=== 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: eb0d6eed04221145363480ad2019c815711cc597 ################# 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 2022.05.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.0.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.0, # SIRIUS 7.3.2, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Author: Matthias Krack (26.10.2022) # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 --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=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2022.05.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.0.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.0 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.3.2 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta_prefixed_scalapack.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/eb0d6eed04221145363480ad2019c815711cc597_performance_tests/01 job id: 43464447 --- 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/eb0d6eed04221145363480ad2019c815711cc597_performance_tests/02 job id: 43464449 --- 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/eb0d6eed04221145363480ad2019c815711cc597_performance_tests/03 job id: 43464450 --- 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/eb0d6eed04221145363480ad2019c815711cc597_performance_tests/04 job id: 43464451 --- 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/eb0d6eed04221145363480ad2019c815711cc597_performance_tests/05 job id: 43464452 --- 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/eb0d6eed04221145363480ad2019c815711cc597_performance_tests/06 job id: 43464453 --- 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/eb0d6eed04221145363480ad2019c815711cc597_performance_tests/07 job id: 43464454 --- 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/eb0d6eed04221145363480ad2019c815711cc597_performance_tests/08 job id: 43464455 --- 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/eb0d6eed04221145363480ad2019c815711cc597_performance_tests/09 job id: 43464456 --- 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/eb0d6eed04221145363480ad2019c815711cc597_performance_tests/10 job id: 43464457 --- 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/eb0d6eed04221145363480ad2019c815711cc597_performance_tests/11 job id: 43464458 --- 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/eb0d6eed04221145363480ad2019c815711cc597_performance_tests/12 job id: 43464459 --- 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/eb0d6eed04221145363480ad2019c815711cc597_performance_tests/13 job id: 43464460 --- 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/eb0d6eed04221145363480ad2019c815711cc597_performance_tests/14 job id: 43464462 --- 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/eb0d6eed04221145363480ad2019c815711cc597_performance_tests/15 job id: 43464464 --- 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/eb0d6eed04221145363480ad2019c815711cc597_performance_tests/16 job id: 43464465 --- 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/eb0d6eed04221145363480ad2019c815711cc597_performance_tests/17 job id: 43464466 --- 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/eb0d6eed04221145363480ad2019c815711cc597_performance_tests/18 job id: 43464468 --- 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/eb0d6eed04221145363480ad2019c815711cc597_performance_tests/19 job id: 43464469 --- 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/eb0d6eed04221145363480ad2019c815711cc597_performance_tests/20 job id: 43464470 --- 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/eb0d6eed04221145363480ad2019c815711cc597_performance_tests/21 job id: 43464471 --- 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/eb0d6eed04221145363480ad2019c815711cc597_performance_tests/22 job id: 43464472 --- 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/eb0d6eed04221145363480ad2019c815711cc597_performance_tests/23 job id: 43464473 --- 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/eb0d6eed04221145363480ad2019c815711cc597_performance_tests/24 job id: 43464474 --- 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/eb0d6eed04221145363480ad2019c815711cc597_performance_tests/25 job id: 43464475 --- 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/eb0d6eed04221145363480ad2019c815711cc597_performance_tests/26 job id: 43464476 --- Point --- name: 510 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/1r/12t) --- Point --- name: 511 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ === END TESTS (description) === ===== PLOTS (description) ===== ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2_mem", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md_mem", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md_mem", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md_mem", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls_mem", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" === END PLOTS (description) === ============ RESULTS ============ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/eb0d6eed04221145363480ad2019c815711cc597_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 344 9. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.034 133.425 133.426 farming_run 1 2.0 132.922 132.923 133.399 133.402 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.448866E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 228 1113141. MP_Allreduce 485 2282278. MP_Sync 27 MP_Alltoall 38 9316958. MP_SendRecv 30 829726. MP_ISendRecv 135 235435. MP_Wait 281 MP_comm_split 8 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.029 115.593 115.593 qs_energies 1 2.0 0.000 0.000 115.365 115.368 mp2_main 1 3.0 0.000 0.000 113.353 113.356 mp2_gpw_main 1 4.0 0.019 0.026 112.435 112.438 mp2_ri_gpw_compute_in 1 5.0 0.172 0.182 93.663 93.797 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 55.349 55.482 mp2_eri_3c_integrate_gpw 272 7.0 0.154 0.171 41.654 46.639 get_2c_integrals 1 6.0 0.001 0.016 37.618 38.132 integrate_v_rspace 273 8.0 0.433 0.448 25.049 29.664 pw_transfer 6555 10.6 0.373 0.387 27.475 27.957 fft_wrap_pw1pw2 5465 11.4 0.044 0.046 26.092 26.401 grid_integrate_task_list 273 9.0 20.872 25.918 20.872 25.918 fft_wrap_pw1pw2_100 2178 12.4 1.262 1.465 23.653 23.977 compute_2c_integrals 1 7.0 0.002 0.003 19.799 19.799 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 19.075 19.525 mp2_eri_2c_integrate_gpw 1 9.0 2.355 2.435 19.072 19.523 rpa_ri_compute_en 1 5.0 0.011 0.013 18.663 18.734 cp_fm_cholesky_decompose 12 8.2 17.754 18.276 17.754 18.276 cholesky_decomp 1 7.0 0.000 0.000 16.667 17.193 fft3d_s 5443 13.4 16.175 16.430 16.198 16.450 ao_to_mo_and_store_B_mult_1 272 7.0 10.852 15.552 10.852 15.552 calculate_wavefunction 272 8.0 5.415 5.498 12.529 13.169 rpa_num_int 1 6.0 0.001 0.023 10.517 10.518 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.467 10.473 calc_mat_Q 8 8.0 0.000 0.000 9.334 9.426 contract_S_to_Q 8 9.0 0.000 0.000 8.755 8.845 calc_potential_gpw 544 9.5 0.005 0.006 8.306 8.724 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.333 8.608 parallel_gemm_fm 14 9.1 0.000 0.000 8.347 8.435 parallel_gemm_fm_cosma 14 10.1 8.347 8.435 8.347 8.435 potential_pw2rs 545 10.0 0.107 0.109 7.745 8.359 create_integ_mat 1 6.0 0.021 0.028 7.904 7.905 collocate_single_gaussian 272 10.0 0.041 0.044 7.563 7.817 array2fm 1 7.0 0.000 0.000 6.763 7.222 pw_scatter_s 2720 13.7 4.419 4.619 4.419 4.619 pw_gather_s 2722 13.2 3.861 4.232 3.861 4.232 array2fm_buffer_send 1 8.0 2.916 3.081 2.916 3.081 pw_poisson_solve 545 10.5 1.135 1.180 2.225 2.447 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.437878, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2720.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/eb0d6eed04221145363480ad2019c815711cc597_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 0.030 0.040 396.487 396.488 farming_run 1 2.0 395.712 395.716 396.438 396.442 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.224598E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 703 408373. MP_Allreduce 1821 23730. MP_Sync 38 MP_Alltoall 77 17246853. MP_SendRecv 2171 2843495. MP_ISendRecv 1739 144022. MP_Wait 2051 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.049 209.496 209.497 qs_energies 1 2.0 0.000 0.000 209.233 209.250 scf_env_do_scf 1 3.0 0.000 0.000 106.137 106.138 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.183 105.190 rebuild_ks_matrix 4 6.0 0.000 0.000 105.182 105.188 qs_ks_build_kohn_sham_matrix 4 7.0 0.061 0.068 105.182 105.188 hfx_ks_matrix 4 8.0 0.001 0.001 104.794 104.797 integrate_four_center 4 9.0 0.144 0.461 104.793 104.796 mp2_main 1 3.0 0.000 0.000 102.778 102.794 mp2_gpw_main 1 4.0 0.031 0.043 101.788 101.803 integrate_four_center_main 4 10.0 0.134 0.581 96.767 99.128 integrate_four_center_bin 264 11.0 96.633 99.094 96.633 99.094 init_scf_loop 1 4.0 0.000 0.000 91.981 91.981 mp2_ri_gpw_compute_in 1 5.0 0.066 0.086 74.915 75.919 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.003 54.512 55.515 mp2_eri_3c_integrate_gpw 91 7.0 0.145 0.162 42.273 47.267 integrate_v_rspace 95 8.0 0.398 0.573 28.625 33.457 pw_transfer 2240 10.6 0.147 0.173 29.900 30.438 fft_wrap_pw1pw2 1868 11.4 0.018 0.021 28.919 29.457 ao_to_mo_and_store_B_mult_1 91 7.0 10.559 29.162 10.559 29.162 grid_integrate_task_list 95 9.0 23.900 28.967 23.900 28.967 mp2_ri_gpw_compute_en 1 5.0 0.055 0.069 26.728 28.355 fft_wrap_pw1pw2_100 730 12.4 1.269 1.481 26.637 27.092 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.832 1.906 25.058 25.069 get_2c_integrals 1 6.0 0.000 0.000 20.271 20.337 compute_2c_integrals 1 7.0 0.003 0.004 19.248 19.258 compute_2c_integrals_loop_lm 1 8.0 0.002 0.008 18.820 19.122 mp2_eri_2c_integrate_gpw 1 9.0 1.734 1.850 18.818 19.121 fft3d_s 1823 13.4 18.430 18.812 18.444 18.825 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.155 14.155 mp2_ri_gpw_compute_en_expansio 172 7.0 0.558 0.588 8.827 10.135 calculate_wavefunction 91 8.0 2.030 2.066 9.766 10.001 local_gemm 172 8.0 8.269 9.549 8.269 9.549 potential_pw2rs 186 10.0 0.033 0.035 8.642 9.231 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.215 8.574 mp2_ri_gpw_compute_en_comm 22 7.0 0.518 0.544 8.012 8.396 collocate_single_gaussian 91 10.0 0.018 0.021 7.862 8.078 calc_potential_gpw 182 9.5 0.002 0.002 7.898 8.066 mp2_ri_gpw_compute_en_ener 172 7.0 6.342 6.431 6.342 6.431 mp_sendrecv_dm3 2068 8.0 5.998 6.396 5.998 6.396 mp_sync 38 10.4 2.709 5.574 2.709 5.574 pw_gather_s 912 13.2 4.923 5.446 4.923 5.446 pw_scatter_s 910 13.7 3.925 4.229 3.925 4.229 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.782434, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1512.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/eb0d6eed04221145363480ad2019c815711cc597_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 451.993600E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62385. MP_Allreduce 10249 271. MP_Sync 580 MP_Alltoall 2083 589496. MP_ISendRecv 45220 5520. MP_Wait 60486 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 0.013 0.032 52.142 52.143 qs_mol_dyn_low 1 2.0 0.003 0.003 51.948 51.955 qs_forces 11 3.9 0.002 0.002 51.884 51.885 qs_energies 11 4.9 0.001 0.001 50.425 50.436 scf_env_do_scf 11 5.9 0.000 0.001 44.532 44.533 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 42.357 42.357 dbcsr_multiply_generic 2286 12.5 0.093 0.098 33.373 33.863 qs_scf_new_mos 108 7.5 0.000 0.001 32.015 32.344 qs_scf_loop_do_ot 108 8.5 0.000 0.001 32.014 32.344 ot_scf_mini 108 9.5 0.002 0.002 30.390 30.588 multiply_cannon 2286 13.5 0.184 0.191 25.768 27.552 multiply_cannon_loop 2286 14.5 1.512 1.584 25.011 26.758 velocity_verlet 10 3.0 0.001 0.001 25.534 25.534 ot_mini 108 10.5 0.001 0.001 19.339 19.632 qs_ot_get_derivative 108 11.5 0.001 0.001 16.351 16.565 mp_waitall_1 267858 16.1 8.712 15.036 8.712 15.036 multiply_cannon_metrocomm3 54864 15.5 0.067 0.073 5.846 13.055 multiply_cannon_multrec 54864 15.5 4.210 6.440 7.509 10.884 rebuild_ks_matrix 119 8.3 0.000 0.000 7.979 8.144 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 7.978 8.144 multiply_cannon_sync_h2d 54864 15.5 5.884 7.804 5.884 7.804 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.052 7.207 mp_sum_l 7207 12.9 5.449 7.167 5.449 7.167 qs_ot_get_p 119 10.4 0.001 0.001 6.414 6.724 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.842 5.963 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.316 5.769 init_scf_run 11 5.9 0.000 0.001 4.657 4.657 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.657 4.657 sum_up_and_integrate 119 10.3 0.012 0.015 4.600 4.607 integrate_v_rspace 119 11.3 0.002 0.002 4.587 4.596 dbcsr_mm_accdrv_process 76910 16.1 1.186 1.820 3.221 4.483 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.297 4.368 calculate_rho_elec 119 8.7 0.011 0.016 4.296 4.367 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.554 3.621 multiply_cannon_metrocomm1 54864 15.5 0.052 0.056 1.807 3.399 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.885 3.074 apply_single 119 13.6 0.000 0.000 2.885 3.074 rs_pw_transfer 974 11.9 0.012 0.013 2.953 3.060 calculate_dm_sparse 119 9.5 0.000 0.001 2.846 2.998 ot_diis_step 108 11.5 0.006 0.006 2.676 2.677 cp_dbcsr_syevd 50 12.0 0.002 0.003 2.662 2.663 jit_kernel_multiply 13 15.8 1.971 2.571 1.971 2.571 density_rs2pw 119 9.7 0.004 0.004 2.417 2.531 calculate_first_density_matrix 1 7.0 0.000 0.000 2.514 2.519 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.434 2.502 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.437 2.438 cp_fm_redistribute_end 50 14.0 2.213 2.416 2.217 2.417 cp_fm_diag_elpa_base 50 14.0 0.198 2.379 0.199 2.389 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.241 2.243 init_scf_loop 11 6.9 0.000 0.000 2.154 2.154 acc_transpose_blocks 54864 15.5 0.234 0.248 1.747 2.152 grid_integrate_task_list 119 12.3 2.024 2.122 2.024 2.122 wfi_extrapolate 11 7.9 0.001 0.001 2.085 2.085 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.971 2.023 mp_sum_d 4125 12.0 1.337 1.974 1.337 1.974 potential_pw2rs 119 12.3 0.004 0.004 1.957 1.970 make_m2s 4572 13.5 0.053 0.055 1.847 1.902 pw_transfer 1439 11.6 0.052 0.056 1.818 1.888 make_images 4572 14.5 0.132 0.138 1.766 1.820 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.743 1.814 fft3d_ps 1201 14.6 0.354 0.456 1.520 1.582 mp_alltoall_d11v 2130 13.8 1.357 1.517 1.357 1.517 fft_wrap_pw1pw2_140 487 13.2 0.080 0.094 1.350 1.424 mp_waitany 12084 13.8 1.277 1.423 1.277 1.423 