=== 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: d60029441ac03bb22e0b17392546c4f4f2d728f9 ################# 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=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2022.05.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.0.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.0 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.3.2 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MPI_VERSION=3 DFLAGS += -D__MAX_CONTR=$(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/d60029441ac03bb22e0b17392546c4f4f2d728f9_performance_tests/01 job id: 42881795 --- 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/d60029441ac03bb22e0b17392546c4f4f2d728f9_performance_tests/02 job id: 42881797 --- 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/d60029441ac03bb22e0b17392546c4f4f2d728f9_performance_tests/03 job id: 42881801 --- 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/d60029441ac03bb22e0b17392546c4f4f2d728f9_performance_tests/04 job id: 42881803 --- 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/d60029441ac03bb22e0b17392546c4f4f2d728f9_performance_tests/05 job id: 42881808 --- 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/d60029441ac03bb22e0b17392546c4f4f2d728f9_performance_tests/06 job id: 42881810 --- 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/d60029441ac03bb22e0b17392546c4f4f2d728f9_performance_tests/07 job id: 42881814 --- 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/d60029441ac03bb22e0b17392546c4f4f2d728f9_performance_tests/08 job id: 42881816 --- 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/d60029441ac03bb22e0b17392546c4f4f2d728f9_performance_tests/09 job id: 42881820 --- 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/d60029441ac03bb22e0b17392546c4f4f2d728f9_performance_tests/10 job id: 42881821 --- 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/d60029441ac03bb22e0b17392546c4f4f2d728f9_performance_tests/11 job id: 42881822 --- 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/d60029441ac03bb22e0b17392546c4f4f2d728f9_performance_tests/12 job id: 42881823 --- 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/d60029441ac03bb22e0b17392546c4f4f2d728f9_performance_tests/13 job id: 42881824 --- 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/d60029441ac03bb22e0b17392546c4f4f2d728f9_performance_tests/14 job id: 42881825 --- 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/d60029441ac03bb22e0b17392546c4f4f2d728f9_performance_tests/15 job id: 42881827 --- 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/d60029441ac03bb22e0b17392546c4f4f2d728f9_performance_tests/16 job id: 42881828 --- 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/d60029441ac03bb22e0b17392546c4f4f2d728f9_performance_tests/17 job id: 42881829 --- 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/d60029441ac03bb22e0b17392546c4f4f2d728f9_performance_tests/18 job id: 42881830 --- 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/d60029441ac03bb22e0b17392546c4f4f2d728f9_performance_tests/19 job id: 42881831 --- 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/d60029441ac03bb22e0b17392546c4f4f2d728f9_performance_tests/20 job id: 42881832 --- 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/d60029441ac03bb22e0b17392546c4f4f2d728f9_performance_tests/21 job id: 42881833 --- 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/d60029441ac03bb22e0b17392546c4f4f2d728f9_performance_tests/22 job id: 42881834 --- 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/d60029441ac03bb22e0b17392546c4f4f2d728f9_performance_tests/23 job id: 42881835 --- 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/d60029441ac03bb22e0b17392546c4f4f2d728f9_performance_tests/24 job id: 42881836 --- 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/d60029441ac03bb22e0b17392546c4f4f2d728f9_performance_tests/25 job id: 42881837 --- 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/d60029441ac03bb22e0b17392546c4f4f2d728f9_performance_tests/26 job id: 42881839 --- 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/d60029441ac03bb22e0b17392546c4f4f2d728f9_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 0.023 0.037 133.918 133.919 farming_run 1 2.0 133.353 133.357 133.884 133.887 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.455354E+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 489 2263609. 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 0.013 0.036 115.664 115.664 qs_energies 1 2.0 0.000 0.000 115.455 115.457 mp2_main 1 3.0 0.000 0.000 113.221 113.223 mp2_gpw_main 1 4.0 0.019 0.026 112.198 112.200 mp2_ri_gpw_compute_in 1 5.0 0.171 0.173 93.118 93.466 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 55.571 55.919 mp2_eri_3c_integrate_gpw 272 7.0 0.154 0.169 41.886 47.176 get_2c_integrals 1 6.0 0.000 0.000 36.831 37.375 integrate_v_rspace 273 8.0 0.435 0.447 25.192 30.235 pw_transfer 6555 10.6 0.375 0.388 27.408 27.942 fft_wrap_pw1pw2 5465 11.4 0.044 0.046 26.089 26.622 grid_integrate_task_list 273 9.0 20.995 26.551 20.995 26.551 fft_wrap_pw1pw2_100 2178 12.4 1.163 1.289 23.638 24.179 compute_2c_integrals 1 7.0 0.002 0.002 19.380 19.381 rpa_ri_compute_en 1 5.0 0.000 0.000 18.973 19.300 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.872 19.045 mp2_eri_2c_integrate_gpw 1 9.0 2.402 2.443 18.869 19.041 cp_fm_cholesky_decompose 12 8.2 17.469 17.965 17.469 17.965 cholesky_decomp 1 7.0 0.000 0.000 16.297 16.818 fft3d_s 5443 13.4 16.195 16.592 16.217 16.613 ao_to_mo_and_store_B_mult_1 272 7.0 10.860 15.586 10.860 15.586 calculate_wavefunction 272 8.0 5.477 5.623 12.614 13.203 rpa_num_int 1 6.0 0.000 0.000 10.672 10.682 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.570 10.649 calc_mat_Q 8 8.0 0.000 0.000 9.350 9.440 contract_S_to_Q 8 9.0 0.000 0.000 8.772 8.864 calc_potential_gpw 544 9.5 0.005 0.006 8.232 8.590 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.201 8.434 potential_pw2rs 545 10.0 0.107 0.109 7.711 8.425 parallel_gemm_fm 14 9.1 0.000 0.000 8.360 8.425 parallel_gemm_fm_cosma 14 10.1 8.360 8.425 8.360 8.425 create_integ_mat 1 6.0 0.014 0.027 7.858 7.868 collocate_single_gaussian 272 10.0 0.040 0.043 7.450 7.726 array2fm 1 7.0 0.000 0.000 6.693 7.117 pw_scatter_s 2720 13.7 4.441 4.665 4.441 4.665 pw_gather_s 2722 13.2 3.903 4.232 3.903 4.232 array2fm_buffer_send 1 8.0 2.959 3.097 2.959 3.097 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.198463, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2727.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d60029441ac03bb22e0b17392546c4f4f2d728f9_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 0.033 0.054 401.567 401.567 farming_run 1 2.0 400.506 400.511 401.509 401.521 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827128576 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788821 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.224225E+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 1825 23678. MP_Sync 38 MP_Alltoall 77 9993341. 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 0.015 0.046 211.254 211.255 qs_energies 1 2.0 0.001 0.007 210.946 210.964 scf_env_do_scf 1 3.0 0.000 0.000 107.273 107.273 qs_ks_update_qs_env 5 5.0 0.000 0.000 106.198 106.206 rebuild_ks_matrix 4 6.0 0.000 0.000 106.196 106.205 qs_ks_build_kohn_sham_matrix 4 7.0 0.059 0.068 106.196 106.205 hfx_ks_matrix 4 8.0 0.001 0.001 105.782 105.785 integrate_four_center 4 9.0 0.144 0.451 105.781 105.784 mp2_main 1 3.0 0.010 0.085 103.370 103.383 mp2_gpw_main 1 4.0 0.043 0.119 102.103 102.123 integrate_four_center_main 4 10.0 0.097 0.538 96.453 99.662 integrate_four_center_bin 261 11.0 96.356 99.650 96.356 99.650 init_scf_loop 1 4.0 0.000 0.000 92.846 92.847 mp2_ri_gpw_compute_in 1 5.0 0.080 0.139 75.004 76.079 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.021 54.490 55.568 mp2_eri_3c_integrate_gpw 91 7.0 0.145 0.161 42.080 47.349 integrate_v_rspace 95 8.0 0.397 0.576 28.474 33.533 pw_transfer 2240 10.6 0.143 0.160 29.948 30.395 fft_wrap_pw1pw2 1868 11.4 0.017 0.020 28.941 29.367 ao_to_mo_and_store_B_mult_1 91 7.0 10.730 28.997 10.730 28.997 grid_integrate_task_list 95 9.0 23.728 28.974 23.728 28.974 mp2_ri_gpw_compute_en 1 5.0 0.074 0.173 26.850 28.523 fft_wrap_pw1pw2_100 730 12.4 1.303 1.451 26.644 27.085 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.834 1.890 24.977 24.985 get_2c_integrals 1 6.0 0.004 0.025 20.382 20.446 compute_2c_integrals 1 7.0 0.003 0.005 19.348 19.370 compute_2c_integrals_loop_lm 1 8.0 0.007 0.043 18.949 19.212 mp2_eri_2c_integrate_gpw 1 9.0 1.735 1.973 18.942 19.211 fft3d_s 1823 13.4 18.413 18.865 18.427 18.878 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.425 14.426 calculate_wavefunction 91 8.0 2.022 2.051 9.752 9.940 potential_pw2rs 186 10.0 0.034 0.035 8.678 9.316 mp2_ri_gpw_compute_en_expansio 172 7.0 0.557 0.599 8.786 9.272 local_gemm 172 8.0 8.229 8.699 8.229 8.699 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.296 8.571 mp2_ri_gpw_compute_en_comm 22 7.0 0.507 0.527 7.964 8.475 calc_potential_gpw 182 9.5 0.002 0.002 7.963 8.174 collocate_single_gaussian 91 10.0 0.018 0.035 7.904 8.103 mp_sync 38 10.4 3.955 6.892 3.955 6.892 mp_sendrecv_dm3 2068 8.0 5.983 6.464 5.983 6.464 mp2_ri_gpw_compute_en_ener 172 7.0 6.348 6.459 6.348 6.459 pw_gather_s 912 13.2 4.906 5.478 4.906 5.478 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=102.104706, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1511.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d60029441ac03bb22e0b17392546c4f4f2d728f9_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 451.309568E+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 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 0.015 0.030 53.503 53.504 qs_mol_dyn_low 1 2.0 0.003 0.003 53.242 53.251 qs_forces 11 3.9 0.002 0.002 53.161 53.163 qs_energies 11 4.9 0.001 0.002 51.586 51.599 scf_env_do_scf 11 5.9 0.000 0.001 45.181 45.181 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 42.829 42.829 dbcsr_multiply_generic 2286 12.5 0.095 0.097 33.281 33.748 qs_scf_new_mos 108 7.5 0.000 0.001 31.679 31.949 qs_scf_loop_do_ot 108 8.5 0.000 0.001 31.678 31.949 ot_scf_mini 108 9.5 0.002 0.002 30.042 30.232 multiply_cannon 2286 13.5 0.182 0.192 26.169 28.060 multiply_cannon_loop 2286 14.5 1.475 1.554 25.215 27.212 velocity_verlet 10 3.0 0.001 0.001 26.027 26.034 ot_mini 108 10.5 0.001 0.001 19.090 19.336 qs_ot_get_derivative 108 11.5 0.001 0.001 16.160 16.368 mp_waitall_1 267858 16.1 9.034 15.231 9.034 15.231 multiply_cannon_metrocomm3 54864 15.5 0.069 0.074 5.723 13.017 multiply_cannon_multrec 54864 15.5 4.328 6.684 7.948 11.697 rebuild_ks_matrix 119 8.3 0.000 0.000 8.581 8.708 qs_ks_build_kohn_sham_matrix 119 9.3 0.014 0.023 8.581 8.708 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.557 7.671 qs_ot_get_p 119 10.4 0.001 0.001 6.400 6.695 multiply_cannon_sync_h2d 54864 15.5 5.992 6.596 5.992 6.596 mp_sum_l 7207 12.9 4.516 6.209 4.516 6.209 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.565 6.005 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.167 5.268 dbcsr_mm_accdrv_process 76910 16.1 1.116 1.806 3.543 5.128 sum_up_and_integrate 119 10.3 0.012 0.014 5.109 5.116 integrate_v_rspace 119 11.3 0.002 0.002 5.097 5.105 init_scf_run 11 5.9 0.000 0.001 5.085 5.086 scf_env_initial_rho_setup 11 6.9 0.000 0.000 5.085 5.085 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.704 4.826 calculate_rho_elec 119 8.7 0.011 0.016 4.704 4.826 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.551 3.601 rs_pw_transfer 974 11.9 0.011 0.012 3.451 3.579 jit_kernel_multiply 13 15.8 2.368 3.423 2.368 3.423 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.963 3.200 apply_single 119 13.6 0.000 0.000 2.963 3.200 calculate_dm_sparse 119 9.5 0.000 0.000 2.862 2.978 density_rs2pw 119 9.7 0.004 0.004 2.758 2.856 calculate_first_density_matrix 1 7.0 0.000 0.001 2.771 2.780 cp_dbcsr_syevd 50 12.0 0.002 0.003 2.735 2.736 ot_diis_step 108 11.5 0.006 0.006 2.688 2.689 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.633 2.635 multiply_cannon_metrocomm1 54864 15.5 0.053 0.059 1.680 2.610 potential_pw2rs 119 12.3 0.004 0.004 2.405 2.445 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.422 2.423 cp_fm_redistribute_end 50 14.0 2.190 2.391 2.195 2.393 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.329 2.389 cp_fm_diag_elpa_base 50 14.0 0.195 2.347 0.196 2.356 make_m2s 4572 13.5 0.055 0.057 2.281 2.343 init_scf_loop 11 6.9 0.000 0.000 2.329 2.329 pw_transfer 1439 11.6 0.051 0.055 2.173 2.280 make_images 4572 14.5 0.133 0.140 2.197 2.259 wfi_extrapolate 11 7.9 0.001 0.001 2.236 2.236 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.148 2.211 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 2.098 2.207 acc_transpose_blocks 54864 15.5 0.217 0.240 1.679 2.132 grid_integrate_task_list 119 12.3 2.002 2.109 2.002 2.109 fft3d_ps 1201 14.6 0.363 0.467 1.879 1.979 mp_sum_d 4125 12.0 1.232 1.846 1.232 1.846 mp_alltoall_d11v 2130 13.8 1.605 1.720 1.605 1.720 fft_wrap_pw1pw2_140 487 13.2 0.077 0.091 1.599 1.707 mp_waitany 12084 13.8 1.305 1.492 1.305 1.492 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.407 1.426 grid_collocate_task_list 119 9.7 1.287 1.359 1.287 1.359 make_images_sizes 4572 15.5 0.004 0.004 0.954 1.210 prepare_preconditioner 11 7.9 0.000 0.000 1.175 1.207 make_preconditioner 11 8.9 0.000 0.000 1.175 1.207 mp_alltoall_i44 4572 16.5 0.950 1.206 0.950 1.206 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.116 1.153 mp_alltoall_z22v 1201 16.6 1.050 1.125 1.050 1.125 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.068 1.085 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=53.504000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=430.545455, yerr=0.655555 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d60029441ac03bb22e0b17392546c4f4f2d728f9_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 486.780928E+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 1151403. 