 0.345 0.370 21.270 21.976 multiply_cannon_loop 2055 14.4 0.341 0.343 19.505 19.911 cp_fm_upper_to_full 70 14.2 12.719 18.283 12.719 18.283 rebuild_ks_matrix 110 8.3 0.001 0.001 14.057 14.092 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 14.057 14.091 ot_mini 99 10.5 0.001 0.001 13.357 13.375 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.735 12.766 dbcsr_complete_redistribute 325 12.2 1.030 1.057 7.402 10.601 multiply_cannon_multrec 8220 15.4 4.360 4.521 9.532 9.634 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.385 9.587 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.835 9.013 qs_ot_get_derivative 99 11.5 0.001 0.001 8.772 8.794 mp_alltoall_i22 605 13.7 5.438 8.632 5.438 8.632 mp_waitall_1 84994 16.7 7.497 8.306 7.497 8.306 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.656 7.693 calculate_rho_elec 110 8.6 0.228 0.228 7.655 7.693 sum_up_and_integrate 110 10.3 0.150 0.153 7.257 7.269 integrate_v_rspace 110 11.3 0.004 0.004 7.107 7.119 make_m2s 4110 13.4 0.043 0.044 5.323 5.810 init_scf_run 11 5.9 0.000 0.001 5.642 5.642 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.642 5.642 make_images 4110 14.4 0.885 0.925 5.137 5.624 qs_ot_get_p 110 10.4 0.001 0.001 5.461 5.481 dbcsr_mm_accdrv_process 11614 15.7 3.303 3.654 5.029 5.276 cp_fm_cholesky_invert 11 10.9 5.242 5.246 5.242 5.246 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.685 5.137 apply_single 110 13.6 0.000 0.000 4.685 5.137 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 4.633 5.073 pw_transfer 1331 11.6 0.075 0.076 5.005 5.013 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 4.888 4.897 ot_diis_step 99 11.5 0.015 0.015 4.561 4.561 fft_wrap_pw1pw2_140 451 13.1 1.337 1.343 4.267 4.279 qs_ot_p2m_diag 48 11.0 0.151 0.156 3.966 3.972 multiply_cannon_sync_h2d 8220 15.4 3.945 3.954 3.945 3.954 grid_integrate_task_list 110 12.3 3.664 3.735 3.664 3.735 density_rs2pw 110 9.6 0.004 0.004 3.692 3.735 hybrid_alltoall_any 4261 16.3 0.258 0.552 2.963 3.727 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.671 3.671 make_images_data 4110 15.4 0.042 0.045 2.977 3.595 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.530 3.531 wfi_extrapolate 11 7.9 0.001 0.001 3.285 3.285 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.803 3.248 calculate_dm_sparse 110 9.5 0.001 0.001 3.112 3.133 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.943 2.943 cp_fm_diag_elpa_base 48 14.0 2.401 2.594 2.941 2.942 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.833 2.836 fft3d_ps 1111 14.6 1.300 1.309 2.815 2.823 grid_collocate_task_list 110 9.6 2.631 2.657 2.631 2.657 potential_pw2rs 110 12.3 0.021 0.021 2.380 2.385 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.352 2.364 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.322 2.332 rs_pw_transfer 902 11.9 0.011 0.011 2.234 2.267 qs_env_update_s_mstruct 11 6.9 0.012 0.021 2.195 2.247 calculate_first_density_matrix 1 7.0 0.001 0.001 2.239 2.240 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.103 2.187 cp_fm_cholesky_decompose 22 10.9 1.971 1.995 1.971 1.995 mp_alltoall_d11v 2046 13.8 1.892 1.974 1.892 1.974 qs_create_task_list 11 7.9 0.001 0.001 1.889 1.942 generate_qs_task_list 11 8.9 0.731 0.784 1.889 1.941 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.928 1.932 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.726 1.753 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=87.058000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1196.000000, yerr=63.446470 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b234498075e47956c66564e93abbe409d5e88a7d_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 630.390784E+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 2386084. 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 0.031 204.855 204.856 qs_mol_dyn_low 1 2.0 0.003 0.003 204.487 204.500 qs_forces 11 3.9 0.005 0.005 204.350 204.351 qs_energies 11 4.9 0.002 0.002 198.786 198.798 scf_env_do_scf 11 5.9 0.001 0.001 182.360 182.364 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 162.073 162.075 dbcsr_multiply_generic 2507 12.6 0.174 0.177 123.561 124.382 velocity_verlet 10 3.0 0.001 0.002 122.790 122.791 qs_scf_new_mos 117 7.6 0.001 0.001 122.219 122.446 qs_scf_loop_do_ot 117 8.6 0.001 0.001 122.218 122.445 ot_scf_mini 117 9.6 0.003 0.003 115.652 115.862 multiply_cannon 2507 13.6 0.241 0.250 100.720 102.546 multiply_cannon_loop 2507 14.6 2.069 2.128 98.617 100.539 ot_mini 117 10.6 0.001 0.001 65.017 65.229 multiply_cannon_multrec 60168 15.6 33.345 35.538 41.468 42.983 qs_ot_get_derivative 117 11.6 0.001 0.001 40.272 40.478 rebuild_ks_matrix 128 8.3 0.001 0.001 33.609 34.020 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.017 33.609 34.019 mp_waitall_1 267128 16.5 27.973 30.879 27.973 30.879 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.189 30.534 multiply_cannon_sync_h2d 60168 15.6 27.684 29.791 27.684 29.791 qs_ot_get_p 128 10.4 0.001 0.001 28.761 29.099 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.318 25.091 apply_single 128 13.6 0.001 0.001 24.317 25.090 ot_diis_step 117 11.6 0.007 0.008 24.536 24.537 qs_ot_p2m_diag 83 11.4 0.079 0.091 22.274 22.326 init_scf_loop 11 6.9 0.001 0.001 20.215 20.217 cp_dbcsr_syevd 83 12.4 0.005 0.005 19.831 19.832 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.002 19.176 multiply_cannon_metrocomm3 60168 15.6 0.113 0.118 15.577 18.442 cp_fm_diag_elpa 83 13.4 0.000 0.001 16.930 16.962 cp_fm_diag_elpa_base 83 14.4 16.862 16.894 16.926 16.956 prepare_preconditioner 11 7.9 0.000 0.000 15.610 15.655 make_preconditioner 11 8.9 0.000 0.000 15.610 15.655 make_full_inverse_cholesky 11 9.9 0.000 0.000 14.866 15.068 sum_up_and_integrate 128 10.3 0.090 0.108 14.133 14.146 make_m2s 5014 13.6 0.103 0.112 13.746 14.133 integrate_v_rspace 128 11.3 0.004 0.005 14.042 14.060 make_images 5014 14.6 0.403 0.420 13.572 13.968 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.607 13.792 calculate_rho_elec 128 8.7 0.046 0.065 13.607 13.791 init_scf_run 11 5.9 0.000 0.001 12.318 12.319 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.318 12.318 density_rs2pw 128 9.7 0.006 0.007 6.947 10.135 mp_sum_l 7870 13.0 7.825 9.038 7.825 9.038 wfi_extrapolate 11 7.9 0.001 0.001 8.954 8.954 cp_fm_cholesky_invert 11 10.9 8.886 8.893 8.886 8.893 rs_pw_transfer 1046 11.9 0.017 0.020 5.585 8.796 calculate_dm_sparse 128 9.5 0.001 0.001 8.470 8.548 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.088 8.220 dbcsr_mm_accdrv_process 124484 16.2 3.088 3.229 7.683 8.215 pw_transfer 1547 11.6 0.075 0.095 7.706 7.880 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.808 