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.402 1.416 grid_collocate_task_list 119 9.7 1.289 1.356 1.289 1.356 dbcsr_dot_sd 1205 11.9 0.047 0.057 0.745 1.130 prepare_preconditioner 11 7.9 0.000 0.000 1.037 1.065 make_preconditioner 11 8.9 0.000 0.000 1.037 1.065 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=52.143000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=431.000000, yerr=1.348400 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/eb0d6eed04221145363480ad2019c815711cc597_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 487.620608E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2194560 MPI messages size (bytes): total size 310.646604E+09 min size 0.000000E+00 max size 1.145520E+06 average size 141.553031E+03 MPI breakdown and total messages size (bytes): size <= 128 724648 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 281952 4619501568 32768 < size <= 131072 494448 39143342080 131072 < size <= 4194304 440000 264807943488 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62664. MP_Allreduce 10226 305. MP_Sync 104 MP_Alltoall 2060 1567453. MP_ISendRecv 33558 37093. MP_Wait 40318 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 0.035 0.074 40.489 40.490 qs_mol_dyn_low 1 2.0 0.003 0.003 40.146 40.155 qs_forces 11 3.9 0.002 0.003 40.088 40.089 qs_energies 11 4.9 0.006 0.023 38.209 38.214 scf_env_do_scf 11 5.9 0.001 0.002 32.815 32.816 scf_env_do_scf_inner_loop 108 6.5 0.002 0.008 30.128 30.129 dbcsr_multiply_generic 2286 12.5 0.099 0.103 22.054 22.453 qs_scf_new_mos 108 7.5 0.001 0.001 20.720 20.967 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.720 20.966 ot_scf_mini 108 9.5 0.003 0.003 19.798 19.979 velocity_verlet 10 3.0 0.001 0.001 18.776 18.777 multiply_cannon 2286 13.5 0.210 0.219 16.656 18.477 multiply_cannon_loop 2286 14.5 0.899 0.971 15.427 17.111 ot_mini 108 10.5 0.001 0.001 12.308 12.557 mp_waitall_1 217478 16.2 6.558 11.963 6.558 11.963 qs_ot_get_derivative 108 11.5 0.001 0.001 9.776 9.953 multiply_cannon_metrocomm3 27432 15.5 0.067 0.071 4.341 9.679 multiply_cannon_multrec 27432 15.5 1.960 4.304 5.970 8.699 rebuild_ks_matrix 119 8.3 0.000 0.000 7.559 7.707 qs_ks_build_kohn_sham_matrix 119 9.3 0.014 0.022 7.558 7.707 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.565 6.703 dbcsr_mm_accdrv_process 47894 16.0 3.031 5.385 3.941 5.838 qs_ot_get_p 119 10.4 0.001 0.001 4.574 4.819 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.704 4.548 sum_up_and_integrate 119 10.3 0.030 0.036 4.522 4.541 integrate_v_rspace 119 11.3 0.002 0.002 4.492 4.518 mp_sum_l 7207 12.9 2.254 4.180 2.254 4.180 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.097 4.159 apply_single 119 13.6 0.000 0.000 3.096 4.158 init_scf_run 11 5.9 0.000 0.002 4.011 4.013 scf_env_initial_rho_setup 11 6.9 0.001 0.002 4.011 4.012 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.819 3.851 calculate_rho_elec 119 8.7 0.021 0.024 3.818 3.850 make_m2s 4572 13.5 0.052 0.054 2.717 2.984 qs_ot_p2m_diag 50 11.0 0.009 0.013 2.954 2.975 rs_pw_transfer 974 11.9 0.010 0.010 2.887 2.968 make_images 4572 14.5 0.200 0.237 2.629 2.894 multiply_cannon_sync_h2d 27432 15.5 2.167 2.798 2.167 2.798 init_scf_loop 11 6.9 0.001 0.003 2.662 2.664 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.490 2.491 calculate_first_density_matrix 1 7.0 0.001 0.003 2.486 2.489 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.390 2.481 ot_diis_step 108 11.5 0.010 0.011 2.468 2.469 potential_pw2rs 119 12.3 0.006 0.006 2.246 2.276 density_rs2pw 119 9.7 0.004 0.004 2.169 2.266 calculate_dm_sparse 119 9.5 0.000 0.000 2.110 2.187 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.092 2.093 cp_fm_redistribute_end 50 14.0 1.720 2.063 1.726 2.066 pw_transfer 1439 11.6 0.064 0.067 2.014 2.063 jit_kernel_multiply 10 16.0 0.858 2.032 0.858 2.032 cp_fm_diag_elpa_base 50 14.0 0.329 1.993 0.337 2.026 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.001 2.003 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.924 1.974 grid_integrate_task_list 119 12.3 1.831 1.934 1.831 1.934 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.748 1.788 prepare_preconditioner 11 7.9 0.000 0.000 1.721 1.747 make_preconditioner 11 8.9 0.000 0.002 1.721 1.747 make_images_data 4572 15.5 0.045 0.051 1.225 1.683 fft3d_ps 1201 14.6 0.497 0.549 1.632 1.675 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.606 1.660 hybrid_alltoall_any 4725 16.4 0.051 0.111 1.065 1.570 fft_wrap_pw1pw2_140 487 13.2 0.075 0.084 1.466 1.514 acc_transpose_blocks 27432 15.5 0.109 0.115 1.206 1.507 wfi_extrapolate 11 7.9 0.001 0.001 1.469 1.469 mp_allgather_i34 2286 14.5 0.671 1.460 0.671 1.460 mp_alltoall_d11v 2130 13.8 1.269 1.450 1.269 1.450 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.299 1.309 grid_collocate_task_list 119 9.7 1.240 1.305 1.240 1.305 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.190 1.239 mp_sum_d 4125 12.0 0.667 1.140 0.667 1.140 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.051 1.065 mp_alltoall_z22v 1201 16.6 0.886 0.986 0.886 0.986 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 0.961 0.963 make_images_sizes 4572 15.5 0.005 0.005 0.670 0.943 mp_alltoall_i44 4572 16.5 0.665 0.938 0.665 0.938 acc_transpose_blocks_kernels 27432 16.5 0.182 0.270 0.674 0.892 rs_pw_transfer_PW2RS_50 119 14.3 0.589 0.607 0.838 0.878 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=40.490000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=463.727273, yerr=1.542778 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/eb0d6eed04221145363480ad2019c815711cc597_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 521.523200E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62660. MP_Allreduce 10225 303. MP_Sync 104 MP_Alltoall 1821 1607811. MP_ISendRecv 22134 57667. MP_Wait 33054 MP_comm_split 50 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.028 33.798 33.798 qs_mol_dyn_low 1 2.0 0.003 0.003 33.501 33.509 qs_forces 11 3.9 0.002 0.003 33.431 33.434 qs_energies 11 4.9 0.001 0.001 31.829 31.834 scf_env_do_scf 11 5.9 0.000 0.001 26.703 26.703 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 23.832 23.833 dbcsr_multiply_generic 2286 12.5 0.094 0.097 17.227 17.337 velocity_verlet 10 3.0 0.001 0.001 16.139 16.141 qs_scf_new_mos 108 7.5 0.001 0.001 15.431 15.451 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.431 15.451 ot_scf_mini 108 9.5 0.002 0.002 14.700 14.717 multiply_cannon 2286 13.5 0.198 0.207 13.574 14.270 multiply_cannon_loop 2286 14.5 0.634 0.665 12.721 13.448 ot_mini 108 10.5 0.001 0.001 9.096 9.106 qs_ot_get_derivative 108 11.5 0.001 0.001 7.569 7.586 multiply_cannon_multrec 18288 15.5 1.957 2.921 7.164 7.431 rebuild_ks_matrix 119 8.3 0.000 0.000 6.626 6.656 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.625 6.655 dbcsr_mm_accdrv_process 38222 16.0 4.229 5.662 5.124 5.952 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.869 5.894 sum_up_and_integrate 119 10.3 0.031 0.035 4.224 4.230 mp_waitall_1 169478 16.3 3.330 4.214 3.330 4.214 integrate_v_rspace 119 11.3 0.002 0.003 4.192 4.198 init_scf_run 11 5.9 0.000 0.001 3.925 3.925 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.924 3.924 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.803 3.440 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.435 3.439 calculate_rho_elec 119 8.7 0.030 0.031 3.434 3.439 qs_ot_get_p 119 10.4 0.001 0.001 3.359 3.384 init_scf_loop 11 6.9 0.000 0.000 2.851 2.852 rs_pw_transfer 974 11.9 0.009 0.010 2.599 2.723 calculate_first_density_matrix 1 7.0 0.000 0.000 2.632 2.633 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.057 2.373 apply_single 119 13.6 0.000 0.000 2.057 2.372 multiply_cannon_metrocomm3 18288 15.5 0.044 0.045 1.507 2.292 density_rs2pw 119 9.7 0.004 0.004 2.031 2.153 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.145 2.151 make_m2s 4572 13.5 0.044 0.045 1.986 2.146 make_images 4572 14.5 0.191 0.202 1.903 2.060 prepare_preconditioner 11 7.9 0.000 0.000 2.023 2.026 make_preconditioner 11 8.9 0.000 0.000 2.023 2.026 potential_pw2rs 119 12.3 0.007 0.008 1.990 2.004 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.982 1.984 calculate_dm_sparse 119 9.5 0.000 0.000 1.955 1.967 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.864 1.951 grid_integrate_task_list 119 12.3 1.795 1.892 1.795 1.892 pw_transfer 1439 11.6 0.064 0.067 1.860 1.868 jit_kernel_multiply 8 16.0 0.843 1.863 0.843 1.863 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.859 1.860 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.781 1.806 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.768 1.777 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.706 1.714 multiply_cannon_sync_h2d 18288 15.5 1.388 1.627 1.388 1.627 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.604 1.606 cp_fm_redistribute_end 50 14.0 1.192 1.584 1.193 1.586 mp_sum_l 7207 12.9 1.283 1.579 1.283 1.579 cp_fm_diag_elpa_base 50 14.0 0.374 1.519 0.390 1.559 ot_diis_step 108 11.5 0.011 0.011 1.497 1.497 fft3d_ps 1201 14.6 0.505 0.521 1.454 1.464 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.325 1.330 acc_transpose_blocks 18288 15.5 0.075 0.077 1.291 1.328 fft_wrap_pw1pw2_140 487 13.2 0.086 0.089 1.297 1.305 grid_collocate_task_list 119 9.7 1.205 1.275 1.205 1.275 wfi_extrapolate 11 7.9 0.001 0.001 1.246 1.246 make_images_data 4572 15.5 0.045 0.049 0.888 1.067 hybrid_alltoall_any 4725 16.4 0.055 0.111 0.761 0.965 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.933 0.958 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 0.944 0.947 cp_fm_cholesky_invert 11 10.9 0.894 0.900 0.894 0.900 acc_transpose_blocks_kernels 18288 16.5 0.209 0.221 0.852 0.874 mp_alltoall_d11v 2130 13.8 0.749 0.860 0.749 0.860 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.827 0.833 mp_alltoall_z22v 1201 16.6 0.774 0.831 0.774 0.831 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.649 0.747 mp_waitany 9880 13.7 0.566 0.693 0.566 0.693 mp_allgather_i34 2286 14.5 0.313 0.679 0.313 0.679 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=33.798000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.636364, yerr=1.493789 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/eb0d6eed04221145363480ad2019c815711cc597_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 545.873920E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62659. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 2412273. MP_ISendRecv 16422 74133. MP_Wait 24482 MP_comm_split 50 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.038 0.083 36.392 36.393 qs_mol_dyn_low 1 2.0 0.003 0.004 36.076 36.083 qs_forces 11 3.9 0.002 0.002 35.540 35.540 qs_energies 11 4.9 0.002 0.007 33.801 33.808 scf_env_do_scf 11 5.9 0.001 0.002 28.651 28.653 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 24.970 24.973 velocity_verlet 10 3.0 0.001 0.001 18.912 18.934 dbcsr_multiply_generic 2286 12.5 0.100 0.104 18.554 18.668 qs_scf_new_mos 108 7.5 0.001 0.001 16.586 16.641 qs_scf_loop_do_ot 108 8.5 0.001 0.002 16.586 16.641 ot_scf_mini 108 9.5 0.003 0.003 15.624 15.675 multiply_cannon 2286 13.5 0.230 0.270 14.588 15.048 multiply_cannon_loop 2286 14.5 0.939 0.970 13.547 13.942 ot_mini 108 10.5 0.001 0.001 9.604 9.664 multiply_cannon_multrec 27432 15.5 2.337 3.081 8.584 8.999 qs_ot_get_derivative 108 11.5 0.001 0.001 7.772 7.825 dbcsr_mm_accdrv_process 47916 15.9 5.196 6.914 6.152 7.334 rebuild_ks_matrix 119 8.3 0.000 0.000 6.611 6.664 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.611 6.664 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.855 5.903 sum_up_and_integrate 119 10.3 0.035 0.037 3.860 3.871 integrate_v_rspace 119 11.3 0.002 0.003 3.825 3.836 init_scf_run 11 5.9 0.000 0.002 3.746 3.747 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.746 3.747 init_scf_loop 11 6.9 0.001 0.002 3.659 3.663 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.464 3.500 calculate_rho_elec 119 8.7 0.040 0.046 3.463 3.499 qs_ot_get_p 119 10.4 0.001 0.002 3.321 3.384 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.858 3.310 mp_waitall_1 145218 16.4 2.500 3.001 2.500 3.001 prepare_preconditioner 11 7.9 0.000 0.000 2.764 2.773 make_preconditioner 11 8.9 0.000 0.001 2.764 2.773 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.349 2.690 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.104 2.545 apply_single 119 13.6 0.000 0.000 2.104 2.545 make_m2s 4572 13.5 0.054 0.056 2.406 2.505 make_images 4572 14.5 0.273 0.332 2.298 2.395 rs_pw_transfer 974 11.9 0.009 0.009 2.240 2.347 calculate_first_density_matrix 1 7.0 0.001 0.004 2.291 2.293 calculate_dm_sparse 119 9.5 0.000 0.000 2.083 2.142 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.000 2.027 density_rs2pw 119 9.7 0.003 0.004 1.932 2.027 qs_ot_p2m_diag 50 11.0 0.015 0.023 1.970 1.981 grid_integrate_task_list 119 12.3 1.818 1.897 1.818 1.897 pw_transfer 1439 11.6 0.064 0.067 1.816 1.852 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.841 1.843 jit_kernel_multiply 10 15.9 0.896 1.823 0.896 1.823 ot_diis_step 108 11.5 0.012 0.012 1.788 1.788 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.725 1.764 mp_sum_l 7207 12.9 1.153 1.754 1.153 1.754 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.651 1.663 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.645 1.645 potential_pw2rs 119 12.3 0.008 0.009 1.632 1.639 acc_transpose_blocks 27432 15.5 0.111 0.114 1.470 1.507 fft3d_ps 1201 14.6 0.533 0.585 1.406 1.433 fft_wrap_pw1pw2_140 487 13.2 0.084 0.092 1.379 1.419 wfi_extrapolate 11 7.9 0.001 0.001 1.405 1.405 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.395 1.396 cp_fm_redistribute_end 50 14.0 0.917 1.369 0.919 1.370 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.800 1.359 cp_fm_diag_elpa_base 50 14.0 0.429 1.307 0.449 1.349 grid_collocate_task_list 119 9.7 1.220 1.334 1.220 1.334 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.002 1.242 1.251 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.165 1.181 cp_fm_upper_to_full 72 13.5 0.822 1.146 0.822 1.146 dbcsr_complete_redistribute 329 12.2 0.118 0.149 0.849 1.134 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.121 1.122 make_images_data 4572 15.5 0.045 0.048 0.969 1.104 multiply_cannon_sync_h2d 27432 15.5 0.984 1.046 0.984 1.046 hybrid_alltoall_any 4725 16.4 0.061 0.150 0.830 1.015 mp_alltoall_d11v 2130 13.8 0.860 1.003 0.860 1.003 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.615 0.893 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.792 0.869 acc_transpose_blocks_kernels 27432 16.5 0.268 0.276 0.846 0.868 cp_fm_cholesky_invert 11 10.9 0.848 0.852 0.848 0.852 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.847 0.852 mp_alltoall_z22v 1201 16.6 0.756 0.786 0.756 0.786 mp_alltoall_i22 627 13.8 0.456 0.769 0.456 0.769 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=36.393000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=518.909091, yerr=3.315379 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/eb0d6eed04221145363480ad2019c815711cc597_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 602.320896E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62658. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 3682667. MP_ISendRecv 10710 94533. MP_Wait 16690 MP_comm_split 50 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.029 29.112 29.113 qs_mol_dyn_low 1 2.0 0.003 0.003 28.844 28.852 qs_forces 11 3.9 0.002 0.002 28.753 28.753 qs_energies 11 4.9 0.001 0.001 27.049 27.052 scf_env_do_scf 11 5.9 0.000 0.001 22.084 22.084 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 19.408 19.408 velocity_verlet 10 3.0 0.001 0.001 14.994 14.999 dbcsr_multiply_generic 2286 12.5 0.092 0.094 12.761 12.832 qs_scf_new_mos 108 7.5 0.001 0.001 11.576 11.603 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.576 11.603 ot_scf_mini 108 9.5 0.002 0.002 10.869 10.894 multiply_cannon 2286 13.5 0.229 0.238 9.951 10.487 multiply_cannon_loop 2286 14.5 0.329 0.342 8.974 9.164 multiply_cannon_multrec 9144 15.5 1.631 1.879 5.940 6.146 ot_mini 108 10.5 0.001 0.001 6.079 6.108 rebuild_ks_matrix 119 8.3 0.000 0.000 5.948 5.972 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.948 5.971 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.314 5.335 qs_ot_get_derivative 108 11.5 0.001 0.001 4.754 4.779 dbcsr_mm_accdrv_process 12550 15.8 3.209 4.224 4.208 4.286 sum_up_and_integrate 119 10.3 0.038 0.042 3.677 3.683 integrate_v_rspace 119 11.3 0.003 0.003 3.639 3.646 init_scf_run 11 5.9 0.000 0.001 3.515 3.515 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.515 3.515 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.371 3.377 calculate_rho_elec 119 8.7 0.059 0.061 3.371 3.377 qs_ot_get_p 119 10.4 0.001 0.001 2.935 2.972 init_scf_loop 11 6.9 0.000 0.000 2.654 2.656 mp_waitall_1 121218 16.5 2.053 2.544 2.053 2.544 calculate_first_density_matrix 1 7.0 0.000 0.000 2.305 2.306 make_m2s 4572 13.5 0.034 0.035 1.915 2.047 rs_pw_transfer 974 11.9 0.008 0.008 1.875 1.992 make_images 4572 14.5 0.268 0.301 1.827 1.957 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.934 1.937 prepare_preconditioner 11 7.9 0.000 0.000 1.903 1.907 make_preconditioner 11 8.9 0.000 0.000 1.903 1.907 grid_integrate_task_list 119 12.3 1.846 1.903 1.846 1.903 density_rs2pw 119 9.7 0.003 0.003 1.768 1.879 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.788 1.816 calculate_dm_sparse 119 9.5 0.000 0.000 1.779 1.799 pw_transfer 1439 11.6 0.064 0.066 1.774 1.782 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.737 1.737 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.700 1.713 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.683 1.692 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.690 1.692 jit_kernel_multiply 8 15.7 0.960 1.563 0.960 1.563 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.507 1.515 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.455 1.456 potential_pw2rs 119 12.3 0.010 0.010 1.449 1.453 cp_fm_redistribute_end 50 14.0 0.720 1.428 0.721 1.429 cp_fm_diag_elpa_base 50 14.0 0.661 1.355 0.706 1.413 fft3d_ps 1201 14.6 0.538 0.551 1.349 1.360 fft_wrap_pw1pw2_140 487 13.2 0.082 0.087 1.343 1.353 grid_collocate_task_list 119 9.7 1.273 1.344 1.273 1.344 ot_diis_step 108 11.5 0.012 0.013 1.307 1.307 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.216 1.231 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.222 1.223 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.174 1.199 apply_single 119 13.6 0.000 0.000 1.174 1.199 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.170 1.176 wfi_extrapolate 11 7.9 0.001 0.001 1.160 1.160 hybrid_alltoall_any 4725 16.4 0.062 0.175 0.829 1.073 make_images_data 4572 15.5 0.039 0.042 0.894 1.072 cp_fm_cholesky_invert 11 10.9 0.967 0.970 0.967 0.970 acc_transpose_blocks 9144 15.5 0.038 0.040 0.960 0.968 mp_alltoall_d11v 2130 13.8 0.847 0.926 0.847 0.926 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.869 0.918 multiply_cannon_sync_h2d 9144 15.5 0.716 0.809 0.716 0.809 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.774 0.782 mp_allgather_i34 2286 14.5 0.313 0.762 0.313 0.762 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.758 0.760 qs_env_update_s_mstruct 11 6.9 0.000 0.001 0.680 0.731 acc_transpose_blocks_kernels 9144 16.5 0.116 0.119 0.721 0.724 mp_alltoall_z22v 1201 16.6 0.684 0.723 0.684 0.723 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.368 0.716 dbcsr_complete_redistribute 329 12.2 0.162 0.171 0.583 0.611 jit_kernel_transpose 5 15.6 0.605 0.607 0.605 0.607 mp_sum_l 7207 12.9 0.471 0.599 0.471 0.599 mp_waitany 5200 13.7 0.485 0.592 0.485 0.592 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=29.113000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=569.727273, yerr=4.937627 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/eb0d6eed04221145363480ad2019c815711cc597_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 753.242112E+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_ISendRecv 4998 189067. MP_Wait 8898 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.037 41.122 41.122 qs_mol_dyn_low 1 2.0 0.003 0.003 40.911 40.919 qs_forces 11 3.9 0.002 0.002 40.853 40.854 qs_energies 11 4.9 0.001 0.001 38.897 38.902 scf_env_do_scf 11 5.9 0.001 0.001 33.217 33.217 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 25.128 25.129 velocity_verlet 10 3.0 0.001 0.001 23.350 23.355 dbcsr_multiply_generic 2286 12.5 0.099 0.101 17.571 17.752 qs_scf_new_mos 108 7.5 0.001 0.001 16.059 16.154 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.058 16.154 ot_scf_mini 108 9.5 0.002 0.002 14.963 15.062 multiply_cannon 2286 13.5 0.303 0.310 13.548 14.525 multiply_cannon_loop 2286 14.5 0.342 0.347 12.252 13.152 ot_mini 108 10.5 0.001 0.001 8.912 9.029 multiply_cannon_multrec 9144 15.5 3.349 4.795 8.504 8.590 init_scf_loop 11 6.9 0.000 0.000 8.060 8.062 prepare_preconditioner 11 7.9 0.000 0.000 7.121 7.135 make_preconditioner 11 8.9 0.000 0.000 7.121 7.135 rebuild_ks_matrix 119 8.3 0.000 0.000 6.930 7.078 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 6.930 7.077 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.629 7.010 qs_ot_get_derivative 108 11.5 0.001 0.001 6.870 6.969 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.280 6.414 dbcsr_mm_accdrv_process 12550 15.8 3.956 5.583 5.032 6.328 cp_fm_upper_to_full 72 14.2 3.230 4.661 3.230 4.661 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.784 3.790 calculate_rho_elec 119 8.7 0.118 0.121 3.783 3.789 sum_up_and_integrate 119 10.3 0.064 0.065 3.736 3.742 integrate_v_rspace 119 11.3 0.003 0.003 3.672 3.678 init_scf_run 11 5.9 0.000 0.001 3.633 3.633 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.633 3.633 mp_waitall_1 97218 16.6 2.692 3.608 2.692 3.608 qs_ot_get_p 119 10.4 0.001 0.001 3.286 3.426 dbcsr_complete_redistribute 329 12.2 0.289 0.295 2.090 2.957 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.476 2.919 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.744 2.614 make_m2s 4572 13.5 0.038 0.038 2.400 2.606 make_images 4572 14.5 0.351 0.381 2.280 2.486 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.213 2.468 apply_single 119 13.6 0.000 0.000 2.213 2.468 mp_alltoall_i22 627 13.8 1.528 2.367 1.528 2.367 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.486 2.345 calculate_dm_sparse 119 9.5 0.000 0.000 2.159 2.179 calculate_first_density_matrix 1 7.0 0.000 0.000 2.172 2.174 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 1.311 2.167 pw_transfer 1439 11.6 0.066 0.067 2.097 2.100 grid_integrate_task_list 119 12.3 2.027 2.037 2.027 2.037 ot_diis_step 108 11.5 0.014 0.014 2.012 2.013 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.009 2.011 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.001 2.004 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.936 1.937 density_rs2pw 119 9.7 0.003 0.004 1.854 1.864 mp_sum_l 7207 12.9 1.088 1.824 1.088 1.824 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.735 1.786 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.767 1.768 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.731 1.732 jit_kernel_multiply 10 15.4 1.049 1.719 1.049 1.719 fft_wrap_pw1pw2_140 487 13.2 0.087 0.089 1.656 1.660 fft3d_ps 1201 14.6 0.566 0.577 1.635 1.639 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.596 1.624 cp_fm_cholesky_invert 11 10.9 1.587 1.590 1.587 1.590 hybrid_alltoall_any 4725 16.4 0.086 0.148 1.220 1.509 rs_pw_transfer 974 11.9 0.009 0.009 1.467 1.496 grid_collocate_task_list 119 9.7 1.464 1.475 1.464 1.475 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.423 1.424 make_images_data 4572 15.5 0.042 0.046 1.167 1.422 cp_fm_diag_elpa_base 50 14.0 1.274 1.325 1.421 1.421 wfi_extrapolate 11 7.9 0.001 0.001 1.401 1.401 mp_alltoall_d11v 2130 13.8 1.232 1.249 1.232 1.249 potential_pw2rs 119 12.3 0.014 0.014 1.243 1.245 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.215 1.225 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.160 1.182 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.103 1.120 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.988 1.053 multiply_cannon_sync_h2d 9144 15.5 1.044 1.048 1.044 1.048 qs_create_task_list 11 7.9 0.001 0.001 0.945 0.958 generate_qs_task_list 11 8.9 0.374 0.394 0.945 0.957 mp_alltoall_z22v 1201 16.6 0.933 0.956 0.933 0.956 acc_transpose_blocks 9144 15.5 0.038 0.039 0.898 0.902 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.863 0.877 copy_dbcsr_to_fm 153 11.3 0.002 0.002 0.818 0.864 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=41.122000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=704.454545, yerr=16.030446 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/eb0d6eed04221145363480ad2019c815711cc597_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 502.681600E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65372. MP_Allreduce 9840 486. MP_Sync 100 MP_Alltoall 1938 1711591. MP_ISendRecv 41800 9096. MP_Wait 58168 MP_comm_split 48 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.037 0.067 84.988 85.003 qs_mol_dyn_low 1 2.0 0.005 0.012 84.587 84.597 qs_forces 11 3.9 0.002 0.003 84.507 84.509 qs_energies 11 4.9 0.001 0.006 81.633 81.650 scf_env_do_scf 11 5.9 0.002 0.004 72.690 72.692 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 66.773 66.774 dbcsr_multiply_generic 2055 12.4 0.104 0.109 53.043 53.353 qs_scf_new_mos 99 7.5 0.000 0.001 49.281 49.382 qs_scf_loop_do_ot 99 8.5 0.001 0.001 49.281 49.381 ot_scf_mini 99 9.5 0.002 0.003 46.807 46.870 multiply_cannon 2055 13.4 0.178 0.187 42.855 43.905 velocity_verlet 10 3.0 0.001 0.001 43.510 43.511 multiply_cannon_loop 2055 14.4 1.566 1.599 41.778 42.700 ot_mini 99 10.5 0.001 0.001 28.448 28.507 qs_ot_get_derivative 99 11.5 0.001 0.002 21.521 21.588 multiply_cannon_multrec 49320 15.4 12.101 12.858 17.129 17.858 rebuild_ks_matrix 110 8.3 0.000 0.001 14.807 14.917 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.018 14.807 14.916 mp_waitall_1 241148 16.1 12.473 13.781 12.473 13.781 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.008 13.098 multiply_cannon_sync_h2d 49320 15.4 9.843 10.372 9.843 10.372 qs_ot_get_p 110 10.4 0.001 0.002 9.742 9.838 multiply_cannon_metrocomm3 49320 15.4 0.079 0.083 7.056 8.107 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.536 8.098 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.211 7.687 apply_single 110 13.6 0.000 0.000 7.211 7.687 sum_up_and_integrate 110 10.3 0.037 0.044 7.272 7.287 integrate_v_rspace 110 11.3 0.003 0.003 7.235 7.259 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.803 6.857 init_scf_run 11 5.9 0.000 0.002 6.830 6.831 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.830 6.831 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.502 6.628 calculate_rho_elec 110 8.6 0.020 0.024 6.501 6.627 ot_diis_step 99 11.5 0.005 0.006 6.624 6.625 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.272 6.302 mp_sum_l 6514 12.8 5.455 6.222 5.455 6.222 init_scf_loop 11 6.9 0.001 0.004 5.889 5.890 dbcsr_mm_accdrv_process 87628 16.1 2.067 2.194 4.907 5.233 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.229 5.229 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.699 4.700 cp_fm_redistribute_end 48 14.0 4.090 4.671 4.095 4.673 cp_fm_diag_elpa_base 48 14.0 0.571 4.577 0.575 4.603 make_m2s 4110 13.4 0.060 0.065 4.294 4.394 make_images 4110 14.4 0.177 0.193 4.199 4.301 rs_pw_transfer 902 11.9 0.012 0.013 4.091 4.277 wfi_extrapolate 11 7.9 0.001 0.001 4.126 4.126 calculate_dm_sparse 110 9.5 0.000 0.001 3.839 3.920 multiply_cannon_metrocomm1 49320 15.4 0.058 0.062 2.907 3.876 density_rs2pw 110 9.6 0.004 0.004 3.552 3.716 prepare_preconditioner 11 7.9 0.000 0.000 3.655 3.676 make_preconditioner 11 8.9 0.000 0.002 3.655 3.676 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 3.548 3.553 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.434 3.498 grid_integrate_task_list 110 12.3 3.238 3.413 3.238 3.413 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.350 3.399 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.202 3.258 pw_transfer 1331 11.6 0.055 0.068 3.109 3.188 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 3.021 3.101 potential_pw2rs 110 12.3 0.006 0.007 2.820 2.846 fft3d_ps 1111 14.6 0.750 0.837 2.556 2.635 calculate_first_density_matrix 1 7.0 0.001 0.002 2.616 2.620 mp_alltoall_d11v 2046 13.8 2.269 2.610 2.269 2.610 jit_kernel_multiply 13 15.9 2.555 2.594 2.555 2.594 fft_wrap_pw1pw2_140 451 13.1 0.170 0.189 2.515 2.593 acc_transpose_blocks 49320 15.4 0.227 0.235 2.198 2.282 grid_collocate_task_list 110 9.6 2.088 2.233 2.088 2.233 mp_waitany 14300 13.8 1.915 2.134 1.915 2.134 make_images_data 4110 15.4 0.044 0.047 1.936 2.084 mp_sum_d 3879 11.9 1.595 2.084 1.595 2.084 cp_fm_cholesky_invert 11 10.9 2.006 2.010 2.006 2.010 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.943 1.959 hybrid_alltoall_any 4261 16.3 0.083 0.479 1.674 1.911 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=85.003000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=477.363636, yerr=2.603240 