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 0.044 0.074 45.394 45.396 qs_mol_dyn_low 1 2.0 0.005 0.017 44.584 44.602 qs_forces 11 3.9 0.002 0.009 44.488 44.489 qs_energies 11 4.9 0.003 0.008 42.673 42.684 scf_env_do_scf 11 5.9 0.001 0.001 36.760 36.760 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 33.434 33.435 dbcsr_multiply_generic 2286 12.5 0.102 0.105 23.840 24.398 qs_scf_new_mos 108 7.5 0.001 0.001 22.854 23.123 qs_scf_loop_do_ot 108 8.5 0.001 0.001 22.853 23.122 ot_scf_mini 108 9.5 0.003 0.003 21.867 22.060 velocity_verlet 10 3.0 0.002 0.010 21.643 21.645 multiply_cannon 2286 13.5 0.209 0.218 17.282 19.180 multiply_cannon_loop 2286 14.5 0.907 0.984 15.832 18.041 ot_mini 108 10.5 0.001 0.002 13.647 13.911 mp_waitall_1 217478 16.2 7.962 13.630 7.962 13.630 qs_ot_get_derivative 108 11.5 0.001 0.001 10.964 11.149 multiply_cannon_metrocomm3 27432 15.5 0.068 0.070 4.690 10.290 multiply_cannon_multrec 27432 15.5 1.922 4.459 5.999 8.825 rebuild_ks_matrix 119 8.3 0.000 0.000 8.130 8.305 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.017 8.130 8.304 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.219 7.380 dbcsr_mm_accdrv_process 47894 16.0 3.115 5.320 4.008 5.947 qs_ot_get_p 119 10.4 0.001 0.001 5.035 5.259 sum_up_and_integrate 119 10.3 0.025 0.030 5.010 5.016 integrate_v_rspace 119 11.3 0.002 0.002 4.985 4.992 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 4.112 4.955 mp_sum_l 7207 12.9 2.761 4.947 2.761 4.947 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.499 4.562 calculate_rho_elec 119 8.7 0.021 0.024 4.498 4.561 init_scf_run 11 5.9 0.000 0.001 4.518 4.519 scf_env_initial_rho_setup 11 6.9 0.000 0.002 4.518 4.518 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.155 4.259 apply_single 119 13.6 0.000 0.000 3.155 4.259 rs_pw_transfer 974 11.9 0.009 0.011 3.709 3.865 make_m2s 4572 13.5 0.053 0.055 3.367 3.724 make_images 4572 14.5 0.199 0.234 3.276 3.631 init_scf_loop 11 6.9 0.001 0.002 3.284 3.285 qs_ot_p2m_diag 50 11.0 0.008 0.012 3.062 3.090 density_rs2pw 119 9.7 0.004 0.004 2.809 2.912 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.722 2.826 multiply_cannon_sync_h2d 27432 15.5 2.120 2.727 2.120 2.727 potential_pw2rs 119 12.3 0.006 0.006 2.662 2.691 calculate_first_density_matrix 1 7.0 0.001 0.004 2.656 2.659 ot_diis_step 108 11.5 0.011 0.013 2.587 2.588 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.557 2.557 jit_kernel_multiply 11 16.3 0.841 2.406 0.841 2.406 pw_transfer 1439 11.6 0.063 0.066 2.301 2.376 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.304 2.307 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.211 2.288 prepare_preconditioner 11 7.9 0.000 0.000 2.209 2.239 make_preconditioner 11 8.9 0.001 0.002 2.208 2.239 calculate_dm_sparse 119 9.5 0.000 0.000 2.154 2.230 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.091 2.150 make_images_data 4572 15.5 0.045 0.052 1.405 2.071 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.974 2.023 fft3d_ps 1201 14.6 0.504 0.562 1.922 1.993 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.976 1.978 cp_fm_redistribute_end 50 14.0 1.607 1.924 1.612 1.926 grid_integrate_task_list 119 12.3 1.822 1.914 1.822 1.914 hybrid_alltoall_any 4725 16.4 0.051 0.112 1.221 1.913 cp_fm_diag_elpa_base 50 14.0 0.303 1.834 0.311 1.866 mp_alltoall_d11v 2130 13.8 1.644 1.848 1.644 1.848 wfi_extrapolate 11 7.9 0.001 0.003 1.796 1.796 mp_allgather_i34 2286 14.5 0.889 1.738 0.889 1.738 fft_wrap_pw1pw2_140 487 13.2 0.074 0.081 1.481 1.559 make_images_sizes 4572 15.5 0.005 0.005 1.152 1.519 mp_alltoall_i44 4572 16.5 1.147 1.514 1.147 1.514 acc_transpose_blocks 27432 15.5 0.107 0.112 1.181 1.475 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.368 1.413 grid_collocate_task_list 119 9.7 1.242 1.330 1.242 1.330 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.303 1.319 mp_sum_d 4125 12.0 0.813 1.280 0.813 1.280 mp_alltoall_z22v 1201 16.6 1.046 1.167 1.046 1.167 dbcsr_complete_redistribute 329 12.2 0.078 0.100 0.973 1.054 rs_pw_transfer_PW2RS_50 119 14.3 0.587 0.606 0.996 1.045 cp_fm_cholesky_invert 11 10.9 1.037 1.044 1.037 1.044 qs_energies_init_hamiltonians 11 5.9 0.006 0.028 1.039 1.041 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.976 0.994 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=45.396000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=463.909091, yerr=1.564059 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d60029441ac03bb22e0b17392546c4f4f2d728f9_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 524.328960E+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 0.058 0.101 33.429 33.431 qs_mol_dyn_low 1 2.0 0.003 0.003 32.778 32.786 qs_forces 11 3.9 0.002 0.002 32.701 32.712 qs_energies 11 4.9 0.003 0.006 31.047 31.059 scf_env_do_scf 11 5.9 0.001 0.002 26.118 26.118 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 23.429 23.430 dbcsr_multiply_generic 2286 12.5 0.094 0.097 16.730 16.825 velocity_verlet 10 3.0 0.001 0.001 15.740 15.741 qs_scf_new_mos 108 7.5 0.001 0.001 15.176 15.202 qs_scf_loop_do_ot 108 8.5 0.001 0.002 15.176 15.202 ot_scf_mini 108 9.5 0.002 0.003 14.449 14.467 multiply_cannon 2286 13.5 0.196 0.207 13.146 13.968 multiply_cannon_loop 2286 14.5 0.640 0.669 12.308 13.179 ot_mini 108 10.5 0.001 0.001 9.047 9.066 qs_ot_get_derivative 108 11.5 0.001 0.001 7.544 7.561 multiply_cannon_multrec 18288 15.5 1.931 2.939 6.753 7.072 rebuild_ks_matrix 119 8.3 0.000 0.000 6.574 6.596 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.574 6.596 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.801 5.823 dbcsr_mm_accdrv_process 38222 16.0 4.121 5.508 4.739 5.679 mp_waitall_1 169478 16.3 3.368 4.479 3.368 4.479 sum_up_and_integrate 119 10.3 0.030 0.031 4.168 4.174 integrate_v_rspace 119 11.3 0.002 0.003 4.138 4.146 init_scf_run 11 5.9 0.000 0.001 3.661 3.662 scf_env_initial_rho_setup 11 6.9 0.000 0.001 3.661 3.661 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.880 3.532 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.374 3.383 calculate_rho_elec 119 8.7 0.030 0.031 3.373 3.382 qs_ot_get_p 119 10.4 0.001 0.001 3.184 3.208 init_scf_loop 11 6.9 0.000 0.001 2.672 2.673 rs_pw_transfer 974 11.9 0.009 0.009 2.560 2.665 multiply_cannon_metrocomm3 18288 15.5 0.044 0.045 1.446 2.417 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.064 2.382 apply_single 119 13.6 0.000 0.000 2.064 2.382 calculate_first_density_matrix 1 7.0 0.005 0.023 2.375 2.377 make_m2s 4572 13.5 0.045 0.046 2.024 2.184 make_images 4572 14.5 0.189 0.202 1.938 2.096 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.034 2.039 density_rs2pw 119 9.7 0.004 0.004 1.956 2.022 potential_pw2rs 119 12.3 0.007 0.008 1.936 1.944 grid_integrate_task_list 119 12.3 1.796 1.900 1.796 1.900 jit_kernel_multiply 10 16.3 0.567 1.866 0.567 1.866 prepare_preconditioner 11 7.9 0.000 0.000 1.841 1.843 make_preconditioner 11 8.9 0.000 0.001 1.841 1.843 calculate_dm_sparse 119 9.5 0.000 0.000 1.825 1.836 pw_transfer 1439 11.6 0.063 0.067 1.817 1.828 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.696 1.780 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.764 1.765 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.726 1.739 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.684 1.685 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.659 1.667 multiply_cannon_sync_h2d 18288 15.5 1.394 1.532 1.394 1.532 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.521 1.525 mp_sum_l 7207 12.9 1.172 1.518 1.172 1.518 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.514 1.515 cp_fm_redistribute_end 50 14.0 1.121 1.490 1.122 1.491 ot_diis_step 108 11.5 0.011 0.011 1.484 1.484 cp_fm_diag_elpa_base 50 14.0 0.353 1.430 0.367 1.468 fft3d_ps 1201 14.6 0.509 0.525 1.412 1.424 fft_wrap_pw1pw2_140 487 13.2 0.086 0.090 1.321 1.333 grid_collocate_task_list 119 9.7 1.202 1.274 1.202 1.274 acc_transpose_blocks 18288 15.5 0.074 0.076 1.219 1.241 wfi_extrapolate 11 7.9 0.001 0.001 1.230 1.231 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.195 1.198 make_images_data 4572 15.5 0.046 0.050 0.914 1.108 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 1.025 1.036 hybrid_alltoall_any 4725 16.4 0.055 0.112 0.788 1.004 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.913 0.935 mp_alltoall_d11v 2130 13.8 0.756 0.861 0.756 0.861 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.844 0.846 mp_alltoall_z22v 1201 16.6 0.741 0.805 0.741 0.805 acc_transpose_blocks_kernels 18288 16.5 0.207 0.218 0.784 0.798 cp_fm_cholesky_invert 11 10.9 0.769 0.773 0.769 0.773 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.660 0.750 mp_waitany 9880 13.7 0.578 0.704 0.578 0.704 mp_allgather_i34 2286 14.5 0.297 0.682 0.297 0.682 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=33.431000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=497.727273, yerr=2.699862 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d60029441ac03bb22e0b17392546c4f4f2d728f9_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 545.366016E+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 0.016 0.030 39.591 39.591 qs_mol_dyn_low 1 2.0 0.003 0.003 39.234 39.243 qs_forces 11 3.9 0.002 0.002 38.876 38.878 qs_energies 11 4.9 0.001 0.001 37.055 37.063 scf_env_do_scf 11 5.9 0.000 0.001 31.327 31.328 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 27.118 27.118 velocity_verlet 10 3.0 0.001 0.001 20.355 20.369 dbcsr_multiply_generic 2286 12.5 0.101 0.103 19.832 19.952 qs_scf_new_mos 108 7.5 0.001 0.001 18.075 18.139 qs_scf_loop_do_ot 108 8.5 0.001 0.001 18.074 18.138 ot_scf_mini 108 9.5 0.002 0.003 17.081 17.137 multiply_cannon 2286 13.5 0.233 0.276 15.377 15.809 multiply_cannon_loop 2286 14.5 0.945 0.974 14.232 14.640 ot_mini 108 10.5 0.001 0.001 10.510 10.567 multiply_cannon_multrec 27432 15.5 2.368 3.153 9.087 9.613 qs_ot_get_derivative 108 11.5 0.001 0.001 8.650 8.707 dbcsr_mm_accdrv_process 47916 15.9 5.420 7.275 6.624 7.995 rebuild_ks_matrix 119 8.3 0.000 0.000 7.037 7.083 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.037 7.083 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.241 6.283 init_scf_run 11 5.9 0.000 0.001 4.341 4.341 scf_env_initial_rho_setup 11 6.9 0.000 0.000 4.341 4.341 sum_up_and_integrate 119 10.3 0.035 0.037 4.230 4.238 integrate_v_rspace 119 11.3 0.002 0.002 4.195 4.203 init_scf_loop 11 6.9 0.000 0.000 4.173 4.174 mp_waitall_1 145218 16.4 3.263 4.071 3.263 4.071 qs_ot_get_p 119 10.4 0.001 0.001 3.821 3.887 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.835 3.875 calculate_rho_elec 119 8.7 0.040 0.046 3.834 3.875 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.339 3.843 prepare_preconditioner 11 7.9 0.000 0.000 3.229 3.237 make_preconditioner 11 8.9 0.000 0.000 3.229 3.237 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.804 3.154 make_m2s 4572 13.5 0.055 0.057 2.729 2.797 rs_pw_transfer 974 11.9 0.009 0.009 2.675 2.774 calculate_first_density_matrix 1 7.0 0.000 0.000 2.725 2.727 make_images 4572 14.5 0.271 0.329 2.619 2.685 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.172 2.628 apply_single 119 13.6 0.000 0.000 2.172 2.628 density_rs2pw 119 9.7 0.004 0.004 2.262 2.376 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.287 2.288 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.262 2.272 calculate_dm_sparse 119 9.5 0.000 0.000 2.193 2.250 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.148 2.175 pw_transfer 1439 11.6 0.063 0.066 2.065 2.097 jit_kernel_multiply 10 15.8 1.144 2.047 1.144 2.047 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.975 2.007 potential_pw2rs 119 12.3 0.008 0.009 1.982 1.988 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.918 1.919 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.885 1.900 grid_integrate_task_list 119 12.3 1.834 1.893 1.834 1.893 mp_sum_l 7207 12.9 1.308 1.848 1.308 1.848 ot_diis_step 108 11.5 0.012 0.012 1.815 1.816 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.963 1.779 fft3d_ps 1201 14.6 0.535 0.587 1.657 1.687 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.604 1.605 cp_fm_redistribute_end 50 14.0 1.050 1.565 1.051 1.566 wfi_extrapolate 11 7.9 0.001 0.001 1.556 1.556 fft_wrap_pw1pw2_140 487 13.2 0.083 0.089 1.502 1.536 cp_fm_diag_elpa_base 50 14.0 0.489 1.493 0.510 1.535 acc_transpose_blocks 27432 15.5 0.110 0.111 1.486 1.517 dbcsr_complete_redistribute 329 12.2 0.190 0.268 1.156 1.448 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.332 1.343 grid_collocate_task_list 119 9.7 1.220 1.327 1.220 1.327 make_images_data 4572 15.5 0.045 0.048 1.098 1.244 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.204 1.220 cp_fm_upper_to_full 72 13.5 0.856 1.192 0.856 1.192 mp_alltoall_d11v 2130 13.8 1.011 1.154 1.011 1.154 cp_fm_cholesky_invert 11 10.9 1.141 1.146 1.141 1.146 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.833 1.118 hybrid_alltoall_any 4725 16.4 0.062 0.151 0.948 1.093 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.081 1.082 multiply_cannon_sync_h2d 27432 15.5 0.980 1.046 0.980 1.046 mp_alltoall_z22v 1201 16.6 1.004 1.028 1.004 1.028 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.889 0.892 acc_transpose_blocks_kernels 27432 16.5 0.265 0.273 0.867 0.892 