7.872 multiply_cannon_metrocomm1 60168 15.6 0.088 0.093 5.969 7.738 make_images_data 5014 15.6 0.069 0.075 6.773 7.726 fft_wrap_pw1pw2 1291 12.7 0.010 0.014 7.502 7.674 grid_integrate_task_list 128 12.3 7.085 7.571 7.085 7.571 hybrid_alltoall_any 5200 16.5 0.294 2.248 5.932 7.039 fft_wrap_pw1pw2_140 523 13.2 1.283 1.325 6.608 6.791 cp_dbcsr_sm_fm_multiply 37 9.5 0.003 0.003 6.676 6.682 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.805 6.009 mp_waitany 16020 13.9 2.633 5.812 2.633 5.812 grid_collocate_task_list 128 9.7 4.555 5.778 4.555 5.778 mp_alltoall_d11v 2415 14.1 4.205 5.704 4.205 5.704 fft3d_ps 1291 14.7 2.129 2.769 5.254 5.555 rs_pw_transfer_RS2PW_140 139 11.5 0.282 0.301 2.109 5.321 cp_fm_cholesky_decompose 22 10.9 4.565 4.580 4.565 4.580 potential_pw2rs 128 12.3 0.009 0.011 4.508 4.527 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=204.856000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=596.636364, yerr=6.568142 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b234498075e47956c66564e93abbe409d5e88a7d_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 828.657664E+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 4010 58106. MP_Allreduce 11023 966. MP_Sync 87 MP_Alltoall 1969 5722441. 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 0.036 0.077 188.828 188.830 qs_mol_dyn_low 1 2.0 0.003 0.004 188.267 188.414 qs_forces 11 3.9 0.005 0.005 188.181 188.182 qs_energies 11 4.9 0.003 0.010 181.345 181.358 scf_env_do_scf 11 5.9 0.001 0.001 165.066 165.075 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 132.653 132.656 velocity_verlet 10 3.0 0.002 0.002 118.662 118.664 dbcsr_multiply_generic 2507 12.6 0.184 0.189 97.001 98.136 qs_scf_new_mos 117 7.6 0.001 0.001 92.643 93.038 qs_scf_loop_do_ot 117 8.6 0.001 0.001 92.642 93.037 ot_scf_mini 117 9.6 0.004 0.005 87.865 88.355 multiply_cannon 2507 13.6 0.479 0.531 77.261 81.849 multiply_cannon_loop 2507 14.6 1.249 1.285 73.959 76.285 ot_mini 117 10.6 0.001 0.001 49.407 49.857 mp_waitall_1 214728 16.6 24.179 38.786 24.179 38.786 multiply_cannon_multrec 30084 15.6 22.055 25.973 31.630 35.850 rebuild_ks_matrix 128 8.3 0.001 0.001 33.432 33.947 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.022 33.432 33.947 init_scf_loop 11 6.9 0.001 0.005 32.326 32.328 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.039 30.493 multiply_cannon_metrocomm3 30084 15.6 0.090 0.097 15.576 28.449 qs_ot_get_derivative 117 11.6 0.001 0.002 27.498 27.976 prepare_preconditioner 11 7.9 0.000 0.000 27.875 27.933 make_preconditioner 11 8.9 0.001 0.002 27.875 27.933 make_full_inverse_cholesky 11 9.9 0.000 0.000 26.577 27.095 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 21.941 22.986 apply_single 128 13.6 0.001 0.001 21.940 22.985 ot_diis_step 117 11.6 0.014 0.014 21.735 21.737 multiply_cannon_sync_h2d 30084 15.6 19.301 21.732 19.301 21.732 qs_ot_get_p 128 10.4 0.001 0.002 20.274 20.905 cp_fm_cholesky_invert 11 10.9 16.334 16.346 16.334 16.346 make_m2s 5014 13.6 0.090 0.096 14.155 15.817 make_images 5014 14.6 1.168 1.360 13.956 15.617 qs_ot_p2m_diag 83 11.4 0.187 0.216 15.519 15.551 sum_up_and_integrate 128 10.3 0.116 0.133 14.692 14.733 integrate_v_rspace 128 11.3 0.004 0.005 14.575 14.619 cp_dbcsr_syevd 83 12.4 0.005 0.006 14.394 14.395 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.960 13.990 calculate_rho_elec 128 8.7 0.089 0.106 13.960 13.989 init_scf_run 11 5.9 0.000 0.001 11.455 11.456 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.455 11.456 cp_fm_diag_elpa 83 13.4 0.001 0.001 11.299 11.327 cp_fm_diag_elpa_base 83 14.4 11.043 11.116 11.293 11.324 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 10.821 11.118 make_images_data 5014 15.6 0.065 0.071 8.624 10.552 multiply_cannon_metrocomm4 27577 15.6 0.099 0.113 3.764 10.522 mp_irecv_dv 69486 16.3 3.568 10.129 3.568 10.129 hybrid_alltoall_any 5200 16.5 0.343 1.498 7.311 9.890 density_rs2pw 128 9.7 0.006 0.008 7.347 9.790 dbcsr_mm_accdrv_process 62242 16.2 4.411 5.083 9.025 9.619 pw_transfer 1547 11.6 0.086 0.095 8.792 8.862 fft_wrap_pw1pw2 1291 12.7 0.011 0.011 8.567 8.636 wfi_extrapolate 11 7.9 0.001 0.001 8.355 8.355 rs_pw_transfer 1046 11.9 0.015 0.017 5.617 8.084 fft_wrap_pw1pw2_140 523 13.2 1.338 1.364 7.599 7.677 grid_integrate_task_list 128 12.3 7.175 7.559 7.175 7.559 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.130 6.918 cp_fm_cholesky_decompose 22 10.9 6.723 6.806 6.723 6.806 calculate_dm_sparse 128 9.5 0.001 0.001 6.436 6.585 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.126 6.135 fft3d_ps 1291 14.7 2.820 2.987 5.872 5.940 grid_collocate_task_list 128 9.7 4.701 5.852 4.701 5.852 mp_sum_l 7870 13.0 3.839 5.736 3.839 5.736 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.251 5.405 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.283 5.361 mp_alltoall_d11v 2415 14.1 4.063 4.994 4.063 4.994 mp_allgather_i34 2507 14.6 1.864 4.964 1.864 4.964 potential_pw2rs 128 12.3 0.015 0.018 4.921 4.949 mp_waitany 11748 13.9 2.485 4.934 2.485 4.934 rs_pw_transfer_RS2PW_140 139 11.5 0.361 0.409 2.091 4.527 mp_sum_d 4474 12.1 2.671 4.013 2.671 4.013 dbcsr_complete_redistribute 395 12.7 0.777 0.866 3.071 3.895 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.784 3.841 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=188.830000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=789.727273, yerr=2.259964 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b234498075e47956c66564e93abbe409d5e88a7d_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 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 92.796578E+12 0.0% 0.0% 100.0% flops max/rank 2.906046E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500608 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 957.374464E+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 925172869264 16777216 < size 28448 745747251200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4003 58198. MP_Allreduce 11005 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 0.016 0.032 173.523 173.523 qs_mol_dyn_low 1 2.0 0.003 0.003 173.122 173.129 qs_forces 11 3.9 0.004 0.005 172.988 172.996 qs_energies 11 4.9 0.001 0.002 166.383 166.393 scf_env_do_scf 11 5.9 0.001 0.001 150.920 150.921 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 116.103 116.104 velocity_verlet 10 3.0 0.002 0.002 111.118 111.120 dbcsr_multiply_generic 2485 12.5 0.174 0.179 80.415 81.506 qs_scf_new_mos 116 7.6 0.001 0.001 78.325 78.715 qs_scf_loop_do_ot 116 8.6 0.001 0.001 78.324 78.714 ot_scf_mini 116 9.6 0.003 0.004 74.234 74.659 multiply_cannon 2485 13.5 0.494 0.519 60.794 