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/eb0d6eed04221145363480ad2019c815711cc597_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 589.987840E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1972800 MPI messages size (bytes): total size 1.077520E+12 min size 0.000000E+00 max size 4.537280E+06 average size 546.188250E+03 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 222984 1826684928 8192 < size <= 32768 520356 13399818240 32768 < size <= 131072 372336 35386294272 131072 < size <= 4194304 787758 788321309808 4194304 < size <= 16777216 54450 238588003280 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65587. MP_Allreduce 9839 562. MP_Sync 100 MP_Alltoall 1717 2742099. MP_ISendRecv 20680 26400. MP_Wait 32692 MP_comm_split 48 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.027 0.048 70.721 70.722 qs_mol_dyn_low 1 2.0 0.003 0.004 70.393 70.403 qs_forces 11 3.9 0.002 0.003 70.328 70.330 qs_energies 11 4.9 0.001 0.001 67.020 67.025 scf_env_do_scf 11 5.9 0.000 0.001 58.089 58.093 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 49.988 49.989 dbcsr_multiply_generic 2055 12.4 0.112 0.118 38.415 38.688 velocity_verlet 10 3.0 0.001 0.001 37.232 37.233 qs_scf_new_mos 99 7.5 0.001 0.001 33.722 33.850 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.722 33.849 multiply_cannon 2055 13.4 0.222 0.246 31.371 32.718 ot_scf_mini 99 9.5 0.003 0.003 32.016 32.131 multiply_cannon_loop 2055 14.4 0.923 0.952 30.049 31.187 ot_mini 99 10.5 0.001 0.001 18.747 18.882 multiply_cannon_multrec 24660 15.4 7.551 9.129 13.977 15.475 rebuild_ks_matrix 110 8.3 0.000 0.001 13.792 13.957 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 13.791 13.956 qs_ot_get_derivative 99 11.5 0.001 0.001 12.858 12.984 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.173 12.320 mp_waitall_1 186928 16.3 8.456 11.367 8.456 11.367 multiply_cannon_metrocomm3 24660 15.4 0.068 0.069 5.458 8.447 init_scf_loop 11 6.9 0.000 0.000 8.063 8.064 multiply_cannon_sync_h2d 24660 15.4 6.862 7.691 6.862 7.691 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.542 7.234 apply_single 110 13.6 0.000 0.001 6.542 7.234 qs_ot_get_p 110 10.4 0.001 0.001 6.594 6.728 sum_up_and_integrate 110 10.3 0.055 0.058 6.698 6.708 integrate_v_rspace 110 11.3 0.002 0.002 6.642 6.655 dbcsr_mm_accdrv_process 52282 16.1 4.826 5.683 6.269 6.580 init_scf_run 11 5.9 0.000 0.002 6.536 6.536 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.536 6.536 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.041 6.054 calculate_rho_elec 110 8.6 0.039 0.047 6.040 6.053 prepare_preconditioner 11 7.9 0.000 0.000 6.022 6.038 make_preconditioner 11 8.9 0.000 0.000 6.022 6.038 ot_diis_step 99 11.5 0.010 0.010 5.826 5.827 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.591 5.747 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.793 5.485 make_m2s 4110 13.4 0.056 0.059 4.389 4.845 make_images 4110 14.4 0.401 0.444 4.283 4.736 qs_ot_p2m_diag 48 11.0 0.028 0.044 4.574 4.594 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.132 4.133 wfi_extrapolate 11 7.9 0.001 0.001 3.631 3.631 pw_transfer 1331 11.6 0.066 0.071 3.483 3.629 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.549 3.551 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.377 3.524 cp_fm_redistribute_end 48 14.0 2.644 3.518 2.646 3.518 cp_fm_diag_elpa_base 48 14.0 0.832 3.383 0.867 3.472 density_rs2pw 110 9.6 0.004 0.004 3.290 3.453 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.374 3.444 grid_integrate_task_list 110 12.3 3.149 3.358 3.149 3.358 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.345 3.348 rs_pw_transfer 902 11.9 0.012 0.013 3.177 3.340 calculate_dm_sparse 110 9.5 0.001 0.001 3.244 3.280 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.957 3.008 make_images_data 4110 15.4 0.047 0.051 2.461 2.940 fft_wrap_pw1pw2_140 451 13.1 0.199 0.216 2.751 2.900 cp_fm_cholesky_invert 11 10.9 2.874 2.883 2.874 2.883 fft3d_ps 1111 14.6 1.074 1.274 2.735 2.871 mp_sum_l 6514 12.8 2.051 2.852 2.051 2.852 hybrid_alltoall_any 4261 16.3 0.103 0.445 2.171 2.833 calculate_first_density_matrix 1 7.0 0.000 0.001 2.818 2.821 jit_kernel_multiply 13 16.3 1.093 2.722 1.093 2.722 potential_pw2rs 110 12.3 0.008 0.009 2.531 2.556 grid_collocate_task_list 110 9.6 2.121 2.287 2.121 2.287 mp_alltoall_d11v 2046 13.8 1.931 2.226 1.931 2.226 multiply_cannon_metrocomm4 22605 15.4 0.074 0.079 0.827 2.173 qs_ot_get_orbitals 99 10.5 0.001 0.001 2.022 2.047 mp_irecv_dv 57340 16.2 0.703 1.956 0.703 1.956 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 1.859 1.861 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.785 1.799 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.775 1.789 mp_allgather_i34 2055 14.4 0.650 1.706 0.650 1.706 acc_transpose_blocks 24660 15.4 0.110 0.112 1.537 1.559 dbcsr_complete_redistribute 325 12.2 0.242 0.300 1.237 1.506 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.368 1.473 mp_alltoall_z22v 1111 16.6 1.390 1.470 1.390 1.470 cp_fm_cholesky_decompose 22 10.9 1.453 1.459 1.453 1.459 mp_sum_d 3879 11.9 0.989 1.433 0.989 1.433 mp_waitany 10164 13.8 1.221 1.416 1.221 1.416 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=70.722000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=556.181818, yerr=7.767276 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/eb0d6eed04221145363480ad2019c815711cc597_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 659.845120E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65578. MP_Allreduce 9838 559. MP_Sync 100 MP_Alltoall 1496 4511006. MP_ISendRecv 13640 27424. MP_Wait 32318 MP_comm_split 48 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.033 62.688 62.689 qs_mol_dyn_low 1 2.0 0.003 0.003 62.363 62.383 qs_forces 11 3.9 0.002 0.002 62.296 62.297 qs_energies 11 4.9 0.002 0.008 59.031 59.036 scf_env_do_scf 11 5.9 0.001 0.001 50.881 50.881 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 41.667 41.667 velocity_verlet 10 3.0 0.001 0.001 34.130 34.132 dbcsr_multiply_generic 2055 12.4 0.108 0.110 29.662 30.064 qs_scf_new_mos 99 7.5 0.001 0.001 26.311 26.419 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.311 26.418 ot_scf_mini 99 9.5 0.002 0.003 25.044 25.176 multiply_cannon 2055 13.4 0.212 0.234 22.638 24.057 multiply_cannon_loop 2055 14.4 0.616 0.637 21.283 22.838 ot_mini 99 10.5 0.001 0.001 14.439 14.582 rebuild_ks_matrix 110 8.3 0.000 0.001 12.563 12.767 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.015 12.563 12.767 mp_waitall_1 146766 16.3 8.343 11.331 8.343 11.331 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.083 11.266 multiply_cannon_multrec 16440 15.4 3.721 4.819 9.438 10.453 qs_ot_get_derivative 99 11.5 0.001 0.001 9.839 9.973 init_scf_loop 11 6.9 0.001 0.004 9.168 9.168 multiply_cannon_metrocomm3 16440 15.4 0.043 0.045 4.883 8.019 prepare_preconditioner 11 7.9 0.000 0.000 7.330 7.364 make_preconditioner 11 8.9 0.000 0.002 7.330 7.364 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.649 6.994 sum_up_and_integrate 110 10.3 0.061 0.062 6.705 6.720 integrate_v_rspace 110 11.3 0.002 0.003 6.644 6.658 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.099 6.117 calculate_rho_elec 110 8.6 0.058 0.058 6.098 6.116 qs_ot_get_p 110 10.4 0.001 0.001 5.615 5.793 dbcsr_mm_accdrv_process 34862 16.1 4.498 4.884 5.574 5.691 init_scf_run 11 5.9 0.000 0.002 5.669 5.669 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.669 5.669 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.073 5.559 apply_single 110 13.6 0.000 0.000 5.073 5.558 make_m2s 4110 13.4 0.050 0.052 4.550 4.904 make_images 4110 14.4 0.393 0.512 4.438 4.791 ot_diis_step 99 11.5 0.010 0.011 4.560 4.560 multiply_cannon_sync_h2d 16440 15.4 3.668 4.091 3.668 4.091 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.319 3.995 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.807 3.814 density_rs2pw 110 9.6 0.004 0.004 3.288 3.500 pw_transfer 1331 11.6 0.066 0.071 3.460 3.482 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.439 3.440 grid_integrate_task_list 110 12.3 3.231 3.378 3.231 3.378 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.353 3.376 rs_pw_transfer 902 11.9 0.010 0.011 3.133 3.344 wfi_extrapolate 11 7.9 0.001 0.001 3.137 3.137 make_images_data 4110 15.4 0.043 0.047 2.618 3.099 cp_fm_cholesky_invert 11 10.9 3.005 3.011 3.005 3.011 hybrid_alltoall_any 4261 16.3 0.105 0.375 2.332 2.998 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.871 2.873 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.832 2.835 cp_fm_redistribute_end 48 14.0 1.760 2.804 1.762 2.805 fft_wrap_pw1pw2_140 451 13.1 0.208 0.212 2.773 2.800 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.718 2.791 cp_fm_diag_elpa_base 48 14.0 0.978 2.659 1.036 2.767 fft3d_ps 1111 14.6 1.069 1.086 2.667 2.683 mp_sum_l 6514 12.8 1.860 2.660 1.860 2.660 calculate_dm_sparse 110 9.5 0.001 0.001 2.596 2.628 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.487 2.550 multiply_cannon_metrocomm4 14385 15.4 0.044 0.049 0.916 2.548 mp_alltoall_d11v 2046 13.8 2.047 2.460 2.047 2.460 potential_pw2rs 110 12.3 0.011 0.011 2.446 2.459 calculate_first_density_matrix 1 7.0 0.000 0.003 2.444 2.446 mp_irecv_dv 48980 15.7 0.847 2.424 0.847 2.424 grid_collocate_task_list 110 9.6 2.176 2.327 2.176 2.327 dbcsr_complete_redistribute 325 12.2 0.317 0.350 1.607 2.071 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.975 1.976 cp_fm_upper_to_full 70 13.6 1.392 1.862 1.392 1.862 mp_allgather_i34 2055 14.4 0.680 1.823 0.680 1.823 cp_fm_cholesky_decompose 22 10.9 1.740 1.763 1.740 1.763 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.686 1.708 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.521 1.534 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.373 1.505 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.035 1.486 jit_kernel_multiply 8 16.3 0.687 1.481 0.687 1.481 mp_waitany 17072 13.8 1.226 1.468 1.226 1.468 mp_alltoall_z22v 1111 16.6 1.384 1.413 1.384 1.413 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.372 1.386 rs_gather_matrices 110 12.3 0.138 0.150 0.910 1.331 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=62.689000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=623.909091, yerr=9.510325 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/eb0d6eed04221145363480ad2019c815711cc597_performance_tests/12/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 601.317074E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 4916280 0.0% 0.0% 100.0% average stack size 0.0 0.0 200.2 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 732.143616E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65576. MP_Allreduce 9838 600. MP_Sync 100 MP_Alltoall 1496 5863162. MP_ISendRecv 10120 43184. MP_Wait 25102 MP_comm_split 48 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.036 0.047 67.945 67.945 qs_mol_dyn_low 1 2.0 0.003 0.003 67.351 67.359 qs_forces 11 3.9 0.002 0.002 67.275 67.275 qs_energies 11 4.9 0.001 0.004 63.824 63.829 scf_env_do_scf 11 5.9 0.001 0.005 55.245 55.248 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 42.854 42.855 velocity_verlet 10 3.0 0.001 0.001 38.261 38.263 dbcsr_multiply_generic 2055 12.4 0.114 0.117 30.733 31.036 qs_scf_new_mos 99 7.5 0.001 0.001 27.958 28.055 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.957 28.054 ot_scf_mini 99 9.5 0.003 0.003 26.299 26.399 multiply_cannon 2055 13.4 0.247 0.263 23.036 24.175 multiply_cannon_loop 2055 14.4 0.886 0.908 21.411 22.185 ot_mini 99 10.5 0.001 0.001 15.085 15.189 multiply_cannon_multrec 24660 15.4 4.184 6.651 12.722 13.881 rebuild_ks_matrix 110 8.3 0.000 0.000 12.316 12.422 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.315 12.421 init_scf_loop 11 6.9 0.000 0.002 12.340 12.340 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.874 10.964 qs_ot_get_derivative 99 11.5 0.001 0.001 10.855 10.954 prepare_preconditioner 11 7.9 0.000 0.000 10.570 10.588 make_preconditioner 11 8.9 0.000 0.002 10.570 10.588 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.790 10.250 dbcsr_mm_accdrv_process 52304 16.0 6.983 8.283 8.391 9.375 mp_waitall_1 126806 16.4 5.257 7.552 5.257 7.552 sum_up_and_integrate 110 10.3 0.068 0.071 6.642 6.655 integrate_v_rspace 110 11.3 0.003 0.003 6.574 6.589 make_m2s 4110 13.4 0.059 0.061 5.862 6.311 qs_ot_get_p 110 10.4 0.001 0.001 6.075 6.235 make_images 4110 14.4 0.577 0.700 5.722 6.167 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.843 5.851 calculate_rho_elec 110 8.6 0.077 0.081 5.842 5.850 init_scf_run 11 5.9 0.000 0.001 5.820 5.820 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.819 5.820 cp_fm_upper_to_full 70 13.8 3.285 4.668 3.285 4.668 ot_diis_step 99 11.5 0.011 0.011 4.188 4.188 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.086 4.174 apply_single 110 13.6 0.000 0.000 4.086 4.174 qs_ot_p2m_diag 48 11.0 0.055 0.064 4.115 4.131 dbcsr_complete_redistribute 325 12.2 0.418 0.462 2.816 3.950 multiply_cannon_metrocomm3 24660 15.4 0.035 0.036 1.653 3.651 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.585 3.587 pw_transfer 1331 11.6 0.065 0.073 3.514 3.540 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.465 3.515 make_images_data 4110 15.4 0.046 0.050 3.007 3.511 hybrid_alltoall_any 4261 16.3 0.120 0.459 2.580 3.489 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.408 3.437 grid_integrate_task_list 110 12.3 3.267 3.428 3.267 3.428 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.357 3.415 multiply_cannon_sync_h2d 24660 15.4 3.179 3.402 3.179 3.402 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.201 3.313 density_rs2pw 110 9.6 0.004 0.004 2.985 3.184 cp_fm_cholesky_invert 11 10.9 3.141 3.150 3.141 3.150 wfi_extrapolate 11 7.9 0.001 0.001 3.143 3.143 calculate_dm_sparse 110 9.5 0.001 0.001 3.057 3.092 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.950 2.951 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.916 2.919 rs_pw_transfer 902 11.9 0.010 0.011 2.691 2.916 cp_fm_redistribute_end 48 14.0 1.449 2.884 1.451 2.886 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.770 2.862 cp_fm_diag_elpa_base 48 14.0 1.351 2.751 1.430 2.862 mp_alltoall_i22 605 13.7 1.665 2.836 1.665 2.836 fft_wrap_pw1pw2_140 451 13.1 0.201 0.213 2.776 2.806 fft3d_ps 1111 14.6 1.063 1.092 2.711 2.735 calculate_first_density_matrix 1 7.0 0.000 0.000 2.582 2.583 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.499 2.535 potential_pw2rs 110 12.3 0.012 0.013 2.387 2.399 