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.793 0.868 mp_alltoall_i22 627 13.8 0.494 0.805 0.494 0.805 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=39.591000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=517.545455, yerr=3.257541 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d60029441ac03bb22e0b17392546c4f4f2d728f9_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 595.460096E+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 0.023 0.047 29.548 29.549 qs_mol_dyn_low 1 2.0 0.003 0.003 29.347 29.361 qs_forces 11 3.9 0.002 0.002 29.284 29.286 qs_energies 11 4.9 0.001 0.002 27.552 27.556 scf_env_do_scf 11 5.9 0.000 0.001 22.791 22.791 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 19.958 19.959 velocity_verlet 10 3.0 0.001 0.001 15.283 15.286 dbcsr_multiply_generic 2286 12.5 0.093 0.096 12.766 12.867 qs_scf_new_mos 108 7.5 0.001 0.001 11.927 11.955 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.926 11.955 ot_scf_mini 108 9.5 0.002 0.002 11.209 11.235 multiply_cannon 2286 13.5 0.232 0.239 9.788 10.362 multiply_cannon_loop 2286 14.5 0.334 0.344 8.746 8.985 ot_mini 108 10.5 0.001 0.001 6.119 6.153 rebuild_ks_matrix 119 8.3 0.000 0.000 6.089 6.125 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.089 6.125 multiply_cannon_multrec 9144 15.5 1.583 1.846 5.751 5.962 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.436 5.471 qs_ot_get_derivative 108 11.5 0.001 0.001 4.722 4.746 dbcsr_mm_accdrv_process 12550 15.8 3.031 3.897 4.067 4.159 sum_up_and_integrate 119 10.3 0.037 0.041 3.788 3.792 integrate_v_rspace 119 11.3 0.002 0.003 3.750 3.754 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.457 3.467 calculate_rho_elec 119 8.7 0.059 0.061 3.457 3.466 init_scf_run 11 5.9 0.000 0.001 3.309 3.309 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.309 3.309 qs_ot_get_p 119 10.4 0.001 0.001 3.183 3.225 mp_waitall_1 121218 16.5 2.247 2.851 2.247 2.851 init_scf_loop 11 6.9 0.000 0.000 2.812 2.814 make_m2s 4572 13.5 0.035 0.036 2.030 2.299 make_images 4572 14.5 0.267 0.301 1.939 2.207 calculate_first_density_matrix 1 7.0 0.000 0.000 2.100 2.101 qs_ot_p2m_diag 50 11.0 0.022 0.023 2.080 2.084 rs_pw_transfer 974 11.9 0.008 0.008 1.969 2.062 prepare_preconditioner 11 7.9 0.000 0.000 2.018 2.023 make_preconditioner 11 8.9 0.000 0.000 2.018 2.023 grid_integrate_task_list 119 12.3 1.857 1.964 1.857 1.964 density_rs2pw 119 9.7 0.003 0.004 1.841 1.931 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.888 1.917 pw_transfer 1439 11.6 0.063 0.066 1.851 1.859 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.854 1.855 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.760 1.769 jit_kernel_multiply 10 15.8 0.998 1.767 0.998 1.767 calculate_dm_sparse 119 9.5 0.000 0.000 1.744 1.766 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.682 1.697 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.564 1.566 potential_pw2rs 119 12.3 0.010 0.010 1.540 1.545 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.496 1.498 cp_fm_redistribute_end 50 14.0 0.742 1.467 0.743 1.468 cp_fm_diag_elpa_base 50 14.0 0.679 1.387 0.723 1.446 fft3d_ps 1201 14.6 0.539 0.553 1.427 1.434 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.376 1.386 fft_wrap_pw1pw2_140 487 13.2 0.082 0.086 1.375 1.385 grid_collocate_task_list 119 9.7 1.278 1.373 1.278 1.373 ot_diis_step 108 11.5 0.012 0.013 1.371 1.372 make_images_data 4572 15.5 0.039 0.042 0.961 1.286 hybrid_alltoall_any 4725 16.4 0.063 0.177 0.922 1.280 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.260 1.272 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.230 1.270 apply_single 119 13.6 0.000 0.000 1.230 1.270 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.222 1.225 wfi_extrapolate 11 7.9 0.001 0.001 1.158 1.158 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.127 1.132 cp_fm_cholesky_invert 11 10.9 1.051 1.055 1.051 1.055 mp_alltoall_d11v 2130 13.8 0.937 1.008 0.937 1.008 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.866 0.917 acc_transpose_blocks 9144 15.5 0.038 0.039 0.891 0.897 mp_allgather_i34 2286 14.5 0.372 0.857 0.372 0.857 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.828 0.836 mp_alltoall_z22v 1201 16.6 0.762 0.792 0.762 0.792 multiply_cannon_sync_h2d 9144 15.5 0.708 0.791 0.708 0.791 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.778 0.780 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.357 0.777 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.678 0.729 dbcsr_complete_redistribute 329 12.2 0.163 0.171 0.630 0.660 mp_sum_l 7207 12.9 0.519 0.654 0.519 0.654 acc_transpose_blocks_kernels 9144 16.5 0.115 0.119 0.651 0.654 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=29.549000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=565.636364, yerr=3.418472 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d60029441ac03bb22e0b17392546c4f4f2d728f9_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 752.254976E+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 0.016 0.037 42.807 42.807 qs_mol_dyn_low 1 2.0 0.003 0.003 42.564 42.572 qs_forces 11 3.9 0.001 0.002 42.505 42.506 qs_energies 11 4.9 0.003 0.007 40.518 40.522 scf_env_do_scf 11 5.9 0.001 0.001 34.719 34.719 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 26.367 26.368 velocity_verlet 10 3.0 0.001 0.001 24.163 24.168 dbcsr_multiply_generic 2286 12.5 0.101 0.102 18.239 18.512 qs_scf_new_mos 108 7.5 0.001 0.001 16.984 17.085 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.984 17.085 ot_scf_mini 108 9.5 0.002 0.002 15.861 15.962 multiply_cannon 2286 13.5 0.308 0.319 13.923 15.019 multiply_cannon_loop 2286 14.5 0.347 0.352 12.563 13.652 ot_mini 108 10.5 0.001 0.001 9.326 9.454 multiply_cannon_multrec 9144 15.5 3.344 4.741 8.523 8.628 init_scf_loop 11 6.9 0.000 0.000 8.321 8.322 prepare_preconditioner 11 7.9 0.000 0.000 7.358 7.372 make_preconditioner 11 8.9 0.000 0.000 7.358 7.372 qs_ot_get_derivative 108 11.5 0.001 0.001 7.214 7.318 rebuild_ks_matrix 119 8.3 0.000 0.000 7.139 7.299 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.138 7.299 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.902 7.243 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.449 6.595 dbcsr_mm_accdrv_process 12550 15.8 4.154 5.630 5.055 6.361 cp_fm_upper_to_full 72 14.2 3.222 4.616 3.222 4.616 mp_waitall_1 97218 16.6 3.089 4.391 3.089 4.391 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.937 3.946 calculate_rho_elec 119 8.7 0.118 0.121 3.936 3.946 sum_up_and_integrate 119 10.3 0.064 0.065 3.888 3.894 integrate_v_rspace 119 11.3 0.003 0.004 3.824 3.830 qs_ot_get_p 119 10.4 0.001 0.001 3.648 3.797 init_scf_run 11 5.9 0.000 0.001 3.720 3.720 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.720 3.720 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.644 3.039 dbcsr_complete_redistribute 329 12.2 0.292 0.299 2.080 2.906 make_m2s 4572 13.5 0.038 0.039 2.602 2.814 make_images 4572 14.5 0.351 0.378 2.480 2.692 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.720 2.549 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 1.488 2.541 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.218 2.512 apply_single 119 13.6 0.000 0.000 2.218 2.512 mp_alltoall_i22 627 13.8 1.440 2.304 1.440 2.304 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.451 2.272 qs_ot_p2m_diag 50 11.0 0.043 0.043 2.233 2.234 calculate_first_density_matrix 1 7.0 0.000 0.000 2.223 2.224 pw_transfer 1439 11.6 0.066 0.067 2.214 2.217 calculate_dm_sparse 119 9.5 0.000 0.000 2.194 2.209 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.118 2.121 ot_diis_step 108 11.5 0.014 0.014 2.078 2.078 grid_integrate_task_list 119 12.3 2.055 2.067 2.055 2.067 density_rs2pw 119 9.7 0.004 0.004 1.959 1.973 mp_sum_l 7207 12.9 1.177 1.967 1.177 1.967 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.948 1.949 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.935 1.936 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.861 1.914 qs_energies_init_hamiltonians 11 5.9 0.010 0.016 1.778 1.778 cp_fm_cholesky_invert 11 10.9 1.758 1.762 1.758 1.762 fft3d_ps 1201 14.6 0.567 0.578 1.753 1.756 fft_wrap_pw1pw2_140 487 13.2 0.087 0.089 1.729 1.733 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.662 1.701 rs_pw_transfer 974 11.9 0.009 0.009 1.573 1.597 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.592 1.592 cp_fm_diag_elpa_base 50 14.0 1.444 1.495 1.589 1.590 hybrid_alltoall_any 4725 16.4 0.087 0.148 1.278 1.569 make_images_data 4572 15.5 0.043 0.046 1.275 1.548 grid_collocate_task_list 119 9.7 1.487 1.504 1.487 1.504 wfi_extrapolate 11 7.9 0.001 0.001 1.439 1.439 jit_kernel_multiply 8 15.4 0.874 1.412 0.874 1.412 potential_pw2rs 119 12.3 0.014 0.014 1.354 1.356 mp_alltoall_d11v 2130 13.8 1.289 1.315 1.289 1.315 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.240 1.269 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.223 1.230 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.104 1.123 mp_alltoall_z22v 1201 16.6 1.049 1.070 1.049 1.070 multiply_cannon_sync_h2d 9144 15.5 1.047 1.049 1.047 1.049 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.981 1.018 acc_transpose_blocks 9144 15.5 0.037 0.037 1.001 1.004 qs_create_task_list 11 7.9 0.009 0.010 0.956 0.968 generate_qs_task_list 11 8.9 0.375 0.394 0.947 0.959 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.903 0.918 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=42.807000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=706.818182, yerr=12.415666 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d60029441ac03bb22e0b17392546c4f4f2d728f9_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 500.903936E+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 1067428. 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 0.033 0.069 84.464 84.465 qs_mol_dyn_low 1 2.0 0.005 0.013 84.066 84.080 qs_forces 11 3.9 0.002 0.004 83.976 83.978 qs_energies 11 4.9 0.003 0.009 81.084 81.098 scf_env_do_scf 11 5.9 0.001 0.001 72.050 72.052 scf_env_do_scf_inner_loop 99 6.5 0.002 0.008 66.124 66.126 dbcsr_multiply_generic 2055 12.4 0.107 0.111 52.749 53.084 qs_scf_new_mos 99 7.5 0.000 0.001 48.792 48.920 qs_scf_loop_do_ot 99 8.5 0.001 0.001 48.791 48.920 ot_scf_mini 99 9.5 0.002 0.003 46.377 46.488 multiply_cannon 2055 13.4 0.179 0.189 42.894 44.075 multiply_cannon_loop 2055 14.4 1.567 1.607 41.844 43.061 velocity_verlet 10 3.0 0.002 0.005 42.709 42.710 ot_mini 99 10.5 0.001 0.002 28.098 28.220 qs_ot_get_derivative 99 11.5 0.001 0.001 21.200 21.359 multiply_cannon_multrec 49320 15.4 12.076 12.826 17.172 17.967 rebuild_ks_matrix 110 8.3 0.000 0.001 14.830 15.038 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.019 14.830 15.038 mp_waitall_1 241148 16.1 12.352 13.251 12.352 13.251 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.013 13.203 multiply_cannon_sync_h2d 49320 15.4 9.932 10.659 9.932 10.659 qs_ot_get_p 110 10.4 0.001 0.002 9.635 9.796 multiply_cannon_metrocomm3 49320 15.4 0.079 0.084 7.037 8.422 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.292 7.844 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.194 7.605 apply_single 110 13.6 0.000 0.001 7.194 7.605 sum_up_and_integrate 110 10.3 0.036 0.045 7.270 7.301 integrate_v_rspace 110 11.3 0.002 0.003 7.234 7.274 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.933 7.006 init_scf_run 11 5.9 0.000 0.001 6.917 6.918 scf_env_initial_rho_setup 11 6.9 0.000 0.003 6.917 6.918 ot_diis_step 99 11.5 0.006 0.009 6.587 6.588 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.394 6.508 calculate_rho_elec 110 8.6 0.020 0.025 6.393 6.508 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.171 6.213 init_scf_loop 11 6.9 0.001 0.003 5.884 5.888 mp_sum_l 6514 12.8 5.246 5.843 5.246 5.843 dbcsr_mm_accdrv_process 87628 16.1 2.111 2.207 4.976 5.279 cp_dbcsr_syevd 48 12.0 0.002 0.002 5.130 5.130 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.594 4.596 cp_fm_redistribute_end 48 14.0 3.990 4.559 3.994 4.562 cp_fm_diag_elpa_base 48 14.0 0.560 4.482 0.563 4.508 make_m2s 4110 13.4 0.061 0.065 4.168 4.266 rs_pw_transfer 902 11.9 0.011 0.012 4.057 4.224 make_images 4110 14.4 0.178 0.191 4.071 4.170 wfi_extrapolate 11 7.9 0.001 0.003 4.124 4.124 calculate_dm_sparse 110 9.5 0.000 0.001 3.787 3.907 multiply_cannon_metrocomm1 49320 15.4 0.059 0.062 2.833 3.727 density_rs2pw 110 9.6 0.004 0.004 3.509 3.681 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.614 3.620 prepare_preconditioner 11 7.9 0.000 0.000 3.590 3.611 make_preconditioner 11 8.9 0.000 0.002 3.590 3.611 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.387 3.485 make_full_inverse_cholesky 11 9.9 0.000 0.001 3.363 3.415 grid_integrate_task_list 110 12.3 3.240 3.392 3.240 3.392 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.246 3.293 pw_transfer 1331 11.6 0.054 0.062 3.128 3.222 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 3.040 3.138 potential_pw2rs 110 12.3 0.006 0.007 2.873 2.921 calculate_first_density_matrix 1 7.0 0.001 0.003 2.700 2.706 mp_alltoall_d11v 2046 13.8 2.148 2.679 2.148 2.679 fft3d_ps 1111 14.6 0.750 0.836 2.580 2.662 fft_wrap_pw1pw2_140 451 13.1 0.167 0.187 2.507 2.608 jit_kernel_multiply 13 15.9 2.581 2.608 2.581 2.608 acc_transpose_blocks 49320 15.4 0.222 0.231 2.229 2.341 grid_collocate_task_list 110 9.6 2.084 2.199 2.084 2.199 mp_sum_d 3879 11.9 1.617 2.088 1.617 2.088 mp_waitany 14300 13.8 1.857 2.046 1.857 2.046 make_images_data 4110 15.4 0.043 0.046 1.865 2.006 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.958 1.984 cp_fm_cholesky_invert 11 10.9 1.957 1.961 1.957 1.961 hybrid_alltoall_any 4261 16.3 0.082 0.480 1.634 1.881 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=84.465000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.000000, yerr=2.215647 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d60029441ac03bb22e0b17392546c4f4f2d728f9_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 590.798848E+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 3187944. 