64.619 multiply_cannon_loop 2485 14.5 0.847 0.883 58.016 60.669 ot_mini 116 10.6 0.001 0.001 41.905 42.350 init_scf_loop 11 6.9 0.001 0.001 34.721 34.722 mp_waitall_1 169034 16.6 24.871 34.037 24.871 34.037 rebuild_ks_matrix 127 8.3 0.001 0.001 30.480 31.023 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.018 30.480 31.022 prepare_preconditioner 11 7.9 0.000 0.000 30.706 30.754 make_preconditioner 11 8.9 0.000 0.000 30.706 30.754 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.301 29.759 qs_ks_update_qs_env 127 7.6 0.001 0.001 27.422 27.916 multiply_cannon_multrec 19880 15.5 13.347 16.986 21.928 25.537 multiply_cannon_metrocomm3 19880 15.5 0.057 0.061 15.170 24.330 qs_ot_get_derivative 116 11.6 0.001 0.002 22.288 22.721 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.730 20.722 apply_single 127 13.6 0.001 0.001 19.730 20.721 ot_diis_step 116 11.6 0.017 0.018 19.513 19.513 qs_ot_get_p 127 10.4 0.001 0.001 17.915 18.440 multiply_cannon_sync_h2d 19880 15.5 14.115 15.953 14.115 15.953 make_m2s 4970 13.5 0.082 0.089 14.760 15.435 make_images 4970 14.5 1.166 1.263 14.533 15.202 sum_up_and_integrate 127 10.3 0.131 0.145 14.487 14.514 integrate_v_rspace 127 11.3 0.004 0.005 14.355 14.386 cp_fm_cholesky_invert 11 10.9 14.182 14.190 14.182 14.190 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.143 14.185 calculate_rho_elec 127 8.7 0.131 0.147 14.143 14.185 qs_ot_p2m_diag 82 11.4 0.262 0.269 13.796 13.803 cp_dbcsr_syevd 82 12.4 0.005 0.006 12.815 12.816 init_scf_run 11 5.9 0.000 0.001 10.460 10.460 scf_env_initial_rho_setup 11 6.9 0.000 0.001 10.460 10.460 make_images_data 4970 15.5 0.060 0.068 9.244 10.368 cp_fm_diag_elpa 82 13.4 0.000 0.001 9.727 9.742 cp_fm_diag_elpa_base 82 14.4 9.319 9.473 9.724 9.739 hybrid_alltoall_any 5155 16.4 0.434 1.981 8.025 9.628 multiply_cannon_metrocomm4 17395 15.5 0.063 0.072 3.468 9.290 mp_irecv_dv 49801 16.2 3.346 9.046 3.346 9.046 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 8.667 9.002 pw_transfer 1535 11.6 0.085 0.100 8.746 8.843 density_rs2pw 127 9.7 0.006 0.007 7.035 8.702 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.523 8.623 dbcsr_mm_accdrv_process 41158 16.2 4.463 5.208 8.039 8.150 grid_integrate_task_list 127 12.3 7.247 7.727 7.247 7.727 fft_wrap_pw1pw2_140 519 13.2 1.408 1.445 7.582 7.696 cp_fm_upper_to_full 104 14.8 5.839 7.415 5.839 7.415 wfi_extrapolate 11 7.9 0.001 0.001 7.372 7.372 cp_fm_cholesky_decompose 22 10.9 7.227 7.255 7.227 7.255 rs_pw_transfer 1038 11.9 0.014 0.015 5.074 6.761 dbcsr_complete_redistribute 393 12.7 1.221 1.265 4.665 6.525 calculate_dm_sparse 127 9.5 0.001 0.001 5.690 5.786 grid_collocate_task_list 127 9.7 4.876 5.757 4.876 5.757 fft3d_ps 1281 14.7 2.716 2.930 5.660 5.728 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.429 5.435 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.505 5.363 mp_alltoall_d11v 2401 14.1 4.291 5.259 4.291 5.259 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.501 5.205 potential_pw2rs 127 12.3 0.020 0.022 4.646 4.659 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.490 4.626 mp_sum_l 7804 13.0 3.066 4.594 3.066 4.594 mp_allgather_i34 2485 14.5 1.303 4.443 1.303 4.443 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.386 4.226 mp_waitany 11660 13.9 2.331 4.106 2.331 4.106 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.992 4.027 mp_alltoall_i22 712 14.1 1.989 4.025 1.989 4.025 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 3.771 3.775 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.552 3.601 rs_pw_transfer_RS2PW_140 138 11.5 0.334 0.355 1.912 3.586 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=173.523000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=895.545455, yerr=13.405315 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b234498075e47956c66564e93abbe409d5e88a7d_performance_tests/18/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 4.387043E+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 6024576 0.0% 0.0% 100.0% average stack size 0.0 0.0 1129.8 marketing flops 145.647559E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 1.150198E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1152312 MPI messages size (bytes): total size 2.039354E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.769794E+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 321600 36333158400 131072 < size <= 4194304 721560 792041881600 4194304 < size <= 16777216 70800 669921260160 16777216 < size 30960 541065216000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4024 57899. MP_Allreduce 11057 1088. MP_Sync 87 MP_Alltoall 1724 12509439. MP_SendRecv 5934 75008. MP_ISendRecv 5934 75008. MP_Wait 22612 MP_ISend 15064 244788. MP_IRecv 15064 244788. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.035 187.690 187.691 qs_mol_dyn_low 1 2.0 0.003 0.003 187.349 187.360 qs_forces 11 3.9 0.004 0.005 187.241 187.247 qs_energies 11 4.9 0.001 0.002 180.075 180.088 scf_env_do_scf 11 5.9 0.001 0.001 163.541 163.547 velocity_verlet 10 3.0 0.002 0.002 125.011 125.014 scf_env_do_scf_inner_loop 118 6.6 0.003 0.009 117.436 117.438 dbcsr_multiply_generic 2527 12.6 0.185 0.191 79.769 80.678 qs_scf_new_mos 118 7.6 0.001 0.001 79.552 79.829 qs_scf_loop_do_ot 118 8.6 0.001 0.001 79.551 79.828 ot_scf_mini 118 9.6 0.003 0.003 75.038 75.341 multiply_cannon 2527 13.6 0.561 0.597 54.903 57.922 multiply_cannon_loop 2527 14.6 1.187 1.222 51.322 53.378 init_scf_loop 11 6.9 0.000 0.001 45.984 45.987 ot_mini 118 10.6 0.001 0.001 42.602 42.885 prepare_preconditioner 11 7.9 0.000 0.000 41.842 41.863 make_preconditioner 11 8.9 0.000 0.000 41.842 41.863 make_full_inverse_cholesky 11 9.9 0.000 0.000 35.443 40.513 multiply_cannon_multrec 30324 15.6 14.403 19.382 26.332 30.912 rebuild_ks_matrix 129 8.3 0.001 0.001 30.205 30.449 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.020 30.204 30.449 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.258 27.475 mp_waitall_1 149060 16.7 17.388 27.092 17.388 27.092 qs_ot_get_derivative 118 11.6 0.002 0.002 22.812 23.106 make_m2s 5054 13.6 0.099 0.101 20.490 21.456 make_images 5054 14.6 1.980 2.273 20.183 21.148 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 19.107 19.744 apply_single 129 13.6 0.001 0.001 19.106 19.744 ot_diis_step 118 11.6 0.017 0.018 19.653 19.655 qs_ot_get_p 129 10.4 0.001 0.001 18.564 18.887 cp_fm_upper_to_full 105 14.8 11.580 17.046 11.580 17.046 cp_fm_cholesky_invert 11 10.9 15.704 15.713 15.704 15.713 multiply_cannon_metrocomm3 30324 15.6 0.047 0.050 6.373 15.332 sum_up_and_integrate 129 10.3 0.141 0.152 14.785 14.808 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.646 