grid_collocate_task_list 110 9.6 2.219 2.385 2.219 2.385 jit_kernel_multiply 12 15.6 1.074 2.341 1.074 2.341 mp_alltoall_d11v 2046 13.8 2.033 2.340 2.033 2.340 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.251 2.252 cp_fm_cholesky_decompose 22 10.9 1.904 1.943 1.904 1.943 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.769 1.800 mp_sum_l 6514 12.8 1.164 1.784 1.164 1.784 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.691 1.705 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.597 1.704 mp_allgather_i34 2055 14.4 0.721 1.681 0.721 1.681 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.639 1.654 acc_transpose_blocks 24660 15.4 0.103 0.106 1.503 1.526 mp_alltoall_z22v 1111 16.6 1.432 1.493 1.432 1.493 multiply_cannon_metrocomm4 20550 15.4 0.056 0.060 0.868 1.467 mp_waitany 13376 13.8 1.133 1.430 1.133 1.430 mp_irecv_dv 62702 16.1 0.771 1.384 0.771 1.384 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=67.945000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=689.545455, yerr=9.138964 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/eb0d6eed04221145363480ad2019c815711cc597_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 837.750784E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65574. MP_Allreduce 9838 640. MP_Sync 100 MP_Alltoall 1496 8504061. MP_ISendRecv 6600 54848. MP_Wait 17226 MP_comm_split 48 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.048 0.063 55.054 55.054 qs_mol_dyn_low 1 2.0 0.003 0.003 54.654 54.663 qs_forces 11 3.9 0.002 0.002 54.571 54.571 qs_energies 11 4.9 0.001 0.001 50.974 50.979 scf_env_do_scf 11 5.9 0.000 0.001 42.500 42.502 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 34.751 34.752 velocity_verlet 10 3.0 0.001 0.001 30.605 30.609 dbcsr_multiply_generic 2055 12.4 0.104 0.106 23.222 23.359 qs_scf_new_mos 99 7.5 0.001 0.001 20.632 20.674 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.631 20.674 ot_scf_mini 99 9.5 0.002 0.002 19.378 19.401 multiply_cannon 2055 13.4 0.247 0.262 17.616 19.034 multiply_cannon_loop 2055 14.4 0.321 0.334 16.236 16.541 rebuild_ks_matrix 110 8.3 0.000 0.001 11.462 11.493 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.461 11.492 ot_mini 99 10.5 0.001 0.001 10.505 10.521 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.208 10.237 multiply_cannon_multrec 8220 15.4 3.237 4.475 7.709 8.719 mp_waitall_1 106626 16.5 6.204 7.745 6.204 7.745 init_scf_loop 11 6.9 0.000 0.000 7.699 7.700 qs_ot_get_derivative 99 11.5 0.001 0.001 6.703 6.724 sum_up_and_integrate 110 10.3 0.080 0.082 6.191 6.205 integrate_v_rspace 110 11.3 0.004 0.012 6.110 6.124 prepare_preconditioner 11 7.9 0.000 0.000 6.069 6.073 make_preconditioner 11 8.9 0.000 0.000 6.069 6.073 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.638 5.717 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.617 5.630 calculate_rho_elec 110 8.6 0.115 0.115 5.616 5.630 init_scf_run 11 5.9 0.000 0.003 5.332 5.332 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.332 5.332 dbcsr_mm_accdrv_process 17442 15.9 2.945 4.035 4.343 5.270 qs_ot_get_p 110 10.4 0.001 0.001 4.924 4.945 make_m2s 4110 13.4 0.039 0.040 4.330 4.592 make_images 4110 14.4 0.640 0.689 4.202 4.461 multiply_cannon_metrocomm3 8220 15.4 0.017 0.018 2.959 4.251 ot_diis_step 99 11.5 0.012 0.012 3.779 3.780 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.687 3.715 apply_single 110 13.6 0.000 0.000 3.687 3.714 qs_ot_p2m_diag 48 11.0 0.081 0.085 3.588 3.592 grid_integrate_task_list 110 12.3 3.363 3.511 3.363 3.511 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.284 3.284 pw_transfer 1331 11.6 0.065 0.070 3.256 3.265 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.148 3.160 multiply_cannon_sync_h2d 8220 15.4 2.913 3.041 2.913 3.041 make_images_data 4110 15.4 0.039 0.043 2.476 2.906 hybrid_alltoall_any 4261 16.3 0.200 0.862 2.394 2.893 cp_fm_cholesky_invert 11 10.9 2.863 2.867 2.863 2.867 density_rs2pw 110 9.6 0.004 0.004 2.670 2.862 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.774 2.776 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.760 2.761 cp_fm_redistribute_end 48 14.0 0.697 2.734 0.702 2.735 fft_wrap_pw1pw2_140 451 13.1 0.212 0.217 2.699 2.716 cp_fm_diag_elpa_base 48 14.0 1.850 2.548 2.025 2.705 wfi_extrapolate 11 7.9 0.001 0.001 2.689 2.690 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.672 2.673 calculate_first_density_matrix 1 7.0 0.000 0.000 2.555 2.555 calculate_dm_sparse 110 9.5 0.001 0.001 2.511 2.552 fft3d_ps 1111 14.6 1.122 1.157 2.406 2.419 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.403 2.419 grid_collocate_task_list 110 9.6 2.319 2.411 2.319 2.411 rs_pw_transfer 902 11.9 0.010 0.010 2.176 2.385 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.768 1.982 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.955 1.969 potential_pw2rs 110 12.3 0.015 0.016 1.921 1.925 mp_alltoall_d11v 2046 13.8 1.679 1.903 1.679 1.903 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.826 1.836 cp_fm_cholesky_decompose 22 10.9 1.733 1.743 1.733 1.743 mp_allgather_i34 2055 14.4 0.543 1.661 0.543 1.661 jit_kernel_multiply 8 15.4 1.089 1.657 1.089 1.657 qs_env_update_s_mstruct 11 6.9 0.001 0.002 1.517 1.636 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.586 1.589 dbcsr_complete_redistribute 325 12.2 0.567 0.605 1.475 1.568 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.465 1.476 qs_create_task_list 11 7.9 0.009 0.024 1.247 1.346 generate_qs_task_list 11 8.9 0.379 0.447 1.237 1.345 mp_waitany 9240 13.8 1.060 1.268 1.060 1.268 multiply_cannon_metrocomm4 6165 15.4 0.017 0.018 0.487 1.163 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.141 1.155 mp_irecv_dv 24056 15.7 0.465 1.121 0.465 1.121 multiply_cannon_metrocomm1 8220 15.4 0.020 0.021 0.816 1.112 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=55.054000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=788.272727, yerr=10.762654 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/eb0d6eed04221145363480ad2019c815711cc597_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 1.333727E+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_ISendRecv 4620 360267. MP_Wait 7524 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 0.033 0.040 86.072 86.073 qs_mol_dyn_low 1 2.0 0.003 0.003 85.739 85.748 qs_forces 11 3.9 0.002 0.002 85.674 85.674 qs_energies 11 4.9 0.001 0.001 81.607 81.610 scf_env_do_scf 11 5.9 0.001 0.001 71.693 71.693 velocity_verlet 10 3.0 0.001 0.001 55.170 55.176 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 43.210 43.212 dbcsr_multiply_generic 2055 12.4 0.118 0.121 29.003 29.133 init_scf_loop 11 6.9 0.000 0.000 28.411 28.415 qs_scf_new_mos 99 7.5 0.001 0.001 26.659 26.751 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.658 26.750 prepare_preconditioner 11 7.9 0.000 0.000 26.429 26.440 make_preconditioner 11 8.9 0.000 0.000 26.429 26.440 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.801 25.899 ot_scf_mini 99 9.5 0.002 0.002 24.876 24.947 multiply_cannon 2055 13.4 0.350 0.375 21.742 22.495 multiply_cannon_loop 2055 14.4 0.340 0.346 19.907 20.247 cp_fm_upper_to_full 70 14.2 12.567 17.903 12.567 17.903 ot_mini 99 10.5 0.001 0.001 13.897 13.972 rebuild_ks_matrix 110 8.3 0.001 0.001 13.254 13.346 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 13.253 13.345 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.046 12.129 dbcsr_complete_redistribute 325 12.2 1.026 1.042 7.265 10.306 multiply_cannon_multrec 8220 15.4 4.385 4.699 9.586 9.680 qs_ot_get_derivative 99 11.5 0.001 0.001 9.183 9.252 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.202 9.248 mp_waitall_1 87304 16.6 8.102 9.188 8.102 9.188 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.612 8.631 mp_alltoall_i22 605 13.7 5.255 8.285 5.255 8.285 sum_up_and_integrate 110 10.3 0.151 0.152 6.685 6.701 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.559 6.592 calculate_rho_elec 110 8.6 0.227 0.227 6.558 6.591 integrate_v_rspace 110 11.3 0.004 0.004 6.534 6.551 make_m2s 4110 13.4 0.043 0.043 5.345 5.889 qs_ot_get_p 110 10.4 0.001 0.001 5.709 5.802 make_images 4110 14.4 0.877 0.931 5.157 5.700 init_scf_run 11 5.9 0.000 0.001 5.675 5.675 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.675 5.675 cp_fm_cholesky_invert 11 10.9 5.532 5.536 5.532 5.536 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 4.975 5.353 dbcsr_mm_accdrv_process 11614 15.7 3.251 3.658 5.058 5.319 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.766 5.280 apply_single 110 13.6 0.000 0.000 4.766 5.280 ot_diis_step 99 11.5 0.015 0.016 4.685 4.685 qs_ot_p2m_diag 48 11.0 0.151 0.155 4.109 4.116 multiply_cannon_sync_h2d 8220 15.4 3.948 3.954 3.948 3.954 pw_transfer 1331 11.6 0.074 0.075 3.819 3.826 hybrid_alltoall_any 4261 16.3 0.255 0.555 2.955 3.733 grid_integrate_task_list 110 12.3 3.662 3.723 3.662 3.723 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 3.702 3.709 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.659 3.659 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.629 3.629 make_images_data 4110 15.4 0.041 0.044 2.943 3.619 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.902 3.368 wfi_extrapolate 11 7.9 0.001 0.001 3.233 3.233 fft_wrap_pw1pw2_140 451 13.1 0.215 0.217 3.194 3.200 calculate_dm_sparse 110 9.5 0.001 0.001 3.175 3.197 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.043 3.043 cp_fm_diag_elpa_base 48 14.0 2.522 2.707 3.041 3.041 density_rs2pw 110 9.6 0.004 0.004 3.014 3.029 fft3d_ps 1111 14.6 1.270 1.277 2.911 2.919 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.890 2.894 grid_collocate_task_list 110 9.6 2.631 2.656 2.631 2.656 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.447 2.488 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.369 2.399 calculate_first_density_matrix 1 7.0 0.000 0.000 2.338 2.338 qs_env_update_s_mstruct 11 6.9 0.001 0.001 2.173 2.229 mp_alltoall_d11v 2046 13.8 2.161 2.214 2.161 2.214 rs_pw_transfer 902 11.9 0.010 0.011 2.158 2.199 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.094 2.191 cp_fm_cholesky_decompose 22 10.9 2.102 2.118 2.102 2.118 potential_pw2rs 110 12.3 0.021 0.021 2.092 2.098 qs_create_task_list 11 7.9 0.001 0.001 1.899 1.943 generate_qs_task_list 11 8.9 0.741 0.793 1.898 1.942 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.847 1.907 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.816 1.824 jit_kernel_multiply 10 15.3 1.604 1.776 1.604 1.776 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=86.073000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1205.818182, yerr=51.689313 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/eb0d6eed04221145363480ad2019c815711cc597_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410023282688 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444707676160 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796579E+12 0.0% 0.0% 100.0% flops max/rank 1.086553E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500928 0.0% 0.0% 100.0% number of processed stacks 11851392 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.8 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 627.941376E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10258080 MPI messages size (bytes): total size 4.456716E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.459062E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3542056 94711185408 32768 < size <= 131072 1282176 73356279808 131072 < size <= 4194304 5107038 3151762635464 4194304 < size <= 16777216 259842 1136842803272 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4083 56784. MP_Allreduce 11248 786. MP_Sync 168 MP_Alltoall 2210 2744833. MP_ISendRecv 48260 18752. MP_Wait 66280 MP_comm_split 82 MP_ISend 15900 108037. MP_IRecv 15900 108037. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.032 202.681 202.682 qs_mol_dyn_low 1 2.0 0.003 0.003 202.295 202.341 qs_forces 11 3.9 0.003 0.003 202.209 202.209 qs_energies 11 4.9 0.001 0.001 196.686 196.699 scf_env_do_scf 11 5.9 0.001 0.002 180.245 180.249 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 158.921 158.922 dbcsr_multiply_generic 2485 12.5 0.171 0.177 123.123 123.847 velocity_verlet 10 3.0 0.001 0.001 121.696 121.696 qs_scf_new_mos 116 7.6 0.001 0.001 120.326 120.642 qs_scf_loop_do_ot 116 8.6 0.001 0.001 120.325 120.641 ot_scf_mini 116 9.6 0.003 0.003 113.848 114.145 multiply_cannon 2485 13.5 0.233 0.242 99.826 101.363 multiply_cannon_loop 2485 14.5 2.111 2.188 97.648 99.243 ot_mini 116 10.6 0.001 0.001 65.244 65.604 multiply_cannon_multrec 59640 15.5 32.966 35.611 41.202 43.315 qs_ot_get_derivative 116 11.6 0.001 0.001 40.529 40.849 rebuild_ks_matrix 127 8.3 0.001 0.001 33.107 33.376 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.025 33.106 33.375 mp_waitall_1 288940 16.2 28.485 30.977 28.485 30.977 qs_ks_update_qs_env 127 7.6 0.001 0.001 29.757 29.973 multiply_cannon_sync_h2d 59640 15.5 27.035 28.728 27.035 28.728 qs_ot_get_p 127 10.4 0.001 0.001 26.833 27.106 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 24.227 24.797 apply_single 127 13.6 0.001 0.001 24.226 24.797 ot_diis_step 116 11.6 0.007 0.008 24.363 24.364 init_scf_loop 11 6.9 0.000 0.000 21.246 21.247 qs_ot_p2m_diag 82 11.4 0.076 0.090 20.199 20.271 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 18.798 19.013 cp_dbcsr_syevd 82 12.4 0.004 0.005 17.714 17.715 multiply_cannon_metrocomm3 59640 15.5 0.115 0.120 15.459 17.672 prepare_preconditioner 11 7.9 0.000 0.000 16.667 16.716 make_preconditioner 11 8.9 0.000 0.000 16.667 16.716 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.859 16.043 cp_fm_diag_elpa 82 13.4 0.000 0.001 14.767 14.776 cp_fm_redistribute_end 82 14.4 11.632 14.689 11.645 14.693 cp_fm_diag_elpa_base 82 14.4 3.003 14.455 3.035 14.577 make_m2s 4970 13.5 0.102 0.110 13.903 14.321 make_images 4970 14.5 0.403 0.420 13.732 14.158 sum_up_and_integrate 127 10.3 0.089 0.107 13.985 14.006 integrate_v_rspace 127 11.3 0.003 0.004 13.895 13.920 qs_rho_update_rho_low 127 7.7 0.001 0.001 12.672 12.793 calculate_rho_elec 127 8.7 0.044 0.063 12.672 12.792 init_scf_run 11 5.9 0.000 0.002 12.280 12.280 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.280 12.280 mp_sum_l 7804 13.0 8.144 9.898 8.144 9.898 cp_fm_cholesky_invert 11 10.9 9.482 9.490 9.482 9.490 wfi_extrapolate 11 7.9 0.001 0.001 9.094 9.094 calculate_dm_sparse 127 9.5 0.001 0.001 8.276 8.384 dbcsr_mm_accdrv_process 123452 16.2 3.220 3.382 7.807 8.361 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.909 8.064 