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 0.023 0.035 71.318 71.319 qs_mol_dyn_low 1 2.0 0.003 0.003 70.938 70.948 qs_forces 11 3.9 0.002 0.003 70.862 70.863 qs_energies 11 4.9 0.002 0.007 67.470 67.474 scf_env_do_scf 11 5.9 0.001 0.001 58.648 58.652 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 50.216 50.217 dbcsr_multiply_generic 2055 12.4 0.115 0.119 38.274 38.475 velocity_verlet 10 3.0 0.001 0.001 37.963 37.964 qs_scf_new_mos 99 7.5 0.001 0.001 33.483 33.621 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.482 33.621 multiply_cannon 2055 13.4 0.224 0.247 31.104 32.407 ot_scf_mini 99 9.5 0.003 0.003 31.794 31.938 multiply_cannon_loop 2055 14.4 0.930 0.952 29.702 30.572 ot_mini 99 10.5 0.001 0.001 18.712 18.863 multiply_cannon_multrec 24660 15.4 7.639 9.144 13.850 15.250 rebuild_ks_matrix 110 8.3 0.000 0.001 14.204 14.320 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.017 14.204 14.319 qs_ot_get_derivative 99 11.5 0.001 0.001 12.913 13.059 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.534 12.635 mp_waitall_1 186928 16.3 8.566 11.141 8.566 11.141 init_scf_loop 11 6.9 0.002 0.011 8.390 8.391 multiply_cannon_sync_h2d 24660 15.4 7.008 8.268 7.008 8.268 multiply_cannon_metrocomm3 24660 15.4 0.069 0.071 5.129 7.527 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.544 7.233 apply_single 110 13.6 0.000 0.001 6.543 7.233 sum_up_and_integrate 110 10.3 0.053 0.059 7.034 7.046 integrate_v_rspace 110 11.3 0.002 0.002 6.981 6.994 qs_ot_get_p 110 10.4 0.001 0.001 6.363 6.542 init_scf_run 11 5.9 0.000 0.001 6.368 6.369 scf_env_initial_rho_setup 11 6.9 0.001 0.005 6.368 6.369 dbcsr_mm_accdrv_process 52282 16.1 4.696 5.414 6.052 6.353 prepare_preconditioner 11 7.9 0.000 0.000 6.287 6.306 make_preconditioner 11 8.9 0.000 0.000 6.287 6.306 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.252 6.259 calculate_rho_elec 110 8.6 0.039 0.047 6.252 6.259 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.842 6.008 ot_diis_step 99 11.5 0.010 0.011 5.748 5.749 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.845 5.588 make_m2s 4110 13.4 0.057 0.060 4.610 5.058 make_images 4110 14.4 0.396 0.439 4.499 4.943 qs_ot_p2m_diag 48 11.0 0.028 0.044 4.282 4.303 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.818 3.819 rs_pw_transfer 902 11.9 0.012 0.013 3.585 3.806 pw_transfer 1331 11.6 0.066 0.072 3.636 3.791 density_rs2pw 110 9.6 0.004 0.004 3.511 3.751 wfi_extrapolate 11 7.9 0.001 0.001 3.715 3.715 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.529 3.682 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.334 3.406 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.352 3.355 grid_integrate_task_list 110 12.3 3.142 3.350 3.142 3.350 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.184 3.187 cp_fm_redistribute_end 48 14.0 2.365 3.144 2.368 3.146 cp_fm_diag_elpa_base 48 14.0 0.742 3.008 0.773 3.095 calculate_dm_sparse 110 9.5 0.001 0.001 2.998 3.028 fft3d_ps 1111 14.6 1.074 1.281 2.888 3.027 make_images_data 4110 15.4 0.047 0.051 2.536 3.019 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.946 3.000 fft_wrap_pw1pw2_140 451 13.1 0.198 0.217 2.830 2.980 cp_fm_cholesky_invert 11 10.9 2.939 2.947 2.939 2.947 hybrid_alltoall_any 4261 16.3 0.102 0.443 2.224 2.942 potential_pw2rs 110 12.3 0.008 0.009 2.832 2.861 mp_sum_l 6514 12.8 1.956 2.584 1.956 2.584 calculate_first_density_matrix 1 7.0 0.000 0.003 2.553 2.557 mp_alltoall_d11v 2046 13.8 2.027 2.359 2.027 2.359 grid_collocate_task_list 110 9.6 2.117 2.247 2.117 2.247 jit_kernel_multiply 11 16.3 1.006 2.135 1.006 2.135 qs_ot_get_orbitals 99 10.5 0.001 0.001 2.006 2.025 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.885 1.885 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.844 1.856 multiply_cannon_metrocomm4 22605 15.4 0.073 0.076 0.796 1.823 mp_allgather_i34 2055 14.4 0.727 1.786 0.727 1.786 mp_irecv_dv 57340 16.2 0.673 1.706 0.673 1.706 cp_fm_cholesky_decompose 22 10.9 1.615 1.621 1.615 1.621 mp_alltoall_z22v 1111 16.6 1.530 1.621 1.530 1.621 dbcsr_complete_redistribute 325 12.2 0.240 0.301 1.341 1.618 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.557 1.567 acc_transpose_blocks 24660 15.4 0.108 0.111 1.533 1.553 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.368 1.475 mp_waitany 10164 13.8 1.265 1.453 1.265 1.453 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=71.319000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=556.181818, yerr=6.364935 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d60029441ac03bb22e0b17392546c4f4f2d728f9_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 656.293888E+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 0.184 0.552 65.308 65.309 qs_mol_dyn_low 1 2.0 0.003 0.003 63.898 64.339 qs_forces 11 3.9 0.004 0.004 63.812 63.813 qs_energies 11 4.9 0.019 0.048 60.552 60.559 scf_env_do_scf 11 5.9 0.001 0.001 52.053 52.053 scf_env_do_scf_inner_loop 99 6.5 0.005 0.022 42.227 42.229 velocity_verlet 10 3.0 0.001 0.001 35.010 35.011 dbcsr_multiply_generic 2055 12.4 0.111 0.120 29.994 30.438 qs_scf_new_mos 99 7.5 0.001 0.001 27.038 27.139 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.037 27.138 ot_scf_mini 99 9.5 0.003 0.003 25.748 25.866 multiply_cannon 2055 13.4 0.211 0.231 22.742 23.897 multiply_cannon_loop 2055 14.4 0.620 0.639 21.324 22.474 ot_mini 99 10.5 0.001 0.001 14.519 14.650 rebuild_ks_matrix 110 8.3 0.000 0.000 12.639 12.848 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.638 12.848 mp_waitall_1 146766 16.3 8.159 11.548 8.159 11.548 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.187 11.383 multiply_cannon_multrec 16440 15.4 3.803 4.863 9.551 10.531 qs_ot_get_derivative 99 11.5 0.001 0.001 9.906 10.025 init_scf_loop 11 6.9 0.001 0.002 9.778 9.781 prepare_preconditioner 11 7.9 0.000 0.000 7.935 7.956 make_preconditioner 11 8.9 0.000 0.001 7.935 7.956 multiply_cannon_metrocomm3 16440 15.4 0.043 0.045 4.601 7.723 make_full_inverse_cholesky 11 9.9 0.000 0.000 7.234 7.602 sum_up_and_integrate 110 10.3 0.061 0.061 6.752 6.769 integrate_v_rspace 110 11.3 0.002 0.002 6.691 6.708 qs_ot_get_p 110 10.4 0.001 0.001 6.204 6.367 init_scf_run 11 5.9 0.000 0.001 5.973 5.973 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.973 5.973 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.861 5.881 calculate_rho_elec 110 8.6 0.058 0.058 5.860 5.881 dbcsr_mm_accdrv_process 34862 16.1 4.635 5.192 5.606 5.790 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.099 5.587 apply_single 110 13.6 0.000 0.000 5.099 5.587 make_m2s 4110 13.4 0.051 0.052 4.691 5.216 make_images 4110 14.4 0.393 0.534 4.574 5.100 ot_diis_step 99 11.5 0.011 0.011 4.557 4.557 qs_ot_p2m_diag 48 11.0 0.042 0.044 4.298 4.304 multiply_cannon_sync_h2d 16440 15.4 3.704 4.274 3.704 4.274 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.348 4.008 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.921 3.921 cp_fm_cholesky_invert 11 10.9 3.457 3.466 3.457 3.466 grid_integrate_task_list 110 12.3 3.202 3.442 3.202 3.442 rs_pw_transfer 902 11.9 0.010 0.011 3.142 3.400 pw_transfer 1331 11.6 0.065 0.071 3.351 3.364 make_images_data 4110 15.4 0.044 0.048 2.712 3.355 density_rs2pw 110 9.6 0.004 0.004 3.099 3.333 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.302 3.306 cp_fm_redistribute_end 48 14.0 2.048 3.265 2.050 3.267 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.244 3.260 cp_fm_diag_elpa_base 48 14.0 1.146 3.116 1.207 3.223 wfi_extrapolate 11 7.9 0.001 0.001 3.210 3.210 hybrid_alltoall_any 4261 16.3 0.105 0.377 2.339 3.194 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.128 3.129 mp_sum_l 6514 12.8 2.037 2.831 2.037 2.831 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.728 2.786 fft_wrap_pw1pw2_140 451 13.1 0.208 0.213 2.668 2.683 calculate_first_density_matrix 1 7.0 0.001 0.004 2.676 2.678 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.595 2.655 calculate_dm_sparse 110 9.5 0.001 0.001 2.577 2.615 multiply_cannon_metrocomm4 14385 15.4 0.044 0.048 0.880 2.590 fft3d_ps 1111 14.6 1.065 1.077 2.561 2.577 potential_pw2rs 110 12.3 0.011 0.011 2.534 2.543 mp_irecv_dv 48980 15.7 0.811 2.461 0.811 2.461 mp_alltoall_d11v 2046 13.8 1.984 2.439 1.984 2.439 grid_collocate_task_list 110 9.6 2.172 2.374 2.172 2.374 dbcsr_complete_redistribute 325 12.2 0.323 0.352 1.615 2.069 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.006 2.008 mp_allgather_i34 2055 14.4 0.743 1.995 0.743 1.995 cp_fm_cholesky_decompose 22 10.9 1.914 1.932 1.914 1.932 jit_kernel_multiply 10 16.7 0.584 1.908 0.584 1.908 cp_fm_upper_to_full 70 13.6 1.400 1.871 1.400 1.871 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.680 1.703 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.047 1.504 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.483 1.497 mp_waitany 17072 13.8 1.246 1.492 1.246 1.492 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.367 1.489 acc_transpose_blocks 16440 15.4 0.071 0.073 1.413 1.423 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.389 1.405 rs_gather_matrices 110 12.3 0.138 0.150 0.903 1.355 mp_alltoall_z22v 1111 16.6 1.283 1.309 1.283 1.309 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=65.309000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=621.818182, yerr=8.407985 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d60029441ac03bb22e0b17392546c4f4f2d728f9_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 732.536832E+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 0.019 0.035 66.732 66.733 qs_mol_dyn_low 1 2.0 0.003 0.003 66.429 66.440 qs_forces 11 3.9 0.005 0.009 66.358 66.360 qs_energies 11 4.9 0.001 0.004 62.931 62.939 scf_env_do_scf 11 5.9 0.001 0.001 54.332 54.335 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 41.879 41.880 velocity_verlet 10 3.0 0.001 0.001 38.294 38.296 dbcsr_multiply_generic 2055 12.4 0.116 0.124 29.995 30.244 qs_scf_new_mos 99 7.5 0.001 0.001 26.773 26.876 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.773 26.875 ot_scf_mini 99 9.5 0.003 0.003 25.114 25.221 multiply_cannon 2055 13.4 0.246 0.272 22.443 23.810 multiply_cannon_loop 2055 14.4 0.891 0.904 20.938 21.551 ot_mini 99 10.5 0.001 0.001 14.325 14.457 multiply_cannon_multrec 24660 15.4 4.182 6.740 12.598 13.826 rebuild_ks_matrix 110 8.3 0.000 0.000 12.352 12.467 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.352 12.467 init_scf_loop 11 6.9 0.000 0.002 12.401 12.402 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.942 11.043 prepare_preconditioner 11 7.9 0.000 0.000 10.610 10.625 make_preconditioner 11 8.9 0.000 0.001 10.610 10.625 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.745 10.262 qs_ot_get_derivative 99 11.5 0.001 0.001 10.144 10.258 dbcsr_mm_accdrv_process 52304 16.0 6.793 8.144 8.269 9.204 mp_waitall_1 126806 16.4 5.073 6.915 5.073 6.915 sum_up_and_integrate 110 10.3 0.067 0.070 6.642 6.659 integrate_v_rspace 110 11.3 0.002 0.003 6.575 6.588 make_m2s 4110 13.4 0.060 0.062 5.830 6.210 make_images 4110 14.4 0.572 0.690 5.687 6.064 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.007 6.018 calculate_rho_elec 110 8.6 0.077 0.081 6.007 6.017 qs_ot_get_p 110 10.4 0.001 0.001 5.659 5.784 init_scf_run 11 5.9 0.000 0.001 5.707 5.708 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.707 5.707 cp_fm_upper_to_full 70 13.8 3.371 4.808 3.371 4.808 ot_diis_step 99 11.5 0.011 0.011 4.143 4.144 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.033 4.123 apply_single 110 13.6 0.000 0.000 4.032 4.123 dbcsr_complete_redistribute 325 12.2 0.416 0.481 2.853 4.030 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.711 3.725 pw_transfer 1331 11.6 0.064 0.073 3.552 3.577 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.446 3.474 make_images_data 4110 15.4 0.046 0.050 3.000 3.453 grid_integrate_task_list 110 12.3 3.273 3.452 3.273 3.452 multiply_cannon_sync_h2d 24660 15.4 3.170 3.433 3.170 3.433 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.261 3.419 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.366 3.416 hybrid_alltoall_any 4261 16.3 0.119 0.456 2.503 3.411 multiply_cannon_metrocomm3 24660 15.4 0.036 0.037 1.432 3.265 density_rs2pw 110 9.6 0.004 0.004 3.103 3.222 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.219 3.219 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.111 3.171 wfi_extrapolate 11 7.9 0.001 0.001 3.129 3.129 cp_fm_cholesky_invert 11 10.9 3.035 3.045 3.035 3.045 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.857 3.004 rs_pw_transfer 902 11.9 0.010 0.011 2.854 2.999 calculate_dm_sparse 110 9.5 0.001 0.001 2.963 2.996 mp_alltoall_i22 605 13.7 1.712 2.934 1.712 2.934 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.908 2.911 fft_wrap_pw1pw2_140 451 13.1 0.201 0.212 2.852 2.881 fft3d_ps 1111 14.6 1.067 1.102 2.751 2.766 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.612 2.616 cp_fm_redistribute_end 48 14.0 1.302 2.582 1.303 2.584 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.497 2.552 cp_fm_diag_elpa_base 48 14.0 1.194 2.439 1.274 2.550 calculate_first_density_matrix 1 7.0 0.000 0.000 2.489 2.492 grid_collocate_task_list 110 9.6 2.223 2.389 2.223 2.389 potential_pw2rs 110 12.3 0.012 0.013 2.367 2.381 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.349 2.353 