14.689 calculate_rho_elec 129 8.7 0.177 0.193 14.645 14.688 integrate_v_rspace 129 11.3 0.004 0.004 14.644 14.671 qs_ot_p2m_diag 83 11.4 0.343 0.390 14.427 14.479 cp_dbcsr_syevd 83 12.4 0.005 0.006 13.172 13.173 make_images_data 5054 15.6 0.065 0.069 11.128 12.756 dbcsr_complete_redistribute 395 12.7 1.515 1.643 8.910 12.705 multiply_cannon_sync_h2d 30324 15.6 11.855 12.555 11.855 12.555 dbcsr_mm_accdrv_process 62756 16.2 7.305 8.470 11.500 12.114 hybrid_alltoall_any 5240 16.5 0.530 2.231 9.819 11.703 copy_fm_to_dbcsr 209 11.7 0.002 0.002 7.583 11.385 init_scf_run 11 5.9 0.000 0.001 10.685 10.686 scf_env_initial_rho_setup 11 6.9 0.000 0.001 10.685 10.686 cp_fm_diag_elpa 83 13.4 0.001 0.001 10.137 10.147 cp_fm_diag_elpa_base 83 14.4 9.162 9.473 10.131 10.140 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.377 10.098 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.353 9.571 mp_alltoall_i22 718 14.1 5.623 9.432 5.623 9.432 pw_transfer 1559 11.6 0.088 0.103 9.216 9.286 fft_wrap_pw1pw2 1301 12.7 0.011 0.012 8.987 9.063 density_rs2pw 129 9.7 0.006 0.007 6.998 8.516 fft_wrap_pw1pw2_140 527 13.2 1.576 1.609 8.022 8.116 grid_integrate_task_list 129 12.3 7.574 8.067 7.574 8.067 wfi_extrapolate 11 7.9 0.001 0.001 7.658 7.658 cp_fm_cholesky_decompose 22 10.9 7.412 7.512 7.412 7.512 multiply_cannon_metrocomm4 25270 15.6 0.084 0.096 2.750 6.795 mp_irecv_dv 76685 16.2 2.599 6.525 2.599 6.525 calculate_dm_sparse 129 9.5 0.001 0.001 6.184 6.275 rs_pw_transfer 1054 12.0 0.013 0.014 4.594 6.171 fft3d_ps 1301 14.7 2.856 2.926 5.884 5.927 grid_collocate_task_list 129 9.7 5.107 5.825 5.107 5.825 mp_alltoall_d11v 2423 14.1 4.830 5.703 4.830 5.703 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.382 5.436 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 4.456 4.543 potential_pw2rs 129 12.3 0.024 0.024 4.488 4.503 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 4.425 4.427 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.349 4.419 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.170 4.229 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=187.691000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1069.000000, yerr=23.786933 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b234498075e47956c66564e93abbe409d5e88a7d_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 1.523184E+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 0.032 0.050 173.689 173.691 qs_mol_dyn_low 1 2.0 0.003 0.003 173.292 173.302 qs_forces 11 3.9 0.010 0.011 173.174 173.180 qs_energies 11 4.9 0.017 0.032 165.561 165.570 scf_env_do_scf 11 5.9 0.001 0.001 148.409 148.412 velocity_verlet 10 3.0 0.002 0.002 114.343 114.347 scf_env_do_scf_inner_loop 118 6.6 0.003 0.009 112.804 112.805 dbcsr_multiply_generic 2527 12.6 0.178 0.188 74.677 75.157 qs_scf_new_mos 118 7.6 0.001 0.001 73.981 74.094 qs_scf_loop_do_ot 118 8.6 0.001 0.001 73.981 74.093 ot_scf_mini 118 9.6 0.004 0.004 69.538 69.629 multiply_cannon 2527 13.6 0.593 0.634 54.769 59.320 multiply_cannon_loop 2527 14.6 0.446 0.461 50.036 50.883 ot_mini 118 10.6 0.001 0.001 39.739 39.826 init_scf_loop 11 6.9 0.001 0.001 35.460 35.462 mp_waitall_1 126780 16.7 26.020 33.614 26.020 33.614 prepare_preconditioner 11 7.9 0.000 0.000 31.460 31.479 make_preconditioner 11 8.9 0.000 0.000 31.460 31.479 rebuild_ks_matrix 129 8.3 0.001 0.001 30.163 30.268 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.019 30.163 30.267 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.379 29.656 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.384 27.484 multiply_cannon_multrec 10108 15.6 10.526 14.548 17.880 20.770 ot_diis_step 118 11.6 0.020 0.021 20.127 20.127 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.654 19.974 apply_single 129 13.6 0.001 0.001 19.653 19.973 qs_ot_get_derivative 118 11.6 0.002 0.002 19.545 19.636 multiply_cannon_metrocomm3 10108 15.6 0.023 0.024 12.303 19.313 make_m2s 5054 13.6 0.068 0.071 16.366 18.798 make_images 5054 14.6 2.328 2.820 16.062 18.499 cp_fm_cholesky_invert 11 10.9 17.969 17.975 17.969 17.975 qs_ot_get_p 129 10.4 0.001 0.001 16.697 16.781 qs_rho_update_rho_low 129 7.7 0.001 0.001 15.366 15.403 calculate_rho_elec 129 8.7 0.261 0.271 15.365 15.402 sum_up_and_integrate 129 10.3 0.183 0.193 15.110 15.151 integrate_v_rspace 129 11.3 0.004 0.005 14.926 14.977 qs_ot_p2m_diag 83 11.4 0.496 0.502 12.951 12.969 make_images_data 5054 15.6 0.053 0.061 10.029 12.709 hybrid_alltoall_any 5240 16.5 0.823 3.645 9.869 12.054 multiply_cannon_sync_h2d 10108 15.6 11.728 12.032 11.728 12.032 cp_dbcsr_syevd 83 12.4 0.005 0.006 11.838 11.839 init_scf_run 11 5.9 0.000 0.001 10.304 10.305 scf_env_initial_rho_setup 11 6.9 0.000 0.001 10.304 10.305 pw_transfer 1559 11.6 0.087 0.096 9.503 9.524 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 9.277 9.306 cp_fm_diag_elpa 83 13.4 0.000 0.001 8.880 8.891 cp_fm_diag_elpa_base 83 14.4 8.633 8.718 8.873 8.885 fft_wrap_pw1pw2_140 527 13.2 1.929 1.965 8.237 8.270 grid_integrate_task_list 129 12.3 7.807 8.202 7.807 8.202 cp_fm_cholesky_decompose 22 10.9 7.847 7.965 7.847 7.965 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 7.804 7.877 density_rs2pw 129 9.7 0.006 0.007 6.856 7.754 dbcsr_mm_accdrv_process 20926 16.1 2.669 3.468 6.977 7.613 mp_allgather_i34 2527 14.6 2.911 7.538 2.911 7.538 multiply_cannon_metrocomm1 10108 15.6 0.028 0.029 4.475 7.420 wfi_extrapolate 11 7.9 0.001 0.001 7.389 7.389 calculate_dm_sparse 129 9.5 0.001 0.001 6.086 6.164 grid_collocate_task_list 129 9.7 5.432 6.108 5.432 6.108 mp_alltoall_d11v 2423 14.1 4.983 5.912 4.983 5.912 fft3d_ps 1301 14.7 2.784 2.856 5.722 5.752 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.292 5.293 dbcsr_complete_redistribute 395 12.7 2.131 2.194 5.047 5.289 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.160 5.171 rs_pw_transfer 1054 12.0 0.013 0.014 4.083 5.012 potential_pw2rs 129 12.3 0.026 0.026 4.417 4.429 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.108 4.131 multiply_cannon_metrocomm4 7581 15.6 0.024 0.028 1.877 3.948 mp_irecv_dv 29086 15.9 1.841 3.877 1.841 3.877 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.553 3.866 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.695 3.726 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.691 3.701 copy_fm_to_dbcsr 209 11.7 0.001 0.002 3.312 3.611 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.474 3.560 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 3.476 3.506 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=173.691000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1427.636364, yerr=56.911850 