qs_ot_get_orbitals 116 10.6 0.001 0.001 7.795 7.866 multiply_cannon_metrocomm1 59640 15.5 0.085 0.088 5.949 7.694 make_images_data 4970 15.5 0.069 0.075 6.799 7.626 grid_integrate_task_list 127 12.3 7.044 7.534 7.044 7.534 hybrid_alltoall_any 5155 16.4 0.292 2.254 5.925 7.188 density_rs2pw 127 9.7 0.006 0.007 6.534 7.138 pw_transfer 1535 11.6 0.075 0.109 6.819 7.111 fft_wrap_pw1pw2 1281 12.7 0.010 0.013 6.616 6.884 cp_dbcsr_sm_fm_multiply 37 9.5 0.003 0.003 6.646 6.655 rs_pw_transfer 1038 11.9 0.017 0.019 5.816 6.465 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.789 5.872 fft_wrap_pw1pw2_140 519 13.2 0.444 0.508 5.587 5.760 fft3d_ps 1281 14.7 2.078 2.560 5.413 5.629 mp_alltoall_d11v 2401 14.1 4.255 5.332 4.255 5.332 grid_collocate_task_list 127 9.7 4.690 5.025 4.690 5.025 cp_fm_cholesky_decompose 22 10.9 4.938 4.951 4.938 4.951 potential_pw2rs 127 12.3 0.009 0.011 4.508 4.530 mp_sum_d 4441 12.1 3.502 4.088 3.502 4.088 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=202.682000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=596.090909, yerr=6.302302 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/eb0d6eed04221145363480ad2019c815711cc597_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+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 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 827.777024E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100942E+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 1463834332048 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4077 57314. MP_Allreduce 11232 947. MP_Sync 170 MP_Alltoall 1969 4951717. MP_ISendRecv 24064 47072. MP_Wait 37948 MP_comm_split 83 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 0.043 0.048 193.166 193.167 qs_mol_dyn_low 1 2.0 0.003 0.003 192.676 192.689 qs_forces 11 3.9 0.003 0.003 192.594 192.595 qs_energies 11 4.9 0.001 0.001 185.842 185.852 scf_env_do_scf 11 5.9 0.001 0.001 169.115 169.125 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 135.390 135.393 velocity_verlet 10 3.0 0.001 0.001 120.674 120.676 dbcsr_multiply_generic 2507 12.6 0.184 0.191 99.075 100.446 qs_scf_new_mos 117 7.6 0.001 0.001 96.698 97.269 qs_scf_loop_do_ot 117 8.6 0.001 0.001 96.697 97.269 ot_scf_mini 117 9.6 0.004 0.004 91.871 92.550 multiply_cannon 2507 13.6 0.479 0.530 78.226 82.856 multiply_cannon_loop 2507 14.6 1.245 1.279 74.666 77.190 ot_mini 117 10.6 0.001 0.001 51.075 51.674 mp_waitall_1 226760 16.4 25.570 38.641 25.570 38.641 multiply_cannon_multrec 30084 15.6 21.702 26.184 31.668 36.665 init_scf_loop 11 6.9 0.000 0.000 33.636 33.637 rebuild_ks_matrix 128 8.3 0.001 0.001 32.805 33.394 qs_ks_build_kohn_sham_matrix 128 9.3 0.018 0.025 32.804 33.393 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.562 30.095 qs_ot_get_derivative 117 11.6 0.001 0.002 28.990 29.653 prepare_preconditioner 11 7.9 0.000 0.000 29.283 29.366 make_preconditioner 11 8.9 0.000 0.000 29.283 29.366 multiply_cannon_metrocomm3 30084 15.6 0.092 0.097 16.279 28.919 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.975 28.498 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 22.555 23.869 apply_single 128 13.6 0.001 0.001 22.554 23.869 qs_ot_get_p 128 10.4 0.001 0.001 22.505 23.297 ot_diis_step 117 11.6 0.014 0.015 21.894 21.896 multiply_cannon_sync_h2d 30084 15.6 19.062 21.415 19.062 21.415 qs_ot_p2m_diag 83 11.4 0.187 0.216 17.606 17.651 cp_fm_cholesky_invert 11 10.9 17.050 17.062 17.050 17.062 cp_dbcsr_syevd 83 12.4 0.005 0.006 16.401 16.403 make_m2s 5014 13.6 0.091 0.098 14.409 16.138 make_images 5014 14.6 1.165 1.356 14.206 15.934 sum_up_and_integrate 128 10.3 0.116 0.135 14.410 14.438 integrate_v_rspace 128 11.3 0.003 0.004 14.294 14.324 cp_fm_diag_elpa 83 13.4 0.000 0.001 13.215 13.224 cp_fm_redistribute_end 83 14.4 7.685 13.152 7.698 13.153 cp_fm_diag_elpa_base 83 14.4 5.214 12.682 5.433 13.049 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.952 12.985 calculate_rho_elec 128 8.7 0.088 0.105 12.951 12.985 init_scf_run 11 5.9 0.000 0.002 11.966 11.967 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.966 11.967 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 11.201 11.693 multiply_cannon_metrocomm4 27577 15.6 0.098 0.113 3.847 10.822 make_images_data 5014 15.6 0.066 0.072 8.691 10.697 mp_irecv_dv 69486 16.3 3.651 10.428 3.651 10.428 dbcsr_mm_accdrv_process 62242 16.2 4.661 5.387 9.434 10.031 hybrid_alltoall_any 5200 16.5 0.344 1.511 7.350 10.020 wfi_extrapolate 11 7.9 0.001 0.001 8.438 8.438 pw_transfer 1547 11.6 0.085 0.100 7.810 7.883 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.493 7.676 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.585 7.656 grid_integrate_task_list 128 12.3 7.175 7.541 7.175 7.541 cp_fm_cholesky_decompose 22 10.9 7.327 7.401 7.327 7.401 density_rs2pw 128 9.7 0.006 0.006 6.780 7.190 mp_sum_l 7870 13.0 4.663 6.889 4.663 6.889 calculate_dm_sparse 128 9.5 0.001 0.001 6.666 6.812 fft_wrap_pw1pw2_140 523 13.2 0.470 0.520 6.560 6.628 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.420 6.429 rs_pw_transfer 1046 11.9 0.014 0.016 5.632 6.109 fft3d_ps 1291 14.7 2.774 2.919 5.953 5.999 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.480 5.657 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.416 5.492 grid_collocate_task_list 128 9.7 4.907 5.394 4.907 5.394 mp_allgather_i34 2507 14.6 2.133 5.207 2.133 5.207 potential_pw2rs 128 12.3 0.015 0.017 4.865 4.893 mp_alltoall_d11v 2415 14.1 4.127 4.765 4.127 4.765 dbcsr_complete_redistribute 395 12.7 0.778 0.866 3.275 4.140 mp_sum_d 4456 12.2 2.687 4.087 2.687 4.087 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=193.167000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=789.454545, yerr=1.157084 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/eb0d6eed04221145363480ad2019c815711cc597_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420243808256 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528908111872 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514772E+12 0.0% 0.0% 100.0% flops max/rank 2.928533E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755942624 0.0% 0.0% 100.0% number of processed stacks 3984192 0.0% 0.0% 100.0% average stack size 0.0 0.0 1695.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 943.362048E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1042912 MPI messages size (bytes): total size 2.716210E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.604448E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 281856 36943429632 131072 < size <= 4194304 660064 996105256960 4194304 < size <= 16777216 65632 931531401248 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4103 56951. MP_Allreduce 11296 984. MP_Sync 170 MP_Alltoall 1712 9388896. MP_ISendRecv 15872 75008. MP_Wait 29756 MP_comm_split 83 MP_ISend 11748 275205. MP_IRecv 11748 275205. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.038 0.053 179.868 179.869 qs_mol_dyn_low 1 2.0 0.003 0.003 179.216 179.388 qs_forces 11 3.9 0.003 0.003 179.116 179.121 qs_energies 11 4.9 0.001 0.003 172.637 172.645 scf_env_do_scf 11 5.9 0.001 0.001 156.861 156.863 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 120.699 120.701 velocity_verlet 10 3.0 0.001 0.001 114.713 114.715 qs_scf_new_mos 117 7.6 0.001 0.001 83.793 84.106 qs_scf_loop_do_ot 117 8.6 0.001 0.001 83.792 84.105 dbcsr_multiply_generic 2507 12.6 0.179 0.184 82.308 83.296 ot_scf_mini 117 9.6 0.004 0.004 79.633 79.988 multiply_cannon 2507 13.6 0.500 0.521 62.053 66.025 multiply_cannon_loop 2507 14.6 0.854 0.883 58.922 61.619 ot_mini 117 10.6 0.001 0.001 43.104 43.455 init_scf_loop 11 6.9 0.001 0.003 36.043 36.044 mp_waitall_1 178456 16.5 25.855 35.945 25.855 35.945 prepare_preconditioner 11 7.9 0.000 0.000 32.017 32.064 make_preconditioner 11 8.9 0.000 0.001 32.016 32.064 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.581 31.024 rebuild_ks_matrix 128 8.3 0.001 0.001 30.231 30.695 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.017 30.231 30.694 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.321 27.748 multiply_cannon_multrec 20056 15.6 13.488 16.373 22.290 25.183 multiply_cannon_metrocomm3 20056 15.6 0.060 0.063 15.370 25.141 qs_ot_get_derivative 117 11.6 0.001 0.002 23.308 23.652 qs_ot_get_p 128 10.4 0.001 0.001 21.772 22.207 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.844 20.900 apply_single 128 13.6 0.001 0.001 19.844 20.900 ot_diis_step 117 11.6 0.018 0.022 19.685 19.686 qs_ot_p2m_diag 83 11.4 0.265 0.272 17.458 17.470 cp_dbcsr_syevd 83 12.4 0.005 0.005 16.376 16.377 multiply_cannon_sync_h2d 20056 15.6 14.223 15.972 14.223 15.972 make_m2s 5014 13.6 0.083 0.088 14.954 15.802 make_images 5014 14.6 1.189 1.308 14.723 15.570 cp_fm_cholesky_invert 11 10.9 15.064 15.073 15.064 15.073 sum_up_and_integrate 128 10.3 0.133 0.146 14.348 14.376 integrate_v_rspace 128 11.3 0.003 0.004 14.215 14.244 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.326 13.362 calculate_rho_elec 128 8.7 0.132 0.147 13.325 13.361 cp_fm_diag_elpa 83 13.4 0.001 0.001 13.163 13.165 cp_fm_redistribute_end 83 14.4 4.947 13.095 4.964 13.097 cp_fm_diag_elpa_base 83 14.4 7.660 12.470 8.106 12.988 init_scf_run 11 5.9 0.000 0.001 10.733 10.733 scf_env_initial_rho_setup 11 6.9 0.001 0.002 10.733 10.733 make_images_data 5014 15.6 0.061 0.067 9.164 10.362 hybrid_alltoall_any 5200 16.5 0.432 1.979 7.840 9.599 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.226 9.487 multiply_cannon_metrocomm4 17549 15.6 0.060 0.071 3.448 9.400 mp_irecv_dv 50230 16.2 3.329 9.155 3.329 9.155 dbcsr_mm_accdrv_process 41502 16.2 4.508 5.418 8.262 8.410 pw_transfer 1547 11.6 0.085 0.103 7.917 8.018 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.694 7.801 grid_integrate_task_list 128 12.3 7.299 7.765 7.299 7.765 cp_fm_upper_to_full 105 14.5 5.881 7.706 5.881 7.706 cp_fm_cholesky_decompose 22 10.9 7.594 7.604 7.594 7.604 wfi_extrapolate 11 7.9 0.001 0.001 7.439 7.439 density_rs2pw 128 9.7 0.006 0.006 6.777 7.149 fft_wrap_pw1pw2_140 523 13.2 0.478 0.531 6.664 6.775 dbcsr_complete_redistribute 395 12.7 1.175 1.205 4.761 6.573 fft3d_ps 1291 14.7 2.692 2.888 5.961 6.019 calculate_dm_sparse 128 9.5 0.001 0.001 5.880 5.973 rs_pw_transfer 1046 11.9 0.013 0.014 5.382 5.693 grid_collocate_task_list 128 9.7 5.071 5.555 5.071 5.555 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.521 5.529 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.681 5.351 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.532 5.349 mp_alltoall_d11v 2415 14.1 4.448 4.900 4.448 4.900 mp_sum_l 7870 13.0 3.488 4.823 3.488 4.823 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.604 4.723 potential_pw2rs 128 12.3 0.020 0.022 4.703 4.718 mp_allgather_i34 2507 14.6 1.656 4.643 1.656 4.643 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.416 4.215 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.152 4.191 mp_alltoall_i22 716 14.1 2.002 3.948 2.002 3.948 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.735 3.735 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=179.869000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=896.454545, yerr=11.211712 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/eb0d6eed04221145363480ad2019c815711cc597_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 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 11444702699520 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 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 4.320339E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 5927808 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.2 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.148957E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1133160 MPI messages size (bytes): total size 2.008142E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.772161E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 315952 35695099904 131072 < size <= 4194304 709496 778939400192 4194304 < size <= 16777216 69840 660837692480 16777216 < size 30480 532676608000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4085 57191. MP_Allreduce 11251 1068. MP_Sync 168 MP_Alltoall 1700 12496376. MP_ISendRecv 11684 75008. MP_Wait 28114 MP_comm_split 82 MP_ISend 14840 244848. MP_IRecv 14840 244848. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.052 0.101 189.747 189.748 qs_mol_dyn_low 1 2.0 0.003 0.005 189.215 189.228 qs_forces 11 3.9 0.003 0.004 189.098 189.107 qs_energies 11 4.9 0.002 0.005 182.028 182.040 scf_env_do_scf 11 5.9 0.001 0.001 164.786 164.800 velocity_verlet 10 3.0 0.001 0.001 124.775 124.778 scf_env_do_scf_inner_loop 116 6.6 0.003 0.009 117.515 117.517 qs_scf_new_mos 116 7.6 0.001 0.001 81.866 82.193 qs_scf_loop_do_ot 116 8.6 0.001 0.002 81.865 82.192 dbcsr_multiply_generic 2485 12.5 0.186 0.191 79.869 80.699 ot_scf_mini 116 9.6 0.003 0.004 77.277 77.573 multiply_cannon 2485 13.5 0.556 0.585 54.737 57.422 multiply_cannon_loop 2485 14.5 1.172 1.198 50.708 52.751 init_scf_loop 11 6.9 0.001 0.002 47.144 47.147 prepare_preconditioner 11 7.9 0.000 0.000 43.045 43.064 make_preconditioner 11 8.9 0.000 0.001 43.045 43.064 ot_mini 116 10.6 0.001 0.001 42.717 43.006 make_full_inverse_cholesky 11 9.9 0.000 0.000 36.704 41.622 multiply_cannon_multrec 29820 15.5 13.958 19.205 25.845 31.412 rebuild_ks_matrix 127 8.3 0.001 0.001 29.142 29.449 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.018 29.142 29.448 mp_waitall_1 152434 16.5 18.172 27.693 18.172 27.693 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.298 26.572 qs_ot_get_derivative 116 11.6 0.001 0.002 23.086 23.376 make_m2s 4970 13.5 0.096 0.099 20.670 21.701 make_images 4970 14.5 1.962 2.322 20.368 21.402 qs_ot_get_p 127 10.4 0.001 0.001 20.770 21.082 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 19.111 19.764 apply_single 127 13.6 0.001 0.001 19.111 19.764 ot_diis_step 116 11.6 0.017 0.018 19.514 19.515 cp_fm_cholesky_invert 11 10.9 16.893 16.903 16.893 16.903 qs_ot_p2m_diag 82 11.4 0.338 0.385 16.597 16.649 cp_fm_upper_to_full 104 14.7 10.948 16.183 10.948 16.183 multiply_cannon_metrocomm3 29820 15.5 0.046 0.049 6.670 15.582 cp_dbcsr_syevd 82 12.4 0.005 0.005 15.308 15.309 sum_up_and_integrate 127 10.3 0.140 0.153 14.143 14.168 integrate_v_rspace 127 11.3 0.003 0.004 14.002 14.032 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.214 13.249 calculate_rho_elec 127 8.7 0.174 0.189 13.213 13.249 make_images_data 4970 15.5 0.064 0.068 11.232 13.129 dbcsr_complete_redistribute 393 12.7 1.567 1.711 9.140 12.821 multiply_cannon_sync_h2d 