mp_alltoall_d11v 2046 13.8 2.042 2.257 2.042 2.257 jit_kernel_multiply 12 15.7 1.148 2.189 1.148 2.189 cp_fm_cholesky_decompose 22 10.9 1.873 1.935 1.873 1.935 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.746 1.775 mp_allgather_i34 2055 14.4 0.610 1.765 0.610 1.765 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.599 1.699 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.657 1.669 mp_sum_l 6514 12.8 1.038 1.590 1.038 1.590 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.551 1.569 multiply_cannon_metrocomm4 20550 15.4 0.056 0.059 0.863 1.533 mp_alltoall_z22v 1111 16.6 1.468 1.528 1.468 1.528 acc_transpose_blocks 24660 15.4 0.102 0.104 1.488 1.506 mp_irecv_dv 62702 16.1 0.766 1.454 0.766 1.454 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.291 1.410 mp_waitany 13376 13.8 1.158 1.347 1.158 1.347 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=66.733000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=693.090909, yerr=11.587254 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d60029441ac03bb22e0b17392546c4f4f2d728f9_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 834.899968E+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 0.016 0.033 57.980 57.981 qs_mol_dyn_low 1 2.0 0.003 0.004 57.632 57.643 qs_forces 11 3.9 0.002 0.002 57.563 57.564 qs_energies 11 4.9 0.001 0.001 53.896 53.900 scf_env_do_scf 11 5.9 0.000 0.001 45.124 45.124 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 36.394 36.395 velocity_verlet 10 3.0 0.001 0.001 32.214 32.217 dbcsr_multiply_generic 2055 12.4 0.106 0.108 24.024 24.221 qs_scf_new_mos 99 7.5 0.001 0.001 21.878 21.938 qs_scf_loop_do_ot 99 8.5 0.001 0.001 21.877 21.938 ot_scf_mini 99 9.5 0.002 0.002 20.599 20.640 multiply_cannon 2055 13.4 0.250 0.260 18.174 19.429 multiply_cannon_loop 2055 14.4 0.324 0.337 16.716 17.044 rebuild_ks_matrix 110 8.3 0.000 0.000 11.851 11.910 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 11.850 11.909 ot_mini 99 10.5 0.001 0.001 11.080 11.121 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.574 10.626 multiply_cannon_multrec 8220 15.4 3.296 5.077 7.858 9.122 init_scf_loop 11 6.9 0.000 0.000 8.672 8.673 mp_waitall_1 106626 16.5 6.745 8.458 6.745 8.458 qs_ot_get_derivative 99 11.5 0.001 0.001 7.133 7.176 prepare_preconditioner 11 7.9 0.000 0.000 6.869 6.876 make_preconditioner 11 8.9 0.000 0.000 6.869 6.876 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.429 6.503 sum_up_and_integrate 110 10.3 0.080 0.082 6.413 6.467 integrate_v_rspace 110 11.3 0.003 0.003 6.333 6.384 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.824 5.840 calculate_rho_elec 110 8.6 0.115 0.115 5.824 5.839 init_scf_run 11 5.9 0.000 0.001 5.618 5.618 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.618 5.618 qs_ot_get_p 110 10.4 0.001 0.001 5.488 5.532 dbcsr_mm_accdrv_process 17442 15.9 3.037 4.082 4.432 5.491 make_m2s 4110 13.4 0.039 0.040 4.436 4.687 make_images 4110 14.4 0.635 0.691 4.305 4.553 multiply_cannon_metrocomm3 8220 15.4 0.017 0.018 3.244 4.541 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.999 4.003 ot_diis_step 99 11.5 0.012 0.013 3.918 3.918 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.795 3.859 apply_single 110 13.6 0.000 0.000 3.795 3.859 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.667 3.668 cp_fm_cholesky_invert 11 10.9 3.494 3.497 3.494 3.497 grid_integrate_task_list 110 12.3 3.389 3.493 3.389 3.493 pw_transfer 1331 11.6 0.065 0.070 3.431 3.475 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.323 3.368 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.085 3.087 cp_fm_redistribute_end 48 14.0 0.781 3.047 0.787 3.047 multiply_cannon_sync_h2d 8220 15.4 2.895 3.033 2.895 3.033 cp_fm_diag_elpa_base 48 14.0 2.070 2.835 2.248 3.003 density_rs2pw 110 9.6 0.004 0.004 2.852 2.970 make_images_data 4110 15.4 0.039 0.043 2.501 2.918 calculate_dm_sparse 110 9.5 0.001 0.001 2.826 2.870 fft_wrap_pw1pw2_140 451 13.1 0.212 0.216 2.798 2.848 wfi_extrapolate 11 7.9 0.001 0.001 2.829 2.829 hybrid_alltoall_any 4261 16.3 0.200 0.859 2.400 2.806 calculate_first_density_matrix 1 7.0 0.000 0.000 2.699 2.699 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.683 2.684 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.664 2.665 fft3d_ps 1111 14.6 1.120 1.174 2.582 2.626 rs_pw_transfer 902 11.9 0.010 0.010 2.409 2.525 grid_collocate_task_list 110 9.6 2.329 2.415 2.329 2.415 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.300 2.314 potential_pw2rs 110 12.3 0.015 0.016 2.108 2.142 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.068 2.092 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.771 1.995 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.963 1.983 mp_alltoall_d11v 2046 13.8 1.798 1.944 1.798 1.944 cp_fm_cholesky_decompose 22 10.9 1.882 1.901 1.882 1.901 mp_allgather_i34 2055 14.4 0.642 1.779 0.642 1.779 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.770 1.778 dbcsr_complete_redistribute 325 12.2 0.556 0.582 1.591 1.691 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.508 1.627 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.609 1.616 jit_kernel_multiply 8 15.8 1.087 1.486 1.087 1.486 multiply_cannon_metrocomm4 6165 15.4 0.017 0.019 0.475 1.412 mp_irecv_dv 24056 15.7 0.452 1.371 0.452 1.371 qs_create_task_list 11 7.9 0.000 0.001 1.230 1.329 generate_qs_task_list 11 8.9 0.379 0.446 1.230 1.329 mp_alltoall_z22v 1111 16.6 1.228 1.269 1.228 1.269 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.240 1.265 mp_waitany 9240 13.8 1.103 1.224 1.103 1.224 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=57.981000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=780.363636, yerr=13.956248 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d60029441ac03bb22e0b17392546c4f4f2d728f9_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 1.393746E+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 0.022 0.040 90.172 90.172 qs_mol_dyn_low 1 2.0 0.003 0.003 89.621 89.631 qs_forces 11 3.9 0.026 0.032 89.543 89.544 qs_energies 11 4.9 0.001 0.001 85.422 85.446 scf_env_do_scf 11 5.9 0.001 0.001 75.300 75.300 velocity_verlet 10 3.0 0.001 0.001 57.022 57.029 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 45.719 45.720 dbcsr_multiply_generic 2055 12.4 0.120 0.123 29.880 30.004 init_scf_loop 11 6.9 0.000 0.000 29.495 29.497 qs_scf_new_mos 99 7.5 0.001 0.001 28.688 28.744 qs_scf_loop_do_ot 99 8.5 0.001 0.001 28.687 28.744 prepare_preconditioner 11 7.9 0.000 0.000 27.490 27.499 make_preconditioner 11 8.9 0.000 0.000 27.490 27.499 ot_scf_mini 99 9.5 0.002 0.002 26.885 26.949 make_full_inverse_cholesky 11 9.9 0.000 0.000 21.654 26.948 multiply_cannon 2055 13.4 0.354 0.384 22.234 23.021 multiply_cannon_loop 2055 14.4 0.343 0.347 20.312 20.631 cp_fm_upper_to_full 70 14.2 12.783 18.332 12.783 18.332 ot_mini 99 10.5 0.001 0.001 14.780 14.835 rebuild_ks_matrix 110 8.3 0.001 0.001 13.545 13.613 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 13.545 13.612 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.308 12.369 dbcsr_complete_redistribute 325 12.2 1.018 1.059 7.571 10.798 qs_ot_get_derivative 99 11.5 0.001 0.001 10.105 10.164 multiply_cannon_multrec 8220 15.4 4.386 4.566 9.801 10.143 mp_waitall_1 87304 16.6 8.664 9.935 8.664 9.935 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.457 9.678 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.821 9.019 mp_alltoall_i22 605 13.7 5.456 8.655 5.456 8.655 sum_up_and_integrate 110 10.3 0.150 0.152 6.939 6.951 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.848 6.881 calculate_rho_elec 110 8.6 0.227 0.227 6.847 6.880 qs_ot_get_p 110 10.4 0.001 0.001 6.775 6.837 integrate_v_rspace 110 11.3 0.003 0.003 6.788 6.802 make_m2s 4110 13.4 0.043 0.043 5.671 6.237 cp_fm_cholesky_invert 11 10.9 6.067 6.072 6.067 6.072 make_images 4110 14.4 0.878 0.931 5.481 6.046 init_scf_run 11 5.9 0.000 0.001 5.827 5.827 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.827 5.827 dbcsr_mm_accdrv_process 11614 15.7 3.442 4.125 5.272 5.742 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 5.157 5.742 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.728 5.230 apply_single 110 13.6 0.000 0.000 4.728 5.230 qs_ot_p2m_diag 48 11.0 0.151 0.155 4.993 5.000 ot_diis_step 99 11.5 0.016 0.016 4.641 4.641 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.527 4.528 pw_transfer 1331 11.6 0.073 0.074 4.129 4.135 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 4.013 4.019 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.522 3.981 multiply_cannon_sync_h2d 8220 15.4 3.955 3.964 3.955 3.964 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.842 3.843 hybrid_alltoall_any 4261 16.3 0.258 0.564 3.085 3.841 cp_fm_diag_elpa_base 48 14.0 3.292 3.499 3.840 3.841 make_images_data 4110 15.4 0.042 0.045 3.119 3.834 grid_integrate_task_list 110 12.3 3.671 3.723 3.671 3.723 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.664 3.688 fft_wrap_pw1pw2_140 451 13.1 0.214 0.217 3.395 3.406 wfi_extrapolate 11 7.9 0.001 0.001 3.401 3.401 density_rs2pw 110 9.6 0.004 0.004 3.290 3.308 fft3d_ps 1111 14.6 1.271 1.285 3.222 3.232 calculate_dm_sparse 110 9.5 0.001 0.001 3.176 3.201 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.944 2.947 grid_collocate_task_list 110 9.6 2.639 2.678 2.639 2.678 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.621 2.656 rs_pw_transfer 902 11.9 0.010 0.010 2.392 2.434 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.379 2.406 mp_alltoall_d11v 2046 13.8 2.303 2.362 2.303 2.362 calculate_first_density_matrix 1 7.0 0.000 0.000 2.310 2.311 potential_pw2rs 110 12.3 0.021 0.021 2.290 2.302 cp_fm_cholesky_decompose 22 10.9 2.284 2.300 2.284 2.300 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.201 2.263 jit_kernel_multiply 10 15.3 1.630 2.237 1.630 2.237 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.090 2.151 qs_create_task_list 11 7.9 0.001 0.001 1.907 1.951 generate_qs_task_list 11 8.9 0.742 0.793 1.907 1.950 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.893 1.924 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.842 1.848 mp_allgather_i34 2055 14.4 0.700 1.833 0.700 1.833 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=90.172000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1251.272727, yerr=66.723433 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d60029441ac03bb22e0b17392546c4f4f2d728f9_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 628.129792E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175955383376 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4085 56760. MP_Allreduce 11253 785. MP_Sync 170 MP_Alltoall 2226 2824831. MP_ISendRecv 48640 18752. MP_Wait 66796 MP_comm_split 83 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.023 0.059 210.518 210.519 qs_mol_dyn_low 1 2.0 0.003 0.004 210.005 210.052 qs_forces 11 3.9 0.003 0.004 209.910 209.911 qs_energies 11 4.9 0.001 0.002 204.283 204.301 scf_env_do_scf 11 5.9 0.001 0.001 187.094 187.097 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 164.504 164.506 velocity_verlet 10 3.0 0.001 0.001 126.511 126.512 dbcsr_multiply_generic 2507 12.6 0.178 0.181 125.793 126.500 qs_scf_new_mos 117 7.6 0.001 0.001 124.718 124.975 qs_scf_loop_do_ot 117 8.6 0.001 0.001 124.717 124.975 ot_scf_mini 117 9.6 0.003 0.003 118.084 118.295 multiply_cannon 2507 13.6 0.238 0.244 101.621 104.078 multiply_cannon_loop 2507 14.6 2.140 2.218 99.267 101.801 ot_mini 117 10.6 0.001 0.001 66.869 67.123 multiply_cannon_multrec 60168 15.6 33.080 34.881 41.752 43.868 qs_ot_get_derivative 117 11.6 0.001 0.001 42.007 42.239 rebuild_ks_matrix 128 8.3 0.001 0.001 33.850 34.118 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 33.849 34.117 mp_waitall_1 291448 16.2 30.016 32.765 30.016 32.765 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.459 30.709 qs_ot_get_p 128 10.4 0.001 0.001 29.269 29.547 multiply_cannon_sync_h2d 60168 15.6 27.195 28.823 27.195 28.823 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.300 25.079 apply_single 128 13.6 0.001 0.001 24.300 25.079 ot_diis_step 117 11.6 0.008 0.008 24.506 24.507 init_scf_loop 11 6.9 0.000 0.000 22.502 22.503 qs_ot_p2m_diag 83 11.4 0.077 0.091 22.317 22.385 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.516 19.671 cp_dbcsr_syevd 83 12.4 0.004 0.005 19.615 19.617 multiply_cannon_metrocomm3 60168 15.6 0.116 0.124 15.869 18.355 prepare_preconditioner 11 7.9 0.000 0.000 17.836 17.867 make_preconditioner 11 8.9 0.000 0.000 17.836 17.867 make_full_inverse_cholesky 11 9.9 0.000 0.000 17.066 17.270 cp_fm_diag_elpa 83 13.4 0.000 0.001 16.402 16.410 cp_fm_redistribute_end 83 14.4 12.911 16.294 12.925 16.300 cp_fm_diag_elpa_base 83 14.4 3.319 16.002 3.357 16.128 sum_up_and_integrate 128 10.3 0.089 0.107 14.619 14.636 integrate_v_rspace 128 11.3 0.003 0.004 14.530 14.549 make_m2s 5014 13.6 0.104 0.110 14.170 14.533 make_images 5014 14.6 0.403 0.426 13.987 14.359 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.332 13.438 calculate_rho_elec 128 8.7 0.045 0.063 13.332 13.437 init_scf_run 11 5.9 0.000 0.001 12.977 12.978 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.977 12.977 cp_fm_cholesky_invert 11 10.9 10.298 10.309 10.298 10.309 mp_sum_l 7870 13.0 8.745 10.015 8.745 10.015 wfi_extrapolate 11 7.9 0.001 0.001 9.376 9.376 dbcsr_mm_accdrv_process 124484 16.2 3.316 3.519 8.240 8.802 calculate_dm_sparse 128 9.5 0.001 0.001 8.611 8.719 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.220 8.397 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.954 8.040 multiply_cannon_metrocomm1 60168 15.6 0.086 0.089 6.324 7.835 pw_transfer 1547 11.6 0.075 0.107 7.224 7.566 make_images_data 5014 15.6 0.068 0.073 6.759 7.563 density_rs2pw 128 9.7 0.006 0.007 7.153 7.551 grid_integrate_task_list 128 12.3 7.080 7.493 7.080 7.493 fft_wrap_pw1pw2 1291 12.7 0.010 0.014 7.020 7.335 rs_pw_transfer 1046 11.9 0.016 0.018 6.584 7.085 hybrid_alltoall_any 5200 16.5 0.293 2.274 5.907 6.968 cp_dbcsr_sm_fm_multiply 37 9.5 0.003 0.003 6.884 6.891 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.992 6.113 fft3d_ps 1291 14.7 2.096 2.573 5.819 6.070 fft_wrap_pw1pw2_140 523 13.2 0.439 0.508 5.611 5.816 mp_alltoall_d11v 2415 14.1 4.249 5.613 4.249 5.613 cp_fm_cholesky_decompose 22 10.9 5.342 5.358 5.342 5.358 potential_pw2rs 128 12.3 0.009 0.011 5.127 5.170 grid_collocate_task_list 128 9.7 4.726 5.122 4.726 5.122 mp_sum_d 4460 12.2 3.656 4.315 3.656 4.315 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=210.519000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=594.545455, yerr=6.638231 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d60029441ac03bb22e0b17392546c4f4f2d728f9_performance_tests/16/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 2.166472E+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 5925696 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.6 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 826.392576E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2385600 MPI messages size (bytes): total size 4.069300E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.705776E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70188 2295595008 32768 < size <= 131072 716032 54973693952 131072 < size <= 4194304 1363760 1386318135296 4194304 < size <= 16777216 153648 1453842923456 16777216 < size 67056 1171888537600 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4089 57145. MP_Allreduce 11262 946. MP_Sync 168 MP_Alltoall 1955 4550726. MP_ISendRecv 23876 47072. MP_Wait 37656 MP_comm_split 82 MP_ISend 11660 212488. MP_IRecv 11660 212488. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.029 0.058 194.367 194.372 qs_mol_dyn_low 1 2.0 0.003 0.004 193.892 193.915 qs_forces 11 3.9 0.003 0.003 193.794 193.796 qs_energies 11 4.9 0.002 0.005 186.981 186.995 scf_env_do_scf 11 5.9 0.001 0.005 170.301 170.311 scf_env_do_scf_inner_loop 116 6.6 0.009 0.052 136.136 136.139 velocity_verlet 10 3.0 0.001 0.001 121.014 121.019 dbcsr_multiply_generic 2485 12.5 0.186 0.192 97.626 98.687 qs_scf_new_mos 116 7.6 0.001 0.001 96.714 97.191 qs_scf_loop_do_ot 116 8.6 0.005 0.035 96.713 97.190 ot_scf_mini 116 9.6 0.007 0.030 91.922 92.487 multiply_cannon 2485 13.5 0.478 0.522 77.090 81.546 multiply_cannon_loop 2485 14.5 1.243 1.284 73.709 76.058 ot_mini 116 10.6 0.001 0.001 50.037 50.540 mp_waitall_1 224796 16.4 25.596 38.759 25.596 38.759 multiply_cannon_multrec 29820 15.5 21.988 26.204 31.560 36.344 init_scf_loop 11 6.9 0.001 0.005 34.070 34.072 rebuild_ks_matrix 127 8.3 0.001 0.001 33.245 33.797 qs_ks_build_kohn_sham_matrix 127 9.3 0.021 0.042 33.244 33.796 qs_ks_update_qs_env 127 7.6 0.001 0.001 29.926 30.425 prepare_preconditioner 11 7.9 0.000 0.000 29.462 29.554 make_preconditioner 11 8.9 0.000 0.003 29.462 29.554 qs_ot_get_derivative 116 11.6 0.001 0.002 28.310 28.871 make_full_inverse_cholesky 11 9.9 0.000 0.001 28.152 28.687 multiply_cannon_metrocomm3 29820 15.5 0.092 0.097 15.435 28.178 qs_ot_get_p 127 10.4 0.001 0.002 23.780 24.457 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 21.820 22.890 apply_single 127 13.6 0.001 0.001 21.820 22.890 multiply_cannon_sync_h2d 29820 15.5 19.337 22.030 19.337 22.030 ot_diis_step 116 11.6 0.014 0.015 21.551 21.553 qs_ot_p2m_diag 82 11.4 0.185 0.213 18.642 18.676 cp_dbcsr_syevd 82 12.4 0.005 0.005 17.453 17.454 cp_fm_cholesky_invert 11 10.9 17.210 17.224 17.210 17.224 make_m2s 4970 13.5 0.089 0.092 14.672 16.224 make_images 4970 14.5 1.137 1.324 14.463 16.014 sum_up_and_integrate 127 10.3 0.116 0.133 14.798 14.829 integrate_v_rspace 127 11.3 0.003 0.003 14.682 14.720 cp_fm_diag_elpa 82 13.4 0.000 0.001 14.202 14.211 cp_fm_redistribute_end 82 14.4 8.255 14.122 8.267 14.123 cp_fm_diag_elpa_base 82 14.4 5.610 13.657 5.835 14.015 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.464 13.498 calculate_rho_elec 127 8.7 0.087 0.104 13.463 13.498 init_scf_run 11 5.9 0.000 0.001 11.738 11.740 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.738 11.740 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 11.060 11.491 make_images_data 4970 15.5 0.064 0.072 8.919 10.755 multiply_cannon_metrocomm4 27335 15.5 0.096 0.111 3.673 10.532 mp_irecv_dv 68888 16.3 3.482 10.144 3.482 10.144 hybrid_alltoall_any 5155 16.4 0.341 1.498 7.574 10.050 dbcsr_mm_accdrv_process 61726 16.2 4.460 5.234 9.029 9.645 wfi_extrapolate 11 7.9 0.001 0.001 8.557 8.557 pw_transfer 1535 11.6 0.084 0.102 8.075 8.148 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 7.852 7.916 density_rs2pw 127 9.7 0.006 0.006 7.219 7.572 grid_integrate_task_list 127 12.3 7.110 7.568 7.110 7.568 cp_fm_cholesky_decompose 22 10.9 7.301 7.384 7.301 7.384 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.460 7.136 fft_wrap_pw1pw2_140 519 13.2 0.463 0.508 6.724 6.784 rs_pw_transfer 1038 11.9 0.014 0.016 6.363 6.721 calculate_dm_sparse 127 9.5 0.001 0.001 6.447 6.599 fft3d_ps 1281 14.7 2.756 2.907 6.242 6.280 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.226 6.236 mp_sum_l 7804 13.0 4.112 5.808 4.112 5.808 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.300 5.475 qs_ot_get_orbitals 116 10.6 0.001 0.001 5.314 5.375 grid_collocate_task_list 127 9.7 4.871 5.356 4.871 5.356 potential_pw2rs 127 12.3 0.015 0.017 5.283 5.321 mp_allgather_i34 2485 14.5 1.962 5.193 1.962 5.193 mp_alltoall_d11v 2401 14.1 4.392 5.169 4.392 5.169 dbcsr_complete_redistribute 393 12.7 0.771 0.884 3.328 4.214 mp_sum_d 4444 12.2 2.747 4.012 2.747 4.012 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=194.372000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=788.000000, yerr=1.279204 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d60029441ac03bb22e0b17392546c4f4f2d728f9_performance_tests/17/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 2.906045E+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 3951168 0.0% 0.0% 100.0% average stack size 0.0 0.0 1697.1 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 936.992768E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1033760 MPI messages size (bytes): total size 2.695213E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.607194E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 279168 36591108096 131072 < size <= 4194304 654272 987691483136 4194304 < size <= 16777216 65184 925172769472 16777216 < size 28448 745747251200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4089 57139. MP_Allreduce 11261 985. MP_Sync 168 MP_Alltoall 1700 9383497. MP_ISendRecv 15748 75008. MP_Wait 29528 MP_comm_split 82 MP_ISend 11660 275234. MP_IRecv 11660 275234. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.074 0.198 181.393 181.395 qs_mol_dyn_low 1 2.0 0.003 0.005 180.689 180.719 qs_forces 11 3.9 0.009 0.018 180.590 180.594 qs_energies 11 4.9 0.021 0.042 174.015 174.035 scf_env_do_scf 11 5.9 0.003 0.013 157.869 157.871 scf_env_do_scf_inner_loop 116 6.6 0.009 0.025 121.035 121.038 velocity_verlet 10 3.0 0.002 0.011 114.764 114.766 dbcsr_multiply_generic 2485 12.5 0.183 0.189 83.509 84.620 qs_scf_new_mos 116 7.6 0.001 0.001 84.084 84.490 qs_scf_loop_do_ot 116 8.6 0.002 0.007 84.083 84.489 ot_scf_mini 116 9.6 0.004 0.009 79.941 80.371 multiply_cannon 2485 13.5 0.493 0.536 62.126 66.302 multiply_cannon_loop 2485 14.5 0.851 0.880 59.041 62.105 ot_mini 116 10.6 0.001 0.001 44.507 44.950 init_scf_loop 11 6.9 0.015 0.115 36.728 36.728 mp_waitall_1 176908 16.5 27.070 36.105 27.070 36.105 prepare_preconditioner 11 7.9 0.000 0.000 32.469 32.524 make_preconditioner 11 8.9 0.000 0.002 32.469 32.524 make_full_inverse_cholesky 11 9.9 0.000 0.000 30.061 31.490 rebuild_ks_matrix 127 8.3 0.001 0.001 30.299 30.821 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.018 30.299 30.820 qs_ks_update_qs_env 127 7.6 0.001 0.001 27.349 27.830 multiply_cannon_metrocomm3 19880 15.5 0.060 0.064 16.140 25.271 multiply_cannon_multrec 19880 15.5 13.185 16.670 21.912 25.202 qs_ot_get_derivative 116 11.6 0.001 0.002 24.606 25.038 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 20.805 22.243 apply_single 127 13.6 0.001 0.001 20.805 22.243 qs_ot_get_p 127 10.4 0.001 0.001 20.827 21.313 ot_diis_step 116 11.6 0.018 0.018 19.787 19.788 qs_ot_p2m_diag 82 11.4 0.262 0.269 16.290 16.303 make_m2s 4970 13.5 0.081 0.086 15.300 16.092 make_images 4970 14.5 1.163 1.265 15.067 15.861 multiply_cannon_sync_h2d 19880 15.5 14.010 15.705 14.010 15.705 cp_fm_cholesky_invert 11 10.9 15.394 15.404 15.394 15.404 cp_dbcsr_syevd 82 12.4 0.005 0.005 15.214 15.216 sum_up_and_integrate 127 10.3 0.132 0.145 14.437 14.461 integrate_v_rspace 127 11.3 0.003 0.004 14.305 14.331 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.311 13.348 calculate_rho_elec 127 8.7 0.130 0.145 13.310 13.347 cp_fm_diag_elpa 82 13.4 0.000 0.001 11.938 11.942 cp_fm_redistribute_end 82 14.4 4.494 11.857 4.511 11.860 cp_fm_diag_elpa_base 82 14.4 6.926 11.256 7.326 11.734 init_scf_run 11 5.9 0.000 0.001 11.009 11.010 scf_env_initial_rho_setup 11 6.9 0.004 0.027 11.009 11.010 make_images_data 4970 15.5 0.060 0.067 9.483 10.735 hybrid_alltoall_any 5155 16.4 0.428 1.969 8.215 9.727 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.191 9.501 multiply_cannon_metrocomm4 17395 15.5 0.059 0.068 3.501 9.375 dbcsr_mm_accdrv_process 41158 16.2 4.508 6.068 8.200 9.314 mp_irecv_dv 49801 16.2 3.382 9.129 3.382 9.129 pw_transfer 1535 11.6 0.084 0.105 7.974 8.085 cp_fm_cholesky_decompose 22 10.9 7.866 7.893 7.866 7.893 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 7.753 7.871 grid_integrate_task_list 127 12.3 7.231 7.742 7.231 7.742 wfi_extrapolate 11 7.9 0.001 0.001 7.531 7.531 cp_fm_upper_to_full 104 14.5 5.720 7.403 5.720 7.403 density_rs2pw 127 9.7 0.005 0.006 6.803 7.262 fft_wrap_pw1pw2_140 519 13.2 0.472 0.526 6.720 6.849 dbcsr_complete_redistribute 393 12.7 1.183 1.213 4.848 6.677 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 5.090 6.544 mp_sum_l 7804 13.0 4.346 6.322 4.346 6.322 fft3d_ps 1281 14.7 2.676 2.870 6.037 6.094 rs_pw_transfer 1038 11.9 0.013 0.014 5.556 6.007 calculate_dm_sparse 127 9.5 0.001 0.001 5.746 5.847 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.580 5.584 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.567 5.394 grid_collocate_task_list 127 9.7 5.030 5.378 5.030 5.378 mp_alltoall_d11v 2401 14.1 4.415 5.292 4.415 5.292 potential_pw2rs 127 12.3 0.020 0.022 4.826 4.851 mp_allgather_i34 2485 14.5 1.619 4.819 1.619 4.819 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.588 4.744 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.118 4.156 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.388 4.130 mp_alltoall_i22 712 14.1 2.109 4.015 2.109 4.015 qs_energies_init_hamiltonians 11 5.9 0.002 0.011 3.781 3.789 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=181.395000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=886.181818, yerr=11.271994 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d60029441ac03bb22e0b17392546c4f4f2d728f9_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456039424 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613072052224 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176077312 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176077312 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.233020E+12 0.0% 0.0% 100.0% flops max/rank 4.387242E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383904 0.0% 0.0% 100.0% number of processed stacks 6026880 0.0% 0.0% 100.0% average stack size 0.0 0.0 1129.3 marketing flops 145.651870E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 1.133408E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1153224 MPI messages size (bytes): total size 2.039489E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.768511E+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 322096 36390305792 131072 < size <= 4194304 721976 792118951936 4194304 < size <= 16777216 70800 669922227920 16777216 < size 30960 541065216000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4123 56680. MP_Allreduce 11346 1064. MP_Sync 172 MP_Alltoall 1724 12509605. MP_ISendRecv 11868 75008. MP_Wait 28546 MP_comm_split 84 MP_ISend 15064 244788. MP_IRecv 15064 244788. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.038 0.189 193.933 193.934 qs_mol_dyn_low 1 2.0 0.003 0.003 193.282 193.296 qs_forces 11 3.9 0.005 0.008 193.163 193.170 qs_energies 11 4.9 0.002 0.006 186.084 186.096 scf_env_do_scf 11 5.9 0.001 0.002 168.974 168.984 velocity_verlet 10 3.0 0.001 0.001 128.230 128.233 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 120.992 120.995 qs_scf_new_mos 118 7.6 0.001 0.001 84.475 84.757 qs_scf_loop_do_ot 118 8.6 0.001 0.002 84.474 84.756 dbcsr_multiply_generic 2529 12.6 0.195 0.218 81.585 82.232 ot_scf_mini 118 9.6 0.003 0.004 79.761 80.027 multiply_cannon 2529 13.6 0.563 0.597 55.791 58.786 multiply_cannon_loop 2529 14.6 1.201 1.226 51.427 52.849 init_scf_loop 11 6.9 0.007 0.054 47.855 47.860 ot_mini 118 10.6 0.001 0.001 43.702 43.972 prepare_preconditioner 11 7.9 0.000 0.000 43.602 43.632 make_preconditioner 11 8.9 0.002 0.020 43.602 43.632 make_full_inverse_cholesky 11 9.9 0.000 0.001 37.161 42.277 multiply_cannon_multrec 30348 15.6 14.294 19.621 26.314 31.035 rebuild_ks_matrix 129 8.3 0.001 0.001 29.653 29.882 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.019 29.652 29.881 mp_waitall_1 155106 16.5 18.592 29.664 18.592 29.664 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.864 27.072 qs_ot_get_derivative 118 11.6 0.001 0.002 23.698 23.977 make_m2s 5058 13.6 0.096 0.099 21.209 22.852 make_images 5058 14.6 1.950 2.278 20.898 22.544 qs_ot_get_p 129 10.4 0.001 0.001 21.988 22.266 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 19.459 19.923 apply_single 129 13.6 0.001 0.001 19.458 19.923 ot_diis_step 118 11.6 0.018 0.018 19.862 19.864 qs_ot_p2m_diag 84 11.4 0.347 0.394 17.533 17.585 cp_fm_upper_to_full 106 14.7 11.378 16.854 11.378 16.854 cp_fm_cholesky_invert 11 10.9 16.842 16.852 16.842 16.852 cp_dbcsr_syevd 84 12.4 0.005 0.005 16.215 16.216 multiply_cannon_metrocomm3 30348 15.6 0.048 0.050 6.339 15.711 sum_up_and_integrate 129 10.3 0.142 0.153 14.395 14.418 integrate_v_rspace 129 11.3 0.003 0.004 14.253 14.281 make_images_data 5058 15.6 0.064 0.067 11.696 13.997 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.536 13.576 calculate_rho_elec 129 8.7 0.176 0.192 13.535 13.575 hybrid_alltoall_any 5245 16.5 0.529 2.225 10.230 13.082 dbcsr_complete_redistribute 397 12.7 1.509 1.630 9.217 13.059 cp_fm_diag_elpa 84 13.4 0.000 0.001 12.994 12.997 cp_fm_redistribute_end 84 14.4 2.208 12.900 2.224 12.905 multiply_cannon_sync_h2d 30348 15.6 11.931 12.858 11.931 12.858 cp_fm_diag_elpa_base 84 14.4 10.039 12.278 10.640 12.777 dbcsr_mm_accdrv_process 62780 16.2 7.470 8.544 11.594 12.081 copy_fm_to_dbcsr 210 11.7 0.002 0.002 7.760 11.621 init_scf_run 11 5.9 0.000 0.001 10.908 10.909 scf_env_initial_rho_setup 11 6.9 0.000 0.001 10.908 10.909 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.415 10.181 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 9.978 10.179 mp_alltoall_i22 720 14.1 5.728 9.523 5.728 9.523 pw_transfer 1559 11.6 0.086 0.103 8.172 8.247 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 7.945 8.023 grid_integrate_task_list 129 12.3 7.611 7.950 7.611 7.950 cp_fm_cholesky_decompose 22 10.9 7.816 7.930 7.816 7.930 wfi_extrapolate 11 7.9 0.001 0.001 7.701 7.701 multiply_cannon_metrocomm4 25290 15.6 0.075 0.085 2.778 7.163 density_rs2pw 129 9.7 0.005 0.006 6.544 7.079 fft_wrap_pw1pw2_140 527 13.2 0.484 0.490 6.966 7.065 mp_irecv_dv 76751 16.2 2.633 6.889 2.633 6.889 calculate_dm_sparse 129 9.5 0.001 0.001 6.393 6.484 fft3d_ps 1301 14.7 2.816 2.872 6.164 6.230 mp_alltoall_d11v 2429 14.1 5.095 5.730 5.095 5.730 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.554 5.601 grid_collocate_task_list 129 9.7 5.251 5.557 5.251 5.557 rs_pw_transfer 1054 12.0 0.013 0.015 4.875 5.378 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.510 4.586 potential_pw2rs 129 12.3 0.022 0.023 4.503 4.539 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.413 4.460 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 4.421 4.425 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.248 4.298 mp_allgather_i34 2529 14.6 2.451 4.157 2.451 4.157 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=193.934000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1074.636364, yerr=15.967426 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d60029441ac03bb22e0b17392546c4f4f2d728f9_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430457200640 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986252263424 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992001093632 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613077360640 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239162695680 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239162695680 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.232994E+12 0.0% 0.0% 100.0% flops max/rank 5.910120E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806382528 0.0% 0.0% 100.0% number of processed stacks 1976928 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.9 marketing flops 145.650931E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 1.510240E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 242784 MPI messages size (bytes): total size 1.341806E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.526746E+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 115008 60297314304 4194304 < size <= 16777216 105840 554906419200 16777216 < size 20352 726592214928 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 9010 51. MP_Alltoall 9724 794507. MP_ISend 40420 2100460. MP_IRecv 40420 2099564. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 57630. MP_Allreduce 11104 1171. MP_Sync 88 MP_Alltoall 1724 18848066. MP_ISendRecv 7740 122880. MP_Wait 20114 MP_ISend 10760 423501. MP_IRecv 10760 423501. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.020 0.036 180.916 180.917 qs_mol_dyn_low 1 2.0 0.003 0.004 180.505 180.519 qs_forces 11 3.9 0.003 0.006 180.390 180.394 qs_energies 11 4.9 0.001 0.002 173.036 173.046 scf_env_do_scf 11 5.9 0.001 0.001 155.586 155.598 velocity_verlet 10 3.0 0.001 0.001 118.133 118.139 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 117.514 117.515 qs_scf_new_mos 118 7.6 0.001 0.001 80.769 80.844 qs_scf_loop_do_ot 118 8.6 0.001 0.001 80.769 80.843 ot_scf_mini 118 9.6 0.004 0.004 76.227 76.287 dbcsr_multiply_generic 2529 12.6 0.185 0.189 75.877 76.208 multiply_cannon 2529 13.6 0.588 0.622 55.334 60.129 multiply_cannon_loop 2529 14.6 0.454 0.466 50.373 51.003 ot_mini 118 10.6 0.001 0.001 40.859 40.918 init_scf_loop 11 6.9 0.000 0.000 37.914 37.915 mp_waitall_1 130746 16.6 27.234 34.435 27.234 34.435 prepare_preconditioner 11 7.9 0.000 0.000 33.957 33.993 make_preconditioner 11 8.9 0.000 0.000 33.957 33.993 make_full_inverse_cholesky 11 9.9 0.000 0.000 31.623 31.919 rebuild_ks_matrix 129 8.3 0.001 0.001 29.236 29.311 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.018 29.236 29.311 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.711 26.780 qs_ot_get_p 129 10.4 0.001 0.001 22.235 22.348 multiply_cannon_multrec 10116 15.6 10.479 14.942 17.788 20.771 qs_ot_get_derivative 118 11.6 0.002 0.002 20.654 20.714 ot_diis_step 118 11.6 0.020 0.020 20.123 20.124 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.695 19.901 apply_single 129 13.6 0.001 0.001 19.695 19.901 multiply_cannon_metrocomm3 10116 15.6 0.023 0.025 12.810 19.621 cp_fm_cholesky_invert 11 10.9 19.599 19.606 19.599 19.606 make_m2s 5058 13.6 0.067 0.071 16.640 18.979 make_images 5058 14.6 2.298 2.684 16.331 18.671 qs_ot_p2m_diag 84 11.4 0.502 0.507 18.121 18.143 cp_dbcsr_syevd 84 12.4 0.005 0.006 16.884 16.886 sum_up_and_integrate 129 10.3 0.184 0.193 14.617 14.662 integrate_v_rspace 129 11.3 0.004 0.004 14.433 14.487 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.701 13.741 calculate_rho_elec 129 8.7 0.260 0.271 13.700 13.740 cp_fm_diag_elpa 84 13.4 0.000 0.000 13.516 13.517 cp_fm_diag_elpa_base 84 14.4 13.249 13.340 13.512 13.512 make_images_data 5058 15.6 0.054 0.061 10.110 12.497 multiply_cannon_sync_h2d 10116 15.6 11.645 12.047 11.645 12.047 hybrid_alltoall_any 5245 16.5 0.851 3.849 9.742 12.030 init_scf_run 11 5.9 0.000 0.001 10.605 10.605 scf_env_initial_rho_setup 11 6.9 0.000 0.001 10.605 10.605 cp_fm_cholesky_decompose 22 10.9 8.508 8.582 8.508 8.582 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 8.436 8.484 pw_transfer 1559 11.6 0.084 0.092 8.214 8.236 grid_integrate_task_list 129 12.3 7.809 8.218 7.809 8.218 fft_wrap_pw1pw2 1301 12.7 0.010 0.010 7.992 8.015 mp_allgather_i34 2529 14.6 3.135 7.763 3.135 7.763 dbcsr_mm_accdrv_process 20934 16.1 2.602 3.306 6.939 7.604 wfi_extrapolate 11 7.9 0.001 0.001 7.529 7.530 multiply_cannon_metrocomm1 10116 15.6 0.029 0.030 4.508 7.457 fft_wrap_pw1pw2_140 527 13.2 0.504 0.528 6.933 6.968 density_rs2pw 129 9.7 0.005 0.006 6.382 6.601 calculate_dm_sparse 129 9.5 0.001 0.001 6.215 6.305 fft3d_ps 1301 14.7 2.747 2.819 6.141 6.184 dbcsr_complete_redistribute 397 12.7 2.159 2.282 5.728 6.143 mp_alltoall_d11v 2429 14.1 5.317 6.072 5.317 6.072 grid_collocate_task_list 129 9.7 5.526 5.729 5.526 5.729 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.354 5.367 qs_energies_init_hamiltonians 11 5.9 0.003 0.008 5.245 5.247 rs_pw_transfer 1054 12.0 0.012 0.013 4.540 4.774 potential_pw2rs 129 12.3 0.026 0.027 4.460 4.492 multiply_cannon_metrocomm4 7587 15.6 0.023 0.025 1.857 4.462 mp_irecv_dv 29102 15.9 1.823 4.388 1.823 4.388 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.151 4.194 copy_fm_to_dbcsr 210 11.7 0.002 0.002 3.787 4.145 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.907 3.974 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.557 3.852 copy_dbcsr_to_fm 187 11.8 0.004 0.004 3.691 3.750 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.616 3.635 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=180.917000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1408.272727, yerr=48.215597 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d60029441ac03bb22e0b17392546c4f4f2d728f9_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963542011904 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019182452736 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019182452736 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796564E+12 0.0% 0.0% 100.0% flops max/rank 11.606412E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499488 0.0% 0.0% 100.0% number of processed stacks 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 2.918081E+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 11.342272E+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 716108490000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3981 59443. MP_Allreduce 10950 1516. MP_Sync 86 MP_Alltoall 1700 36954339. 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 0.030 0.084 302.634 302.635 qs_mol_dyn_low 1 2.0 0.003 0.004 301.908 301.928 qs_forces 11 3.9 0.003 0.003 301.799 301.802 qs_energies 11 4.9 0.012 0.023 293.132 293.146 scf_env_do_scf 11 5.9 0.001 0.001 270.404 270.418 velocity_verlet 10 3.0 0.001 0.001 218.004 218.014 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 140.997 140.999 init_scf_loop 11 6.9 0.000 0.000 129.127 129.129 prepare_preconditioner 11 7.9 0.000 0.000 124.248 124.281 make_preconditioner 11 8.9 0.000 0.000 124.248 124.281 make_full_inverse_cholesky 11 9.9 0.000 0.000 100.149 121.324 qs_scf_new_mos 116 7.6 0.001 0.001 98.040 98.268 qs_scf_loop_do_ot 116 8.6 0.001 0.001 98.039 98.267 ot_scf_mini 116 9.6 0.004 0.004 93.158 93.279 dbcsr_multiply_generic 2485 12.5 0.212 0.220 86.384 87.178 cp_fm_upper_to_full 104 14.8 53.135 76.122 53.135 76.122 multiply_cannon 2485 13.5 0.711 0.749 60.556 61.075 multiply_cannon_loop 2485 14.5 0.467 0.482 56.486 58.416 ot_mini 116 10.6 0.001 0.001 46.599 46.760 dbcsr_complete_redistribute 393 12.7 3.987 4.033 30.171 42.851 copy_fm_to_dbcsr 208 11.6 0.001 0.002 26.534 39.243 cp_fm_cholesky_invert 11 10.9 36.807 36.816 36.807 36.816 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 24.053 36.711 mp_waitall_1 104546 16.7 31.074 35.408 31.074 35.408 mp_alltoall_i22 712 14.1 21.816 34.729 21.816 34.729 rebuild_ks_matrix 127 8.3 0.001 0.001 34.183 34.333 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.017 34.183 34.332 qs_ks_update_qs_env 127 7.6 0.001 0.001 31.860 31.995 qs_ot_get_p 127 10.4 0.001 0.001 30.555 30.708 qs_ot_p2m_diag 82 11.4 0.868 0.878 26.107 26.145 qs_ot_get_derivative 116 11.6 0.002 0.002 25.642 25.772 cp_dbcsr_syevd 82 12.4 0.005 0.006 24.279 24.282 multiply_cannon_metrocomm3 9940 15.5 0.024 0.024 20.440 22.311 make_m2s 4970 13.5 0.075 0.077 20.781 22.146 make_images 4970 14.5 3.715 3.830 20.304 21.671 ot_diis_step 116 11.6 0.022 0.022 20.901 20.902 cp_fm_diag_elpa 82 13.4 0.000 0.000 20.789 20.790 cp_fm_diag_elpa_base 82 14.4 16.430 18.060 20.783 20.784 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 20.073 20.458 apply_single 127 13.6 0.001 0.001 20.072 20.458 multiply_cannon_multrec 9940 15.5 10.390 12.089 17.883 18.020 sum_up_and_integrate 127 10.3 0.319 0.321 16.063 16.147 integrate_v_rspace 127 11.3 0.004 0.004 15.743 15.828 qs_rho_update_rho_low 127 7.7 0.001 0.001 15.569 15.624 calculate_rho_elec 127 8.7 0.479 0.480 15.569 15.624 multiply_cannon_sync_h2d 9940 15.5 15.548 15.564 15.548 15.564 hybrid_alltoall_any 5155 16.4 1.290 3.022 11.377 13.688 make_images_data 4970 15.5 0.059 0.063 11.331 13.570 init_scf_run 11 5.9 0.000 0.001 12.566 12.567 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.566 12.566 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 10.029 10.131 cp_fm_cholesky_decompose 22 10.9 9.914 9.960 9.914 9.960 wfi_extrapolate 11 7.9 0.001 0.001 9.253 9.253 dbcsr_mm_accdrv_process 20590 16.0 3.775 5.679 7.259 9.077 pw_transfer 1535 11.6 0.089 0.089 8.945 8.959 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.714 8.727 grid_integrate_task_list 127 12.3 8.487 8.654 8.487 8.654 qs_energies_init_hamiltonians 11 5.9 0.012 0.043 7.982 7.983 mp_alltoall_d11v 2401 14.1 7.529 7.635 7.529 7.635 fft_wrap_pw1pw2_140 519 13.2 0.535 0.536 7.592 7.610 calculate_dm_sparse 127 9.5 0.001 0.001 7.008 7.130 fft3d_ps 1281 14.7 2.741 2.746 6.774 6.782 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.569 6.656 density_rs2pw 127 9.7 0.005 0.005 6.550 6.576 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.406 6.481 grid_collocate_task_list 127 9.7 6.291 6.322 6.291 6.322 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=302.635000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2588.090909, yerr=167.646519 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d60029441ac03bb22e0b17392546c4f4f2d728f9_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 1.261445E+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 