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b234498075e47956c66564e93abbe409d5e88a7d_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456868864 0.0% 0.0% 100.0% flops 32 x 32 x 32 1960652570624 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986252263424 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992007581696 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753954734080 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 11613074706432 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176888320 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176888320 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911117062144 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.230858E+12 0.0% 0.0% 100.0% flops max/rank 11.785799E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806351296 0.0% 0.0% 100.0% number of processed stacks 1979920 0.0% 0.0% 100.0% average stack size 0.0 0.0 3437.7 marketing flops 145.648783E+12 ------------------------------------------------------------------------------- # multiplications 2528 max memory usage/rank 3.116397E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101120 MPI messages size (bytes): total size 1.144936E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.322546E+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 45608 35399925760 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592330608 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4069 58204. MP_Allreduce 11170 1510. MP_Sync 86 MP_Alltoall 1724 36993574. 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 0.036 0.065 296.945 296.951 qs_mol_dyn_low 1 2.0 0.003 0.003 296.409 296.415 qs_forces 11 3.9 0.005 0.005 296.297 296.300 qs_energies 11 4.9 0.005 0.005 287.005 287.010 scf_env_do_scf 11 5.9 0.001 0.001 264.858 264.870 velocity_verlet 10 3.0 0.003 0.004 214.140 214.148 scf_env_do_scf_inner_loop 118 6.6 0.004 0.009 140.484 140.486 init_scf_loop 11 6.9 0.001 0.001 124.122 124.125 prepare_preconditioner 11 7.9 0.000 0.000 119.115 119.137 make_preconditioner 11 8.9 0.000 0.000 119.115 119.137 make_full_inverse_cholesky 11 9.9 0.000 0.000 95.176 116.281 qs_scf_new_mos 118 7.6 0.001 0.001 89.510 89.632 qs_scf_loop_do_ot 118 8.6 0.001 0.001 89.509 89.631 ot_scf_mini 118 9.6 0.004 0.004 84.719 84.763 dbcsr_multiply_generic 2528 12.6 0.214 0.223 83.308 83.883 cp_fm_upper_to_full 104 14.8 53.203 76.097 53.203 76.097 multiply_cannon 2528 13.6 0.715 0.799 58.718 59.211 multiply_cannon_loop 2528 14.6 0.476 0.497 55.132 56.509 ot_mini 118 10.6 0.001 0.001 44.331 44.373 dbcsr_complete_redistribute 393 12.7 4.004 4.053 29.484 42.222 copy_fm_to_dbcsr 208 11.6 0.002 0.002 26.159 38.938 rebuild_ks_matrix 129 8.3 0.001 0.001 38.437 38.476 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.021 38.436 38.475 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 23.893 36.505 qs_ks_update_qs_env 129 7.6 0.001 0.001 35.484 35.522 mp_alltoall_i22 716 14.1 21.724 34.704 21.724 34.704 cp_fm_cholesky_invert 11 10.9 32.999 33.005 32.999 33.005 mp_waitall_1 104540 16.8 28.074 31.871 28.074 31.871 qs_ot_get_p 129 10.4 0.001 0.001 24.853 24.867 qs_ot_get_derivative 118 11.6 0.002 0.002 24.139 24.183 make_m2s 5056 13.6 0.079 0.082 20.031 20.903 qs_ot_p2m_diag 82 11.4 0.868 0.874 20.822 20.850 qs_rho_update_rho_low 129 7.7 0.001 0.001 20.483 20.503 calculate_rho_elec 129 8.7 0.487 0.487 20.482 20.502 make_images 5056 14.6 3.811 3.948 19.553 20.421 ot_diis_step 118 11.6 0.022 0.023 20.166 20.167 multiply_cannon_metrocomm3 10112 15.6 0.023 0.024 18.695 19.811 sum_up_and_integrate 129 10.3 0.324 0.325 19.634 19.716 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.333 19.544 apply_single 129 13.6 0.001 0.001 19.333 19.544 integrate_v_rspace 129 11.3 0.004 0.004 19.310 19.391 cp_dbcsr_syevd 82 12.4 0.006 0.006 19.141 19.142 multiply_cannon_multrec 10112 15.6 10.659 12.419 18.092 18.218 cp_fm_diag_elpa 82 13.4 0.000 0.000 16.033 16.033 cp_fm_diag_elpa_base 82 14.4 11.653 13.255 16.029 16.029 multiply_cannon_sync_h2d 10112 15.6 15.776 15.795 15.776 15.795 hybrid_alltoall_any 5241 16.5 1.310 3.065 10.891 12.693 make_images_data 5056 15.6 0.061 0.066 10.677 12.313 init_scf_run 11 5.9 0.000 0.001 12.124 12.124 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.124 12.124 pw_transfer 1559 11.6 0.094 0.095 11.319 11.330 fft_wrap_pw1pw2 1301 12.7 0.012 0.012 11.082 11.094 fft_wrap_pw1pw2_140 527 13.2 3.095 3.112 9.863 9.872 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.202 9.239 mp_alltoall_d11v 2417 14.1 8.290 9.098 8.290 9.098 dbcsr_mm_accdrv_process 20930 16.1 3.821 5.593 7.193 9.056 wfi_extrapolate 11 7.9 0.001 0.001 8.949 8.949 grid_integrate_task_list 129 12.3 8.641 8.805 8.641 8.805 cp_fm_cholesky_decompose 22 10.9 8.756 8.773 8.756 8.773 density_rs2pw 129 9.7 0.005 0.006 7.929 8.065 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 8.038 8.039 calculate_dm_sparse 129 9.5 0.001 0.001 6.614 6.732 grid_collocate_task_list 129 9.7 6.404 6.436 6.404 6.436 rs_scatter_matrices 140 9.7 3.703 4.696 6.109 6.336 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.250 6.297 fft3d_ps 1301 14.7 2.866 2.871 6.121 6.132 copy_dbcsr_to_fm 185 11.7 0.004 0.004 5.968 6.028 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=296.951000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2801.181818, yerr=170.993258 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b234498075e47956c66564e93abbe409d5e88a7d_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 1.262223E+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 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 0.078 0.107 85.191 85.192 qs_energies 1 2.0 0.000 0.000 84.596 84.614 ls_scf 1 3.0 0.000 0.000 83.653 83.667 dbcsr_multiply_generic 111 6.7 0.015 0.016 72.699 72.854 multiply_cannon 111 7.7 0.017 0.020 55.970 57.214 multiply_cannon_loop 111 8.7 0.212 0.224 52.543 53.849 ls_scf_main 1 4.0 0.000 0.000 52.307 52.308 density_matrix_trs4 2 5.0 0.002 0.003 46.805 46.889 ls_scf_init_scf 1 4.0 0.000 0.000 28.272 28.273 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.170 27.228 mp_waitall_1 11031 10.9 22.510 26.766 22.510 26.766 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.016 25.033 multiply_cannon_multrec 2664 9.7 8.188 8.875 15.612 17.299 multiply_cannon_sync_h2d 2664 9.7 13.549 15.414 13.549 15.414 make_m2s 222 7.7 0.009 0.012 13.061 13.667 make_images 222 8.7 0.102 0.112 13.039 13.646 multiply_cannon_metrocomm1 2664 9.7 0.009 0.011 9.766 12.532 make_images_data 222 9.7 0.005 0.006 7.619 8.167 dbcsr_mm_accdrv_process 4760 10.4 0.512 0.651 7.043 8.042 multiply_cannon_metrocomm3 2664 9.7 