29820 15.5 11.541 12.674 11.541 12.674 cp_fm_diag_elpa 82 13.4 0.000 0.001 12.138 12.141 cp_fm_redistribute_end 82 14.4 2.073 12.049 2.090 12.054 dbcsr_mm_accdrv_process 61748 16.2 7.257 8.328 11.461 12.014 cp_fm_diag_elpa_base 82 14.4 9.349 11.428 9.924 11.917 hybrid_alltoall_any 5155 16.4 0.523 2.196 10.012 11.517 copy_fm_to_dbcsr 208 11.6 0.002 0.002 7.681 11.381 init_scf_run 11 5.9 0.000 0.001 11.153 11.155 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.153 11.154 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.318 9.921 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.593 9.796 mp_alltoall_i22 712 14.1 5.512 9.136 5.512 9.136 pw_transfer 1535 11.6 0.084 0.097 7.988 8.063 cp_fm_cholesky_decompose 22 10.9 7.762 7.875 7.762 7.875 fft_wrap_pw1pw2 1281 12.7 0.010 0.010 7.765 7.845 grid_integrate_task_list 127 12.3 7.450 7.744 7.450 7.744 wfi_extrapolate 11 7.9 0.001 0.001 7.715 7.715 multiply_cannon_metrocomm4 24850 15.5 0.075 0.086 2.753 7.195 mp_irecv_dv 75445 16.2 2.611 6.932 2.611 6.932 fft_wrap_pw1pw2_140 519 13.2 0.479 0.484 6.816 6.913 density_rs2pw 127 9.7 0.005 0.006 6.330 6.654 calculate_dm_sparse 127 9.5 0.001 0.001 6.269 6.347 fft3d_ps 1281 14.7 2.788 2.855 6.003 6.066 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.637 5.729 mp_alltoall_d11v 2401 14.1 5.124 5.597 5.124 5.597 grid_collocate_task_list 127 9.7 5.173 5.470 5.173 5.470 rs_pw_transfer 1038 11.9 0.013 0.013 4.787 5.130 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 4.555 4.556 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.432 4.523 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.440 4.504 potential_pw2rs 127 12.3 0.022 0.023 4.383 4.394 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.203 4.255 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=189.748000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1085.818182, yerr=17.652289 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/eb0d6eed04221145363480ad2019c815711cc597_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 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 11444702699520 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 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 5.820059E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 1944496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3448.5 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.508512E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 238560 MPI messages size (bytes): total size 1.321104E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.537828E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 112800 59139686400 4194304 < size <= 16777216 104112 545846722560 16777216 < size 20064 716108580288 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8852 52. MP_Alltoall 9584 804353. MP_ISend 39716 2104723. MP_IRecv 39716 2103824. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4003 58193. MP_Allreduce 11005 1175. MP_Sync 86 MP_Alltoall 1700 18828189. MP_ISendRecv 7620 122880. MP_Wait 19810 MP_ISend 10600 423612. MP_IRecv 10600 423612. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.044 173.407 173.407 qs_mol_dyn_low 1 2.0 0.003 0.003 173.018 173.030 qs_forces 11 3.9 0.003 0.003 172.913 172.918 qs_energies 11 4.9 0.001 0.001 165.521 165.526 scf_env_do_scf 11 5.9 0.001 0.001 148.151 148.164 velocity_verlet 10 3.0 0.001 0.001 113.306 113.310 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 110.829 110.830 qs_scf_new_mos 116 7.6 0.001 0.001 75.348 75.416 qs_scf_loop_do_ot 116 8.6 0.001 0.001 75.347 75.416 dbcsr_multiply_generic 2485 12.5 0.180 0.185 74.127 74.517 ot_scf_mini 116 9.6 0.003 0.004 70.881 70.916 multiply_cannon 2485 13.5 0.583 0.613 54.157 58.526 multiply_cannon_loop 2485 14.5 0.443 0.454 49.380 50.013 ot_mini 116 10.6 0.001 0.001 39.595 39.625 init_scf_loop 11 6.9 0.000 0.000 37.163 37.164 prepare_preconditioner 11 7.9 0.000 0.000 33.278 33.302 make_preconditioner 11 8.9 0.000 0.000 33.278 33.302 mp_waitall_1 128490 16.6 25.909 33.050 25.909 33.050 make_full_inverse_cholesky 11 9.9 0.000 0.000 31.062 31.323 rebuild_ks_matrix 127 8.3 0.001 0.001 28.519 28.590 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.018 28.518 28.590 qs_ks_update_qs_env 127 7.6 0.001 0.001 25.970 26.035 multiply_cannon_multrec 9940 15.5 10.388 14.216 17.766 21.273 qs_ot_get_derivative 116 11.6 0.001 0.002 19.845 19.880 ot_diis_step 116 11.6 0.019 0.020 19.678 19.679 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.334 19.575 apply_single 127 13.6 0.001 0.001 19.333 19.574 cp_fm_cholesky_invert 11 10.9 19.136 19.141 19.136 19.141 multiply_cannon_metrocomm3 9940 15.5 0.023 0.024 12.177 18.842 make_m2s 4970 13.5 0.067 0.071 16.285 18.553 qs_ot_get_p 127 10.4 0.001 0.001 18.406 18.439 make_images 4970 14.5 2.352 2.837 15.983 18.250 qs_ot_p2m_diag 82 11.4 0.489 0.495 14.668 14.683 sum_up_and_integrate 127 10.3 0.180 0.190 14.166 14.211 integrate_v_rspace 127 11.3 0.004 0.004 13.985 14.039 cp_dbcsr_syevd 82 12.4 0.005 0.005 13.489 13.491 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.168 13.194 calculate_rho_elec 127 8.7 0.257 0.267 13.167 13.193 make_images_data 4970 15.5 0.054 0.060 9.768 12.259 multiply_cannon_sync_h2d 9940 15.5 11.519 12.151 11.519 12.151 hybrid_alltoall_any 5155 16.4 0.805 3.766 9.513 11.721 init_scf_run 11 5.9 0.000 0.001 10.519 10.520 scf_env_initial_rho_setup 11 6.9 0.001 0.002 10.519 10.520 cp_fm_diag_elpa 82 13.4 0.000 0.000 10.443 10.444 cp_fm_diag_elpa_base 82 14.4 10.183 10.262 10.429 10.429 cp_fm_cholesky_decompose 22 10.9 8.368 8.488 8.368 8.488 qs_ot_get_derivative_diag 76 12.4 0.002 0.003 8.018 8.050 grid_integrate_task_list 127 12.3 7.719 7.963 7.719 7.963 pw_transfer 1535 11.6 0.083 0.090 7.890 7.903 fft_wrap_pw1pw2 1281 12.7 0.010 0.010 7.671 7.690 dbcsr_mm_accdrv_process 20590 16.1 2.612 3.421 7.010 7.652 wfi_extrapolate 11 7.9 0.001 0.001 7.443 7.443 mp_allgather_i34 2485 14.5 2.993 7.136 2.993 7.136 multiply_cannon_metrocomm1 9940 15.5 0.027 0.028 4.302 7.003 fft_wrap_pw1pw2_140 519 13.2 0.497 0.516 6.693 6.721 density_rs2pw 127 9.7 0.005 0.006 6.050 6.397 calculate_dm_sparse 127 9.5 0.001 0.001 6.144 6.198 fft3d_ps 1281 14.7 2.700 2.762 5.845 5.883 dbcsr_complete_redistribute 393 12.7 2.164 2.250 5.390 5.801 grid_collocate_task_list 127 9.7 5.445 5.672 5.445 5.672 mp_alltoall_d11v 2401 14.1 4.782 5.458 4.782 5.458 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.305 5.321 qs_energies_init_hamiltonians 11 5.9 0.003 0.005 5.280 5.281 rs_pw_transfer 1038 11.9 0.012 0.013 4.291 4.647 multiply_cannon_metrocomm4 7455 15.5 0.023 0.025 1.840 4.515 mp_irecv_dv 28618 15.9 1.806 4.442 1.806 4.442 potential_pw2rs 127 12.3 0.026 0.027 4.265 4.283 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.194 4.222 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.558 3.877 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.556 3.863 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.751 3.771 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.578 3.674 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.517 3.536 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=173.407000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1388.545455, yerr=62.260981 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/eb0d6eed04221145363480ad2019c815711cc597_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410025439232 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 1957874282496 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963547688960 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 11444716634112 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019200700416 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019200700416 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 92.796621E+12 0.0% 0.0% 100.0% flops max/rank 11.606418E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705505504 0.0% 0.0% 100.0% number of processed stacks 1947808 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.6 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 3.092115E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 99400 MPI messages size (bytes): total size 1.127422E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.342271E+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 44768 34745614336 4194304 < size <= 16777216 43984 376564613120 16777216 < size 10032 716109038208 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 58563. MP_Allreduce 11105 1512. MP_Sync 86 MP_Alltoall 1700 36954362. MP_ISendRecv 3556 218624. MP_Wait 11506 MP_ISend 6360 1080477. MP_IRecv 6360 1080477. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.072 0.126 290.437 290.438 qs_mol_dyn_low 1 2.0 0.003 0.003 289.765 289.777 qs_forces 11 3.9 0.003 0.003 289.671 289.674 qs_energies 11 4.9 0.001 0.001 281.117 281.124 scf_env_do_scf 11 5.9 0.001 0.001 259.126 259.141 velocity_verlet 10 3.0 0.001 0.001 208.849 208.856 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 134.146 134.148 init_scf_loop 11 6.9 0.000 0.000 124.712 124.714 prepare_preconditioner 11 7.9 0.000 0.000 120.031 120.054 make_preconditioner 11 8.9 0.000 0.000 120.031 120.054 make_full_inverse_cholesky 11 9.9 0.000 0.000 96.422 117.160 qs_scf_new_mos 116 7.6 0.001 0.001 91.812 91.928 qs_scf_loop_do_ot 116 8.6 0.001 0.001 91.811 91.927 ot_scf_mini 116 9.6 0.004 0.004 87.024 87.097 dbcsr_multiply_generic 2485 12.5 0.222 0.237 82.752 83.236 cp_fm_upper_to_full 104 14.8 52.532 75.094 52.532 75.094 multiply_cannon 2485 13.5 0.711 0.734 57.974 58.425 multiply_cannon_loop 2485 14.5 0.470 0.502 54.427 55.668 ot_mini 116 10.6 0.001 0.001 44.230 44.289 dbcsr_complete_redistribute 393 12.7 4.007 4.046 29.422 42.067 copy_fm_to_dbcsr 208 11.6 0.001 0.002 25.936 38.483 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 23.564 35.963 cp_fm_cholesky_invert 11 10.9 34.367 34.373 34.367 34.373 mp_alltoall_i22 712 14.1 21.370 33.800 21.370 33.800 rebuild_ks_matrix 127 8.3 0.001 0.001 33.734 33.768 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.017 33.733 33.767 mp_waitall_1 104546 16.7 28.182 32.089 28.182 32.089 qs_ks_update_qs_env 127 7.6 0.001 0.001 31.461 31.496 qs_ot_get_p 127 10.4 0.001 0.001 27.372 27.409 qs_ot_get_derivative 116 11.6 0.002 0.002 24.310 24.385 qs_ot_p2m_diag 82 11.4 0.868 0.873 23.186 23.214 cp_dbcsr_syevd 82 12.4 0.005 0.006 21.480 21.481 make_m2s 4970 13.5 0.076 0.078 19.954 20.777 make_images 4970 14.5 3.741 3.890 19.482 20.310 ot_diis_step 116 11.6 0.022 0.023 19.876 19.877 multiply_cannon_metrocomm3 9940 15.5 0.023 0.024 18.559 19.613 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.132 19.296 apply_single 127 13.6 0.001 0.001 19.132 19.296 cp_fm_diag_elpa 82 13.4 0.000 0.000 18.235 18.236 cp_fm_diag_elpa_base 82 14.4 13.846 15.444 18.229 18.229 multiply_cannon_multrec 9940 15.5 10.413 12.128 17.792 17.931 sum_up_and_integrate 127 10.3 0.319 0.321 15.918 16.017 integrate_v_rspace 127 11.3 0.004 0.004 15.598 15.697 multiply_cannon_sync_h2d 9940 15.5 15.533 15.543 15.533 15.543 qs_rho_update_rho_low 127 7.7 0.001 0.001 15.174 15.204 calculate_rho_elec 127 8.7 0.479 0.480 15.173 15.204 hybrid_alltoall_any 5155 16.4 1.293 3.027 10.835 12.652 make_images_data 4970 15.5 0.059 0.064 10.640 12.402 init_scf_run 11 5.9 0.000 0.001 12.015 12.015 scf_env_initial_rho_setup 11 6.9 0.011 0.037 12.014 12.015 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.409 9.478 cp_fm_cholesky_decompose 22 10.9 9.283 9.303 9.283 9.303 dbcsr_mm_accdrv_process 20590 16.0 3.737 5.506 7.145 8.953 wfi_extrapolate 11 7.9 0.001 0.001 8.876 8.876 pw_transfer 1535 11.6 0.090 0.093 8.675 8.685 grid_integrate_task_list 127 12.3 8.476 8.662 8.476 8.662 fft_wrap_pw1pw2 1281 12.7 0.011 0.011 8.442 8.451 qs_energies_init_hamiltonians 11 5.9 0.053 0.058 7.969 7.970 fft_wrap_pw1pw2_140 519 13.2 0.537 0.543 7.433 7.447 mp_alltoall_d11v 2401 14.1 7.286 7.402 7.286 7.402 calculate_dm_sparse 127 9.5 0.001 0.001 6.639 6.721 fft3d_ps 1281 14.7 2.742 2.769 6.496 6.517 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.444 6.490 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.221 6.327 grid_collocate_task_list 127 9.7 6.285 6.321 6.285 6.321 density_rs2pw 127 9.7 0.005 0.005 6.227 6.264 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=290.438000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2718.090909, yerr=196.120767 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/eb0d6eed04221145363480ad2019c815711cc597_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 1.261654E+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_ISendRecv 570 19200. MP_Wait 1302 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 0.054 0.078 85.084 85.085 qs_energies 1 2.0 0.000 0.000 84.611 84.618 ls_scf 1 3.0 0.000 0.000 83.706 83.712 dbcsr_multiply_generic 111 6.7 0.016 0.017 72.728 72.928 multiply_cannon 111 7.7 0.018 0.020 55.996 57.098 multiply_cannon_loop 111 8.7 0.215 0.232 52.576 53.873 ls_scf_main 1 4.0 0.000 0.000 52.389 52.395 density_matrix_trs4 2 5.0 0.002 0.003 46.932 47.016 ls_scf_init_scf 1 4.0 0.000 0.000 28.282 28.283 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.210 27.261 mp_waitall_1 11316 10.9 22.282 25.348 22.282 25.348 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.078 25.103 multiply_cannon_multrec 2664 9.7 8.214 8.905 15.661 17.307 multiply_cannon_sync_h2d 2664 9.7 13.726 15.251 13.726 15.251 make_m2s 222 7.7 0.009 0.011 13.131 13.658 make_images 222 8.7 0.101 0.111 13.109 13.638 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.553 12.445 make_images_data 222 9.7 0.005 0.006 7.636 8.166 dbcsr_mm_accdrv_process 4760 10.4 0.510 0.610 7.063 8.110 hybrid_alltoall_any 227 10.6 0.219 1.850 6.530 7.868 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.414 7.794 dbcsr_mm_accdrv_process_sort 4760 11.4 6.355 7.320 6.355 7.320 calculate_norms 4752 9.8 5.537 6.216 5.537 6.216 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.070 5.216 mp_sum_l 807 5.4 3.070 4.682 3.070 4.682 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.273 3.496 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.063 3.450 mp_irecv_dv 6231 10.9 2.045 3.418 2.045 3.418 arnoldi_extremal 4 6.8 0.000 0.000 3.342 3.373 arnoldi_normal_ev 4 7.8 0.001 0.003 3.342 3.373 make_images_sizes 222 9.7 0.000 0.000 0.758 3.321 mp_alltoall_i44 222 10.7 0.757 3.320 0.757 3.320 build_subspace 16 8.4 0.009 0.012 3.244 3.246 ls_scf_post 1 4.0 0.000 0.000 3.034 3.042 ls_scf_store_result 1 5.0 0.000 0.000 2.865 2.894 dbcsr_special_finalize 555 9.7 0.005 0.007 2.373 2.865 dbcsr_merge_single_wm 555 10.7 0.452 0.588 2.365 2.856 make_images_pack 222 9.7 2.214 2.650 2.216 2.652 dbcsr_matrix_vector_mult 304 9.0 0.004 0.010 2.332 2.611 dbcsr_sort_data 658 11.4 2.170 2.593 2.170 2.593 dbcsr_matrix_vector_mult_local 304 10.0 2.069 2.496 2.071 2.498 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.197 2.250 buffer_matrices_ensure_size 222 8.7 1.747 2.035 1.747 2.035 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.735 1.743 rebuild_ks_matrix 3 7.3 0.000 0.000 1.726 1.734 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.726 1.734 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.085000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1142.