0.010 0.011 5.465 7.898 hybrid_alltoall_any 227 10.6 0.219 1.847 6.516 7.873 dbcsr_mm_accdrv_process_sort 4760 11.4 6.326 7.232 6.326 7.232 calculate_norms 4752 9.8 5.503 6.139 5.503 6.139 mp_sum_l 807 5.4 3.142 5.312 3.142 5.312 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.046 5.155 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.047 3.958 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.314 3.943 mp_irecv_dv 6231 10.9 2.031 3.928 2.031 3.928 make_images_sizes 222 9.7 0.000 0.000 0.688 3.265 mp_alltoall_i44 222 10.7 0.688 3.264 0.688 3.264 arnoldi_extremal 4 6.8 0.000 0.000 3.209 3.239 arnoldi_normal_ev 4 7.8 0.001 0.005 3.209 3.239 build_subspace 16 8.4 0.009 0.012 3.112 3.114 ls_scf_post 1 4.0 0.000 0.000 3.074 3.088 ls_scf_store_result 1 5.0 0.000 0.000 2.882 2.916 dbcsr_special_finalize 555 9.7 0.005 0.006 2.388 2.815 dbcsr_merge_single_wm 555 10.7 0.454 0.592 2.379 2.806 make_images_pack 222 9.7 2.215 2.638 2.217 2.640 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.312 2.555 dbcsr_sort_data 658 11.4 2.183 2.534 2.183 2.534 dbcsr_matrix_vector_mult_local 304 10.0 2.072 2.466 2.074 2.468 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.267 2.342 buffer_matrices_ensure_size 222 8.7 1.750 2.065 1.750 2.065 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.770 1.771 rebuild_ks_matrix 3 7.3 0.000 0.000 1.761 1.762 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.760 1.762 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.192000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1134.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b234498075e47956c66564e93abbe409d5e88a7d_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 2.104869E+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 0.059 0.115 91.941 91.942 qs_energies 1 2.0 0.000 0.000 91.283 91.299 ls_scf 1 3.0 0.000 0.001 89.905 89.922 dbcsr_multiply_generic 111 6.7 0.018 0.035 75.712 76.120 multiply_cannon 111 7.7 0.030 0.039 53.247 57.664 ls_scf_main 1 4.0 0.000 0.002 55.198 55.204 multiply_cannon_loop 111 8.7 0.117 0.124 49.976 53.480 density_matrix_trs4 2 5.0 0.002 0.004 49.450 49.638 ls_scf_init_scf 1 4.0 0.000 0.002 31.170 31.171 mp_waitall_1 9105 10.9 21.115 30.191 21.115 30.191 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.965 30.072 matrix_sqrt_Newton_Schulz 2 6.5 0.012 0.050 27.461 27.473 multiply_cannon_multrec 1332 9.7 13.197 16.487 22.559 26.951 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.600 20.847 make_m2s 222 7.7 0.008 0.009 15.586 16.284 make_images 222 8.7 1.583 1.969 15.556 16.256 dbcsr_mm_accdrv_process 4041 10.4 0.282 0.498 8.966 10.573 dbcsr_mm_accdrv_process_sort 4041 11.4 8.554 10.127 8.554 10.127 make_images_data 222 9.7 0.004 0.005 8.954 9.889 hybrid_alltoall_any 227 10.6 0.523 2.482 8.328 9.367 mp_sum_l 807 5.4 5.613 8.655 5.613 8.655 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.224 7.929 mp_irecv_dv 3311 11.0 3.203 7.861 3.203 7.861 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.330 6.920 calculate_norms 2376 9.8 6.029 6.747 6.029 6.747 multiply_cannon_sync_h2d 1332 9.7 4.789 5.923 4.789 5.923 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.142 5.342 arnoldi_extremal 4 6.8 0.000 0.000 4.704 4.735 arnoldi_normal_ev 4 7.8 0.001 0.005 4.704 4.735 build_subspace 16 8.4 0.014 0.021 4.437 4.445 ls_scf_post 1 4.0 0.001 0.003 3.537 3.554 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.189 3.415 ls_scf_store_result 1 5.0 0.000 0.000 3.231 3.353 dbcsr_matrix_vector_mult_local 304 10.0 2.787 3.279 2.790 3.281 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.612 2.730 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 1.197 2.607 mp_allgather_i34 111 8.7 0.962 2.468 0.962 2.468 make_images_pack 222 9.7 2.031 2.443 2.034 2.446 dbcsr_sort_data 436 11.2 1.829 2.123 1.829 2.123 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.879 1.881 rebuild_ks_matrix 3 7.3 0.000 0.000 1.865 1.867 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.005 1.865 1.867 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.942000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1751.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b234498075e47956c66564e93abbe409d5e88a7d_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 2.785645E+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 0.044 0.058 94.385 94.385 qs_energies 1 2.0 0.000 0.000 93.747 93.752 ls_scf 1 3.0 0.000 0.001 92.343 92.349 dbcsr_multiply_generic 111 6.7 0.016 0.017 76.869 77.129 ls_scf_main 1 4.0 0.000 0.002 57.996 57.999 multiply_cannon 111 7.7 0.045 0.093 52.674 56.897 multiply_cannon_loop 111 8.7 0.100 0.106 49.089 52.817 density_matrix_trs4 2 5.0 0.003 0.009 51.944 52.097 mp_waitall_1 7281 11.0 23.623 33.311 23.623 33.311 ls_scf_init_scf 1 4.0 0.000 0.001 30.771 30.775 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.603 29.670 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.134 27.149 multiply_cannon_multrec 888 9.7 12.719 15.377 21.399 24.494 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.825 22.188 make_m2s 222 7.7 0.007 0.008 17.380 18.550 make_images 222 8.7 1.984 2.310 17.341 18.513 hybrid_alltoall_any 227 10.6 0.623 2.887 9.567 11.045 make_images_data 222 9.7 0.004 0.005 9.966 10.964 dbcsr_mm_accdrv_process 3754 10.4 0.240 0.418 8.198 9.487 dbcsr_mm_accdrv_process_sort 3754 11.4 7.827 9.069 7.827 9.069 mp_sum_l 807 5.4 5.253 8.657 5.253 8.657 multiply_cannon_sync_h2d 888 9.7 6.065 7.235 6.065 7.235 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.435 7.080 mp_irecv_dv 2335 11.1 2.420 7.036 2.420 7.036 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.042 6.848 multiply_cannon_metrocomm1 888 9.7 0.002 0.003 3.459 6.344 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.992 5.219 arnoldi_extremal 4 6.8 0.000 0.000 5.136 5.156 arnoldi_normal_ev 4 7.8 0.001 0.005 5.136 5.156 build_subspace 16 8.4 0.014 0.020 4.827 4.834 calculate_norms 1584 9.8 4.277 4.657 4.277 4.657 mp_allgather_i34 111 8.7 1.408 3.910 1.408 3.910 dbcsr_matrix_vector_mult 304 9.0 0.009 0.021 3.492 3.828 dbcsr_matrix_vector_mult_local 304 10.0 3.058 3.630 3.060 3.632 ls_scf_post 1 4.0 0.000 0.001 3.577 3.582 ls_scf_store_result 1 5.0 0.000 0.000 3.305 3.404 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.875 2.998 dbcsr_sort_data 325 11.1 1.908 2.157 1.908 2.157 make_images_pack 222 9.7 1.828 2.121 1.831 2.124 make_images_sizes 222 9.7 0.000 0.000 1.027 2.070 mp_alltoall_i44 222 10.7 1.026 2.069 1.026 2.069 dbcsr_data_release 9322 10.9 1.329 2.001 1.329 2.001 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.906 1.908 dbcsr_finalize 304 7.8 0.027 0.032 1.622 1.891 rebuild_ks_matrix 3 7.3 0.000 0.000 1.887 1.889 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.887 1.889 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=94.385000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2242.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b234498075e47956c66564e93abbe409d5e88a7d_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 3.352158E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 13371. MP_Sync 4 MP_Alltoall 47 30278988. MP_SendRecv 69 86400. MP_ISendRecv 69 86400. MP_Wait 531 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.048 0.068 97.427 97.428 qs_energies 1 2.0 0.000 0.000 96.706 96.713 ls_scf 1 3.0 0.000 0.000 95.062 95.070 dbcsr_multiply_generic 111 6.7 0.017 0.020 78.782 79.013 ls_scf_main 1 4.0 0.000 0.003 59.011 59.014 multiply_cannon 111 7.7 0.048 0.102 51.868 56.769 density_matrix_trs4 2 5.0 0.003 0.007 52.876 52.982 multiply_cannon_loop 111 8.7 0.115 0.126 46.775 50.307 ls_scf_init_scf 1 4.0 0.000 0.001 32.810 32.813 ls_scf_init_matrix_S 1 5.0 0.000 0.001 31.621 31.682 mp_waitall_1 6369 11.0 22.723 29.451 22.723 29.451 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.137 29.151 multiply_cannon_multrec 1332 9.7 14.234 17.279 22.256 25.480 make_m2s 222 7.7 0.008 0.009 21.106 22.586 make_images 222 8.7 3.146 3.618 21.055 22.538 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.256 17.697 make_images_data 222 9.7 0.004 0.005 11.786 13.415 hybrid_alltoall_any 227 10.6 0.801 3.779 11.161 13.067 dbcsr_mm_accdrv_process 3641 10.4 0.210 0.397 7.660 9.197 dbcsr_mm_accdrv_process_sort 3641 11.4 7.275 8.784 7.275 8.784 mp_sum_l 807 5.4 4.235 7.447 4.235 7.447 multiply_cannon_sync_h2d 1332 9.7 5.518 6.095 5.518 6.095 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.087 6.014 mp_irecv_dv 3229 10.9 2.064 5.935 2.064 5.935 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.274 5.812 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.519 5.250 arnoldi_extremal 4 6.8 0.000 0.000 5.231 5.243 arnoldi_normal_ev 4 7.8 0.001 0.005 5.231 5.243 build_subspace 16 8.4 0.014 0.021 4.894 4.904 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.566 4.745 mp_allgather_i34 111 8.7 2.198 4.624 2.198 4.624 calculate_norms 2376 9.8 4.198 4.560 4.198 4.560 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.622 3.919 dbcsr_matrix_vector_mult_local 304 10.0 3.218 3.729 3.220 3.731 dbcsr_sort_data 658 11.4 3.112 3.397 3.112 3.397 ls_scf_post 1 4.0 0.001 0.005 3.241 3.247 dbcsr_special_finalize 555 9.7 0.006 0.007 2.872 3.205 dbcsr_merge_single_wm 555 10.7 0.539 0.667 2.864 3.197 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.015 3.083 ls_scf_store_result 1 5.0 0.000 0.000 2.974 3.040 dbcsr_data_release 10477 10.7 1.567 2.396 1.567 2.396 dbcsr_finalize 304 7.8 0.050 0.062 1.801 1.978 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.428000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2748.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b234498075e47956c66564e93abbe409d5e88a7d_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 4.675502E+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 0.049 0.067 99.460 99.461 qs_energies 1 2.0 0.000 0.000 98.654 98.661 ls_scf 1 3.0 0.000 0.000 96.599 96.610 dbcsr_multiply_generic 111 6.7 0.018 0.018 78.047 78.264 ls_scf_main 1 4.0 0.000 0.000 62.241 62.241 multiply_cannon 111 7.7 0.087 0.176 55.346 60.764 density_matrix_trs4 2 5.0 0.002 0.003 55.127 55.223 multiply_cannon_loop 111 8.7 0.071 0.084 50.795 52.429 mp_waitall_1 5436 11.0 26.602 32.331 26.602 32.331 ls_scf_init_scf 1 4.0 0.000 0.000 30.747 30.752 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.485 29.523 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.331 27.338 multiply_cannon_multrec 444 9.7 14.083 16.432 21.229 24.633 make_m2s 222 7.7 0.006 0.006 17.811 20.333 make_images 222 8.7 3.729 4.417 17.747 20.272 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.098 16.104 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.186 14.410 make_images_data 222 9.7 0.003 0.004 10.120 12.498 hybrid_alltoall_any 227 10.6 0.789 3.778 9.802 12.243 multiply_cannon_sync_h2d 444 9.7 6.482 9.012 6.482 9.012 dbcsr_mm_accdrv_process 3003 10.4 0.175 0.344 6.857 8.005 dbcsr_mm_accdrv_process_sort 3003 11.4 6.543 7.661 6.543 7.661 mp_allgather_i34 111 8.7 2.766 7.003 2.766 7.003 arnoldi_extremal 4 6.8 0.000 0.000 5.885 5.893 arnoldi_normal_ev 4 7.8 0.002 0.005 5.885 5.892 build_subspace 16 8.4 0.015 0.019 5.492 5.502 mp_sum_l 807 5.4 2.929 5.342 2.929 5.342 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.586 4.729 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.195 4.411 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.590 4.193 mp_irecv_dv 1241 11.2 1.572 4.171 1.572 4.171 dbcsr_matrix_vector_mult_local 304 10.0 3.693 4.168 3.695 4.170 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.036 4.048 calculate_norms 792 9.8 3.564 3.698 3.564 3.698 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.537 3.618 ls_scf_post 1 4.0 0.000 0.000 3.612 3.617 make_images_sizes 222 9.7 0.000 0.000 1.024 3.574 mp_alltoall_i44 222 10.7 1.024 3.573 1.024 3.573 ls_scf_store_result 1 5.0 0.000 0.000 3.382 3.439 dbcsr_finalize 304 7.8 0.062 0.078 2.207 2.286 dbcsr_merge_all 275 8.9 0.474 0.529 2.052 2.107 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.078 2.080 rebuild_ks_matrix 3 7.3 0.000 0.000 2.046 2.048 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.046 2.048 qs_energies_init_hamiltonians 1 3.0 0.000 0.001 2.039 2.045 dbcsr_data_release 10123 10.8 1.336 1.992 1.336 1.992 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=99.461000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3680.