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/eb0d6eed04221145363480ad2019c815711cc597_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 2.106851E+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_ISendRecv 282 57600. MP_Wait 828 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 0.052 0.084 91.229 91.231 qs_energies 1 2.0 0.000 0.000 90.614 90.631 ls_scf 1 3.0 0.000 0.001 89.295 89.311 dbcsr_multiply_generic 111 6.7 0.016 0.017 75.377 75.711 multiply_cannon 111 7.7 0.029 0.041 53.605 57.211 ls_scf_main 1 4.0 0.000 0.007 55.030 55.044 multiply_cannon_loop 111 8.7 0.116 0.123 50.306 53.285 density_matrix_trs4 2 5.0 0.002 0.003 49.338 49.538 ls_scf_init_scf 1 4.0 0.001 0.005 30.725 30.727 mp_waitall_1 9246 10.9 21.260 30.054 21.260 30.054 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.523 29.631 multiply_cannon_multrec 1332 9.7 13.143 16.696 22.517 27.252 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.117 27.129 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.972 20.751 make_m2s 222 7.7 0.008 0.008 15.136 15.730 make_images 222 8.7 1.595 1.937 15.105 15.701 dbcsr_mm_accdrv_process 4041 10.4 0.291 0.452 8.972 10.577 dbcsr_mm_accdrv_process_sort 4041 11.4 8.566 10.129 8.566 10.129 make_images_data 222 9.7 0.004 0.005 8.658 9.544 hybrid_alltoall_any 227 10.6 0.524 2.430 8.070 8.930 mp_sum_l 807 5.4 5.366 8.145 5.366 8.145 multiply_cannon_metrocomm4 1221 9.7 0.007 0.008 3.239 7.713 mp_irecv_dv 3311 11.0 3.219 7.662 3.219 7.662 calculate_norms 2376 9.8 5.986 6.744 5.986 6.744 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.101 6.327 multiply_cannon_sync_h2d 1332 9.7 4.743 5.912 4.743 5.912 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.038 5.223 arnoldi_extremal 4 6.8 0.000 0.000 4.634 4.652 arnoldi_normal_ev 4 7.8 0.001 0.004 4.634 4.652 build_subspace 16 8.4 0.014 0.021 4.382 4.385 ls_scf_post 1 4.0 0.005 0.023 3.540 3.557 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.139 3.374 ls_scf_store_result 1 5.0 0.000 0.000 3.252 3.360 dbcsr_matrix_vector_mult_local 304 10.0 2.738 3.225 2.740 3.227 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.254 2.721 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.556 2.672 make_images_pack 222 9.7 2.030 2.440 2.033 2.442 mp_allgather_i34 111 8.7 0.984 2.367 0.984 2.367 dbcsr_sort_data 436 11.2 1.782 1.977 1.782 1.977 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.891 1.894 rebuild_ks_matrix 3 7.3 0.000 0.000 1.877 1.879 qs_ks_build_kohn_sham_matrix 3 8.3 0.005 0.014 1.877 1.879 dbcsr_data_new 4174 10.1 1.623 1.854 1.623 1.854 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.231000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1727.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/eb0d6eed04221145363480ad2019c815711cc597_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 2.700882E+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_ISendRecv 186 57600. MP_Wait 732 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 0.063 0.152 94.015 94.016 qs_energies 1 2.0 0.000 0.000 93.371 93.377 ls_scf 1 3.0 0.000 0.000 91.986 91.993 dbcsr_multiply_generic 111 6.7 0.016 0.017 76.717 76.948 multiply_cannon 111 7.7 0.048 0.158 53.207 57.964 ls_scf_main 1 4.0 0.000 0.000 57.575 57.581 multiply_cannon_loop 111 8.7 0.101 0.117 49.582 52.641 density_matrix_trs4 2 5.0 0.002 0.003 51.654 51.843 mp_waitall_1 7374 11.0 23.772 33.395 23.772 33.395 ls_scf_init_scf 1 4.0 0.000 0.000 30.897 30.899 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.749 29.837 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.327 27.339 multiply_cannon_multrec 888 9.7 12.697 14.952 21.381 24.657 multiply_cannon_metrocomm3 888 9.7 0.004 0.005 11.182 22.423 make_m2s 222 7.7 0.007 0.008 16.890 18.175 make_images 222 8.7 1.979 2.288 16.852 18.135 make_images_data 222 9.7 0.004 0.005 9.616 10.653 hybrid_alltoall_any 227 10.6 0.622 2.823 9.248 10.634 dbcsr_mm_accdrv_process 3754 10.4 0.250 0.464 8.209 9.498 dbcsr_mm_accdrv_process_sort 3754 11.4 7.828 9.034 7.828 9.034 mp_sum_l 807 5.4 5.136 8.553 5.136 8.553 multiply_cannon_sync_h2d 888 9.7 6.083 7.425 6.083 7.425 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.440 6.868 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.934 6.857 mp_irecv_dv 2335 11.1 2.426 6.833 2.426 6.833 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.592 6.627 arnoldi_extremal 4 6.8 0.000 0.000 5.136 5.154 arnoldi_normal_ev 4 7.8 0.001 0.005 5.136 5.154 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.913 5.106 build_subspace 16 8.4 0.014 0.020 4.828 4.833 calculate_norms 1584 9.8 4.277 4.602 4.277 4.602 mp_allgather_i34 111 8.7 1.459 3.775 1.459 3.775 dbcsr_matrix_vector_mult 304 9.0 0.005 0.016 3.452 3.769 dbcsr_matrix_vector_mult_local 304 10.0 3.024 3.593 3.026 3.595 ls_scf_post 1 4.0 0.000 0.000 3.514 3.522 ls_scf_store_result 1 5.0 0.000 0.000 3.245 3.325 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.758 2.845 make_images_pack 222 9.7 1.807 2.129 1.810 2.132 dbcsr_sort_data 325 11.1 1.912 2.110 1.912 2.110 dbcsr_data_release 9322 10.9 1.306 1.981 1.306 1.981 make_images_sizes 222 9.7 0.000 0.000 0.912 1.952 mp_alltoall_i44 222 10.7 0.912 1.951 0.912 1.951 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=94.016000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2223.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/eb0d6eed04221145363480ad2019c815711cc597_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 3.376308E+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_ISendRecv 138 86400. MP_Wait 600 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.038 0.050 97.936 97.937 qs_energies 1 2.0 0.003 0.023 97.149 97.174 ls_scf 1 3.0 0.000 0.001 95.480 95.488 dbcsr_multiply_generic 111 6.7 0.017 0.022 79.010 79.328 ls_scf_main 1 4.0 0.000 0.002 59.105 59.107 multiply_cannon 111 7.7 0.054 0.129 51.952 56.625 density_matrix_trs4 2 5.0 0.004 0.029 53.003 53.114 multiply_cannon_loop 111 8.7 0.114 0.125 46.910 49.550 ls_scf_init_scf 1 4.0 0.000 0.002 33.096 33.098 ls_scf_init_matrix_S 1 5.0 0.000 0.001 31.780 31.836 mp_waitall_1 6438 11.0 23.014 30.641 23.014 30.641 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.290 29.303 multiply_cannon_multrec 1332 9.7 14.378 17.678 22.375 26.126 make_m2s 222 7.7 0.008 0.009 21.404 22.787 make_images 222 8.7 3.145 3.605 21.352 22.738 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.313 17.674 make_images_data 222 9.7 0.004 0.004 12.000 13.651 hybrid_alltoall_any 227 10.6 0.802 3.799 11.326 13.182 dbcsr_mm_accdrv_process 3641 10.4 0.218 0.415 7.649 9.172 dbcsr_mm_accdrv_process_sort 3641 11.4 7.275 8.767 7.275 8.767 mp_sum_l 807 5.4 4.074 7.222 4.074 7.222 multiply_cannon_sync_h2d 1332 9.7 5.452 6.178 5.452 6.178 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.081 6.081 mp_irecv_dv 3229 10.9 2.057 5.990 2.057 5.990 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.113 5.620 arnoldi_extremal 4 6.8 0.000 0.000 5.381 5.397 arnoldi_normal_ev 4 7.8 0.001 0.004 5.381 5.397 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.532 5.333 build_subspace 16 8.4 0.014 0.021 5.037 5.045 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.584 4.732 calculate_norms 2376 9.8 4.196 4.540 4.196 4.540 mp_allgather_i34 111 8.7 2.158 4.529 2.158 4.529 dbcsr_matrix_vector_mult 304 9.0 0.006 0.016 3.654 3.907 dbcsr_matrix_vector_mult_local 304 10.0 3.201 3.689 3.203 3.691 dbcsr_sort_data 658 11.4 3.074 3.441 3.074 3.441 ls_scf_post 1 4.0 0.001 0.003 3.278 3.285 dbcsr_special_finalize 555 9.7 0.006 0.006 2.823 3.216 dbcsr_merge_single_wm 555 10.7 0.533 0.660 2.815 3.208 ls_scf_store_result 1 5.0 0.000 0.000 3.021 3.078 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.950 3.021 dbcsr_data_release 10477 10.7 1.582 2.394 1.582 2.394 dbcsr_finalize 304 7.8 0.050 0.062 1.808 1.996 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.959 1.962 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.937000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2723.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/eb0d6eed04221145363480ad2019c815711cc597_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 4.629631E+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_ISendRecv 90 115200. MP_Wait 573 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.041 0.058 99.594 99.595 qs_energies 1 2.0 0.000 0.000 98.831 98.835 ls_scf 1 3.0 0.000 0.000 96.904 96.909 dbcsr_multiply_generic 111 6.7 0.018 0.018 78.339 78.551 ls_scf_main 1 4.0 0.000 0.000 62.431 62.432 multiply_cannon 111 7.7 0.081 0.158 55.719 61.079 density_matrix_trs4 2 5.0 0.002 0.003 55.280 55.377 multiply_cannon_loop 111 8.7 0.070 0.077 51.139 52.866 mp_waitall_1 5481 11.0 26.582 31.447 26.582 31.447 ls_scf_init_scf 1 4.0 0.000 0.000 30.860 30.865 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.631 29.662 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.425 27.438 multiply_cannon_multrec 444 9.7 14.103 16.581 21.252 23.063 make_m2s 222 7.7 0.006 0.006 17.713 20.295 make_images 222 8.7 3.727 4.438 17.650 20.234 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.294 16.692 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.181 14.294 make_images_data 222 9.7 0.003 0.004 9.905 12.425 hybrid_alltoall_any 227 10.6 0.791 3.773 9.625 12.157 dbcsr_mm_accdrv_process 3003 10.4 0.187 0.342 6.856 8.063 multiply_cannon_sync_h2d 444 9.7 6.552 7.954 6.552 7.954 dbcsr_mm_accdrv_process_sort 3003 11.4 6.541 7.716 6.541 7.716 mp_allgather_i34 111 8.7 2.798 7.054 2.798 7.054 arnoldi_extremal 4 6.8 0.000 0.000 6.050 6.063 arnoldi_normal_ev 4 7.8 0.002 0.005 6.050 6.063 build_subspace 16 8.4 0.015 0.020 5.643 5.651 mp_sum_l 807 5.4 2.933 4.999 2.933 4.999 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.736 4.868 dbcsr_matrix_vector_mult 304 9.0 0.007 0.016 4.246 4.428 dbcsr_matrix_vector_mult_local 304 10.0 3.732 4.203 3.734 4.205 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.679 3.962 mp_irecv_dv 1241 11.2 1.661 3.942 1.661 3.942 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.996 3.722 calculate_norms 792 9.8 3.538 3.700 3.538 3.700 ls_scf_post 1 4.0 0.000 0.000 3.613 3.617 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.439 3.536 make_images_sizes 222 9.7 0.000 0.000 1.141 3.447 mp_alltoall_i44 222 10.7 1.141 3.446 1.141 3.446 ls_scf_store_result 1 5.0 0.000 0.000 3.400 3.442 dbcsr_finalize 304 7.8 0.062 0.078 2.217 2.336 dbcsr_merge_all 275 8.9 0.474 0.533 2.064 2.166 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.030 2.031 rebuild_ks_matrix 3 7.3 0.000 0.000 1.997 1.998 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.997 1.998 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=99.595000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3630.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/eb0d6eed04221145363480ad2019c815711cc597_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 8.728011E+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_ISendRecv 84 732600. MP_Wait 309 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 0.085 0.162 108.715 108.715 qs_energies 1 2.0 0.000 0.000 107.189 107.201 ls_scf 1 3.0 0.000 0.000 104.290 104.302 dbcsr_multiply_generic 111 6.7 0.024 0.025 78.012 78.135 ls_scf_main 1 4.0 0.000 0.000 66.286 66.288 density_matrix_trs4 2 5.0 0.002 0.003 57.437 57.485 multiply_cannon 111 7.7 0.163 0.218 50.357 52.426 multiply_cannon_loop 111 8.7 0.068 0.070 46.829 47.756 ls_scf_init_scf 1 4.0 0.000 0.000 34.279 34.280 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.849 32.865 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 30.015 30.020 mp_waitall_1 4569 11.1 22.529 26.298 22.529 26.298 make_m2s 222 7.7 0.007 0.007 24.011 25.031 make_images 222 8.7 4.584 5.005 23.901 24.919 multiply_cannon_multrec 444 9.7 17.901 18.613 22.591 23.246 hybrid_alltoall_any 227 10.6 1.661 3.621 12.975 15.884 make_images_data 222 9.7 0.003 0.004 13.192 15.696 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.731 11.009 multiply_cannon_sync_h2d 444 9.7 8.841 8.890 8.841 8.890 arnoldi_extremal 4 6.8 0.000 0.000 7.684 7.691 arnoldi_normal_ev 4 7.8 0.003 0.009 7.684 7.691 build_subspace 16 8.4 0.026 0.036 7.144 7.156 dbcsr_matrix_vector_mult 304 9.0 0.009 0.025 5.604 5.772 dbcsr_matrix_vector_mult_local 304 10.0 5.119 5.432 5.122 5.435 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.034 5.280 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.792 4.884 dbcsr_mm_accdrv_process 1814 10.4 0.236 0.335 4.517 4.636 dbcsr_mm_accdrv_process_sort 1814 11.4 4.209 4.338 4.209 4.338 ls_scf_post 1 4.0 0.000 0.000 3.724 3.736 make_images_sizes 222 9.7 0.000 0.000 1.469 3.645 mp_alltoall_i44 222 10.7 1.469 3.645 1.469 3.645 mp_allgather_i34 111 8.7 1.145 3.578 1.145 3.578 ls_scf_store_result 1 5.0 0.000 0.000 3.427 3.438 calculate_norms 792 9.8 3.245 3.269 3.245 3.269 dbcsr_finalize 304 7.8 0.083 0.090 3.095 3.154 dbcsr_merge_all 275 8.9 0.892 0.916 2.877 2.929 dbcsr_complete_redistribute 5 7.6 1.449 1.481 2.805 2.908 qs_energies_init_hamiltonians 1 3.0 0.002 0.007 2.869 2.869 dbcsr_data_release 12724 10.6 2.346 2.857 2.346 2.857 matrix_ls_to_qs 2 6.0 0.000 0.000 2.425 2.541 dbcsr_sort_data 325 11.1 2.444 2.518 2.444 2.518 dbcsr_new_transposed 4 7.5 0.244 0.252 2.292 2.314 dbcsr_frobenius_norm 74 6.6 2.056 2.137 2.201 2.241 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.226 2.228 dbcsr_add_d 103 6.2 0.000 0.000 2.137 2.207 dbcsr_add_anytype 103 7.2 0.860 0.892 2.137 2.206 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=108.715000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6845.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: eb0d6eed04221145363480ad2019c815711cc597 Summary: empty Status: OK