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b234498075e47956c66564e93abbe409d5e88a7d_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 8.744182E+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 0.077 0.097 111.580 111.580 qs_energies 1 2.0 0.000 0.000 110.166 110.182 ls_scf 1 3.0 0.000 0.000 107.197 107.212 dbcsr_multiply_generic 111 6.7 0.023 0.027 80.796 80.895 ls_scf_main 1 4.0 0.000 0.000 66.128 66.130 density_matrix_trs4 2 5.0 0.002 0.003 56.881 56.958 multiply_cannon 111 7.7 0.140 0.290 50.583 52.877 multiply_cannon_loop 111 8.7 0.068 0.070 47.141 50.543 ls_scf_init_scf 1 4.0 0.000 0.000 37.376 37.377 ls_scf_init_matrix_S 1 5.0 0.000 0.000 35.841 35.852 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 33.004 33.017 mp_waitall_1 4527 11.1 22.360 26.048 22.360 26.048 make_m2s 222 7.7 0.007 0.007 23.894 24.816 make_images 222 8.7 4.599 5.014 23.785 24.705 multiply_cannon_multrec 444 9.7 17.930 18.782 22.609 23.451 hybrid_alltoall_any 227 10.6 1.662 3.620 12.908 15.706 make_images_data 222 9.7 0.004 0.004 13.100 15.433 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.680 11.399 multiply_cannon_sync_h2d 444 9.7 8.823 8.878 8.823 8.878 arnoldi_extremal 4 6.8 0.000 0.000 7.369 7.382 arnoldi_normal_ev 4 7.8 0.021 0.049 7.369 7.382 build_subspace 16 8.4 0.026 0.036 6.778 6.786 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.430 5.574 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.059 5.294 dbcsr_matrix_vector_mult_local 304 10.0 4.970 5.274 4.973 5.276 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.174 5.266 mp_sum_l 807 5.4 3.975 4.929 3.975 4.929 dbcsr_mm_accdrv_process 1814 10.4 0.255 0.319 4.513 4.638 dbcsr_mm_accdrv_process_sort 1814 11.4 4.213 4.345 4.213 4.345 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.378 4.157 compute_matrix_preconditioner 1 6.0 0.002 0.002 3.829 3.841 ls_scf_post 1 4.0 0.000 0.000 3.692 3.707 make_images_sizes 222 9.7 0.000 0.000 1.431 3.565 mp_alltoall_i44 222 10.7 1.430 3.565 1.430 3.565 ls_scf_store_result 1 5.0 0.000 0.000 3.434 3.443 calculate_norms 792 9.8 3.263 3.312 3.263 3.312 mp_allgather_i34 111 8.7 1.042 3.287 1.042 3.287 acc_transpose_blocks 444 9.7 0.002 0.002 0.566 3.269 acc_transpose_blocks_kernels 444 10.7 0.005 0.005 0.530 3.233 jit_kernel_transpose 1 13.0 0.525 3.229 0.525 3.229 dbcsr_finalize 304 7.8 0.082 0.090 3.094 3.188 dbcsr_merge_all 275 8.9 0.892 0.915 2.877 2.967 qs_energies_init_hamiltonians 1 3.0 0.000 0.001 2.939 2.939 dbcsr_complete_redistribute 5 7.6 1.445 1.474 2.779 2.891 dbcsr_data_release 12724 10.6 2.341 2.858 2.341 2.858 matrix_ls_to_qs 2 6.0 0.000 0.000 2.443 2.584 dbcsr_sort_data 325 11.1 2.446 2.522 2.446 2.522 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.467 2.469 rebuild_ks_matrix 3 7.3 0.000 0.000 2.399 2.401 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.399 2.401 dbcsr_new_transposed 4 7.5 0.243 0.253 2.312 2.334 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=111.580000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6840.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b234498075e47956c66564e93abbe409d5e88a7d_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 586.387456E+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 1188511151. 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 0.024 0.041 223.952 223.954 qs_mol_dyn_low 1 2.0 0.004 0.032 223.124 223.146 qs_forces 5 3.8 0.005 0.024 222.987 222.988 qs_energies 5 4.8 0.003 0.027 219.983 220.026 scf_env_do_scf 5 5.8 0.000 0.001 205.814 205.817 scf_env_do_scf_inner_loop 105 6.6 0.003 0.006 179.465 179.466 qs_scf_new_mos 105 7.6 0.001 0.001 138.949 139.106 qs_scf_loop_do_ot 105 8.6 0.001 0.001 138.949 139.106 dbcsr_multiply_generic 1445 12.2 0.122 0.129 131.251 131.591 ot_scf_mini 105 9.6 0.003 0.004 129.377 129.509 multiply_cannon 1445 13.2 0.275 0.293 112.586 115.113 multiply_cannon_loop 1445 14.2 2.390 2.504 110.851 111.959 velocity_verlet 4 3.0 0.001 0.002 106.272 106.273 ot_mini 105 10.6 0.001 0.001 59.911 60.060 multiply_cannon_multrec 69360 15.2 31.901 37.096 39.466 44.481 mp_waitall_1 488190 16.1 33.741 41.112 33.741 41.112 qs_ot_get_p 112 10.4 0.001 0.001 40.016 40.336 qs_ot_get_derivative 55 11.6 0.001 0.001 38.267 38.390 multiply_cannon_sync_h2d 69360 15.2 30.810 35.191 30.810 35.191 multiply_cannon_metrocomm3 69360 15.2 0.191 0.201 24.721 32.562 qs_ot_p2m_diag 40 11.0 0.020 0.031 29.383 29.450 rebuild_ks_matrix 110 8.4 0.000 0.000 29.178 29.351 qs_ks_build_kohn_sham_matrix 110 9.4 0.012 0.013 29.178 29.351 qs_ks_update_qs_env 112 7.6 0.001 0.001 26.807 26.966 cp_dbcsr_syevd 40 12.0 0.002 0.002 26.452 26.453 init_scf_loop 7 6.6 0.000 0.001 26.319 26.320 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 22.960 23.228 apply_single 62 13.6 0.000 0.000 22.960 23.228 prepare_preconditioner 7 7.6 0.000 0.000 21.437 21.476 make_preconditioner 7 8.6 0.000 0.000 21.437 21.476 cp_fm_syevd 40 13.0 0.000 0.000 21.274 21.413 ot_new_cg_direction 55 11.6 0.001 0.001 20.951 20.952 qs_rho_update_rho_low 110 7.6 0.001 0.001 17.184 17.454 calculate_rho_elec 110 8.6 0.030 0.034 17.183 17.453 cp_fm_redistribute_end 40 14.0 8.323 16.605 8.328 16.606 cp_fm_syevd_base 40 14.0 8.269 16.555 8.269 16.555 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.321 14.500 make_full_inverse_cholesky 7 9.6 0.000 0.000 14.193 14.257 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 14.106 14.224 rs_pw_transfer 690 11.5 0.011 0.013 11.556 12.775 mp_sum_l 4684 12.4 11.870 12.708 11.870 12.708 density_rs2pw 110 9.6 0.006 0.008 11.330 12.551 pw_transfer 1645 12.4 0.084 0.111 11.962 12.222 fft_wrap_pw1pw2 1425 13.5 0.013 0.017 11.819 12.084 init_scf_run 5 5.8 0.000 0.001 11.451 11.452 scf_env_initial_rho_setup 5 6.8 0.000 0.000 11.451 11.452 calculate_dm_sparse 110 9.5 0.000 0.001 11.200 11.350 qs_vxc_create 110 10.4 0.002 0.004 10.398 10.450 fft_wrap_pw1pw2_240 915 15.0 1.197 1.318 10.055 10.292 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 10.074 10.157 cp_fm_cholesky_invert 7 10.6 9.980 9.987 9.980 9.987 check_diag 80 13.5 8.585 8.856 9.414 9.554 sum_up_and_integrate 60 10.3 0.028 0.031 8.810 8.820 integrate_v_rspace 60 11.3 0.002 0.022 8.782 8.792 fft3d_pb 915 16.0 2.383 2.675 8.114 8.418 dbcsr_mm_accdrv_process 154766 15.8 3.968 4.146 7.432 8.246 xc_rho_set_and_dset_create 110 12.4 0.076 0.097 7.275 7.529 multiply_cannon_metrocomm1 69360 15.2 0.089 0.096 4.443 7.351 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.002 7.050 7.065 make_full_single_inverse 7 9.6 0.001 0.001 6.973 7.011 xc_vxc_pw_create 60 11.3 0.039 0.049 6.947 6.999 make_m2s 2890 13.2 0.075 0.085 6.305 6.867 acc_transpose_blocks 69360 15.2 0.361 0.377 6.222 6.861 make_images 2890 14.2 0.244 0.260 6.206 6.771 calculate_first_density_matrix 1 7.0 0.000 0.001 6.690 6.717 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 6.644 6.702 xc_pw_derive 510 13.4 0.005 0.007 6.145 6.230 mp_waitany 7680 13.5 4.490 5.899 4.490 5.899 mp_alltoall_z22v 2340 17.7 5.183 5.456 5.183 5.456 potential_pw2rs 60 12.3 0.003 0.003 4.958 4.985 multiply_cannon_metrocomm4 67915 15.2 0.181 0.197 2.063 4.913 acc_transpose_blocks_kernels 69360 16.2 0.851 0.897 4.330 4.871 grid_collocate_task_list 110 9.6 4.091 4.557 4.091 4.557 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=223.954000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=559.000000, yerr=2.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: b234498075e47956c66564e93abbe409d5e88a7d Summary: empty Status: OK