0.023 0.046 85.219 85.220 qs_energies 1 2.0 0.000 0.000 84.778 84.784 ls_scf 1 3.0 0.000 0.000 83.859 83.866 dbcsr_multiply_generic 111 6.7 0.015 0.015 72.703 72.861 multiply_cannon 111 7.7 0.018 0.021 55.792 57.024 multiply_cannon_loop 111 8.7 0.212 0.228 52.348 53.803 ls_scf_main 1 4.0 0.000 0.000 52.518 52.523 density_matrix_trs4 2 5.0 0.002 0.003 47.079 47.174 ls_scf_init_scf 1 4.0 0.000 0.000 28.328 28.333 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.207 27.267 mp_waitall_1 11316 10.9 22.404 25.352 22.404 25.352 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.042 25.056 multiply_cannon_multrec 2664 9.7 8.196 8.953 15.504 17.224 multiply_cannon_sync_h2d 2664 9.7 13.871 15.792 13.871 15.792 make_m2s 222 7.7 0.008 0.011 13.326 13.808 make_images 222 8.7 0.099 0.109 13.304 13.789 multiply_cannon_metrocomm1 2664 9.7 0.010 0.010 9.323 13.231 make_images_data 222 9.7 0.004 0.005 7.945 8.597 hybrid_alltoall_any 227 10.6 0.215 1.837 6.800 8.115 dbcsr_mm_accdrv_process 4760 10.4 0.510 0.633 6.924 7.916 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.452 7.830 dbcsr_mm_accdrv_process_sort 4760 11.4 6.215 7.106 6.215 7.106 calculate_norms 4752 9.8 5.563 6.284 5.563 6.284 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.938 5.113 mp_sum_l 807 5.4 3.053 4.617 3.053 4.617 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.268 3.527 arnoldi_extremal 4 6.8 0.000 0.000 3.348 3.375 arnoldi_normal_ev 4 7.8 0.001 0.004 3.348 3.375 make_images_sizes 222 9.7 0.000 0.000 0.786 3.372 mp_alltoall_i44 222 10.7 0.786 3.372 0.786 3.372 multiply_cannon_metrocomm4 2442 9.7 0.012 0.014 2.039 3.342 mp_irecv_dv 6231 10.9 2.022 3.310 2.022 3.310 build_subspace 16 8.4 0.009 0.012 3.249 3.252 ls_scf_post 1 4.0 0.000 0.000 3.013 3.020 ls_scf_store_result 1 5.0 0.000 0.000 2.836 2.875 dbcsr_special_finalize 555 9.7 0.005 0.006 2.243 2.726 dbcsr_merge_single_wm 555 10.7 0.458 0.573 2.235 2.718 make_images_pack 222 9.7 2.206 2.625 2.208 2.627 dbcsr_matrix_vector_mult 304 9.0 0.003 0.010 2.364 2.616 dbcsr_matrix_vector_mult_local 304 10.0 2.062 2.456 2.064 2.458 dbcsr_sort_data 658 11.4 2.037 2.427 2.037 2.427 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.250 2.335 buffer_matrices_ensure_size 222 8.7 1.748 2.074 1.748 2.074 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.799 1.805 rebuild_ks_matrix 3 7.3 0.000 0.000 1.790 1.795 qs_ks_build_kohn_sham_matrix 3 8.3 0.008 0.016 1.790 1.795 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.220000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1133.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d60029441ac03bb22e0b17392546c4f4f2d728f9_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 2.104541E+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 0.054 0.167 90.984 90.985 qs_energies 1 2.0 0.000 0.000 90.175 90.199 ls_scf 1 3.0 0.000 0.001 88.849 88.876 dbcsr_multiply_generic 111 6.7 0.015 0.016 74.807 75.139 multiply_cannon 111 7.7 0.029 0.044 52.967 56.677 ls_scf_main 1 4.0 0.000 0.001 54.538 54.547 multiply_cannon_loop 111 8.7 0.116 0.123 49.738 52.908 density_matrix_trs4 2 5.0 0.002 0.003 48.896 49.098 ls_scf_init_scf 1 4.0 0.000 0.001 30.776 30.777 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.517 29.633 mp_waitall_1 9246 10.9 20.795 29.295 20.795 29.295 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.066 27.078 multiply_cannon_multrec 1332 9.7 13.131 16.767 22.326 27.074 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.567 20.420 make_m2s 222 7.7 0.006 0.008 15.127 15.741 make_images 222 8.7 1.566 1.957 15.097 15.712 dbcsr_mm_accdrv_process 4041 10.4 0.284 0.452 8.796 10.385 dbcsr_mm_accdrv_process_sort 4041 11.4 8.390 9.933 8.390 9.933 make_images_data 222 9.7 0.004 0.004 8.596 9.432 hybrid_alltoall_any 227 10.6 0.521 2.480 8.055 8.929 mp_sum_l 807 5.4 5.426 8.188 5.426 8.188 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.246 7.869 mp_irecv_dv 3311 11.0 3.226 7.804 3.226 7.804 calculate_norms 2376 9.8 6.003 6.823 6.003 6.823 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.210 6.583 multiply_cannon_sync_h2d 1332 9.7 4.773 6.049 4.773 6.049 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.046 5.266 arnoldi_extremal 4 6.8 0.000 0.000 4.692 4.713 arnoldi_normal_ev 4 7.8 0.001 0.005 4.692 4.713 build_subspace 16 8.4 0.014 0.021 4.437 4.440 ls_scf_post 1 4.0 0.004 0.030 3.535 3.562 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.169 3.399 ls_scf_store_result 1 5.0 0.000 0.000 3.236 3.355 dbcsr_matrix_vector_mult_local 304 10.0 2.738 3.215 2.740 3.216 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.252 2.784 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.516 2.643 make_images_pack 222 9.7 2.019 2.418 2.021 2.420 mp_allgather_i34 111 8.7 0.929 2.209 0.929 2.209 dbcsr_sort_data 436 11.2 1.828 2.068 1.828 2.068 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.907 1.909 rebuild_ks_matrix 3 7.3 0.000 0.000 1.894 1.896 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.006 1.894 1.896 dbcsr_data_new 4174 10.1 1.611 1.837 1.611 1.837 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=90.985000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1718.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d60029441ac03bb22e0b17392546c4f4f2d728f9_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 2.730815E+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 0.070 0.149 93.644 93.644 qs_energies 1 2.0 0.000 0.000 92.916 92.920 ls_scf 1 3.0 0.000 0.001 91.539 91.542 dbcsr_multiply_generic 111 6.7 0.016 0.024 75.922 76.146 ls_scf_main 1 4.0 0.001 0.017 57.097 57.112 multiply_cannon 111 7.7 0.044 0.119 52.308 56.047 multiply_cannon_loop 111 8.7 0.100 0.105 48.730 52.414 density_matrix_trs4 2 5.0 0.002 0.003 51.255 51.420 mp_waitall_1 7374 11.0 23.241 33.623 23.241 33.623 ls_scf_init_scf 1 4.0 0.000 0.002 30.879 30.891 ls_scf_init_matrix_S 1 5.0 0.000 0.002 29.505 29.559 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.126 27.138 multiply_cannon_multrec 888 9.7 12.681 15.439 21.171 24.360 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.708 22.855 make_m2s 222 7.7 0.006 0.008 17.076 18.340 make_images 222 8.7 1.973 2.255 17.038 18.298 hybrid_alltoall_any 227 10.6 0.624 2.883 9.270 10.847 make_images_data 222 9.7 0.004 0.005 9.675 10.647 dbcsr_mm_accdrv_process 3754 10.4 0.234 0.412 8.028 9.206 dbcsr_mm_accdrv_process_sort 3754 11.4 7.658 8.794 7.658 8.794 mp_sum_l 807 5.4 4.978 8.307 4.978 8.307 multiply_cannon_sync_h2d 888 9.7 5.987 7.437 5.987 7.437 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.426 7.408 mp_irecv_dv 2335 11.1 2.411 7.367 2.411 7.367 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.831 6.493 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.471 6.312 arnoldi_extremal 4 6.8 0.000 0.000 5.277 5.296 arnoldi_normal_ev 4 7.8 0.001 0.005 5.276 5.296 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.870 5.074 build_subspace 16 8.4 0.014 0.021 4.950 4.958 calculate_norms 1584 9.8 4.317 4.742 4.317 4.742 dbcsr_matrix_vector_mult 304 9.0 0.005 0.016 3.503 3.840 mp_allgather_i34 111 8.7 1.423 3.712 1.423 3.712 dbcsr_matrix_vector_mult_local 304 10.0 3.015 3.589 3.017 3.591 ls_scf_post 1 4.0 0.000 0.001 3.564 3.567 ls_scf_store_result 1 5.0 0.000 0.000 3.287 3.392 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.721 2.812 dbcsr_sort_data 325 11.1 1.925 2.148 1.925 2.148 make_images_pack 222 9.7 1.815 2.125 1.817 2.128 make_images_sizes 222 9.7 0.000 0.000 1.046 1.953 mp_alltoall_i44 222 10.7 1.046 1.952 1.046 1.952 dbcsr_data_release 9322 10.9 1.288 1.945 1.288 1.945 dbcsr_finalize 304 7.8 0.026 0.034 1.616 1.941 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.898 1.910 rebuild_ks_matrix 3 7.3 0.000 0.000 1.880 1.892 qs_ks_build_kohn_sham_matrix 3 8.3 0.011 0.023 1.880 1.892 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=93.644000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2196.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d60029441ac03bb22e0b17392546c4f4f2d728f9_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 3.337011E+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 0.039 0.053 100.310 100.311 qs_energies 1 2.0 0.000 0.000 99.653 99.660 ls_scf 1 3.0 0.000 0.000 97.987 97.993 dbcsr_multiply_generic 111 6.7 0.016 0.018 81.184 81.548 ls_scf_main 1 4.0 0.000 0.000 61.060 61.061 multiply_cannon 111 7.7 0.050 0.113 52.492 56.718 density_matrix_trs4 2 5.0 0.002 0.003 54.813 54.926 multiply_cannon_loop 111 8.7 0.115 0.125 47.488 52.121 ls_scf_init_scf 1 4.0 0.000 0.000 33.467 33.469 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.238 32.306 mp_waitall_1 6438 11.0 24.495 30.212 24.495 30.212 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.685 29.700 multiply_cannon_multrec 1332 9.7 14.154 17.473 21.968 24.707 make_m2s 222 7.7 0.006 0.008 21.993 23.401 make_images 222 8.7 3.122 3.579 21.943 23.353 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.795 18.707 make_images_data 222 9.7 0.004 0.004 12.560 14.247 hybrid_alltoall_any 227 10.6 0.794 3.784 12.036 13.927 mp_sum_l 807 5.4 5.142 9.479 5.142 9.479 dbcsr_mm_accdrv_process 3641 10.4 0.202 0.399 7.459 8.969 dbcsr_mm_accdrv_process_sort 3641 11.4 7.088 8.566 7.088 8.566 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.016 7.648 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.088 6.168 multiply_cannon_sync_h2d 1332 9.7 5.543 6.142 5.543 6.142 mp_irecv_dv 3229 10.9 2.065 6.122 2.065 6.122 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.956 5.581 arnoldi_extremal 4 6.8 0.000 0.000 5.512 5.532 arnoldi_normal_ev 4 7.8 0.001 0.005 5.512 5.532 build_subspace 16 8.4 0.014 0.021 5.158 5.167 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.722 4.906 calculate_norms 2376 9.8 4.176 4.535 4.176 4.535 mp_allgather_i34 111 8.7 2.128 4.468 2.128 4.468 dbcsr_matrix_vector_mult 304 9.0 0.006 0.016 3.677 3.962 dbcsr_matrix_vector_mult_local 304 10.0 3.185 3.682 3.187 3.684 ls_scf_post 1 4.0 0.000 0.000 3.460 3.467 dbcsr_sort_data 658 11.4 3.114 3.456 3.114 3.456 ls_scf_store_result 1 5.0 0.000 0.000 3.188 3.262 dbcsr_special_finalize 555 9.7 0.006 0.007 2.857 3.159 dbcsr_merge_single_wm 555 10.7 0.535 0.663 2.848 3.151 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.980 3.040 dbcsr_data_release 10477 10.7 1.589 2.458 1.589 2.458 dbcsr_finalize 304 7.8 0.049 0.061 1.803 2.022 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=100.311000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2689.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d60029441ac03bb22e0b17392546c4f4f2d728f9_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 4.630847E+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 0.059 0.093 99.934 99.934 qs_energies 1 2.0 0.000 0.000 99.066 99.069 ls_scf 1 3.0 0.000 0.000 97.151 97.157 dbcsr_multiply_generic 111 6.7 0.017 0.019 78.526 78.783 ls_scf_main 1 4.0 0.000 0.000 62.411 62.416 multiply_cannon 111 7.7 0.085 0.165 55.415 60.409 density_matrix_trs4 2 5.0 0.002 0.003 55.342 55.450 multiply_cannon_loop 111 8.7 0.069 0.076 50.969 53.396 mp_waitall_1 5481 11.0 26.937 32.217 26.937 32.217 ls_scf_init_scf 1 4.0 0.000 0.000 31.153 31.159 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.910 29.957 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.732 27.745 multiply_cannon_multrec 444 9.7 14.115 16.690 21.147 23.657 make_m2s 222 7.7 0.004 0.005 18.011 20.448 make_images 222 8.7 3.710 4.401 17.950 20.388 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.208 16.112 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.208 14.376 make_images_data 222 9.7 0.003 0.004 10.315 12.647 hybrid_alltoall_any 227 10.6 0.787 3.752 10.052 12.527 multiply_cannon_sync_h2d 444 9.7 6.535 7.947 6.535 7.947 dbcsr_mm_accdrv_process 3003 10.4 0.200 0.392 6.740 7.851 dbcsr_mm_accdrv_process_sort 3003 11.4 6.411 7.507 6.411 7.507 mp_allgather_i34 111 8.7 2.677 7.048 2.677 7.048 arnoldi_extremal 4 6.8 0.000 0.000 6.018 6.032 arnoldi_normal_ev 4 7.8 0.002 0.005 6.018 6.032 build_subspace 16 8.4 0.015 0.019 5.624 5.633 mp_sum_l 807 5.4 3.177 5.106 3.177 5.106 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.590 4.753 dbcsr_matrix_vector_mult 304 9.0 0.007 0.016 4.239 4.441 dbcsr_matrix_vector_mult_local 304 10.0 3.698 4.170 3.700 4.172 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.661 4.009 mp_irecv_dv 1241 11.2 1.642 3.981 1.642 3.981 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.253 3.918 make_images_sizes 222 9.7 0.000 0.000 1.066 3.738 mp_alltoall_i44 222 10.7 1.065 3.737 1.065 3.737 calculate_norms 792 9.8 3.529 3.644 3.529 3.644 ls_scf_post 1 4.0 0.000 0.000 3.586 3.589 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.426 3.529 ls_scf_store_result 1 5.0 0.000 0.000 3.340 3.387 dbcsr_finalize 304 7.8 0.062 0.078 2.196 2.283 dbcsr_merge_all 275 8.9 0.473 0.526 2.044 2.098 dbcsr_data_release 10123 10.8 1.328 2.046 1.328 2.046 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.020 2.021 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=99.934000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3655.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d60029441ac03bb22e0b17392546c4f4f2d728f9_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 8.834527E+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 0.079 0.108 107.111 107.111 qs_energies 1 2.0 0.000 0.000 105.705 105.717 ls_scf 1 3.0 0.000 0.000 102.812 102.824 dbcsr_multiply_generic 111 6.7 0.023 0.027 76.797 76.962 ls_scf_main 1 4.0 0.000 0.000 65.249 65.260 density_matrix_trs4 2 5.0 0.002 0.003 56.400 56.468 multiply_cannon 111 7.7 0.170 0.356 49.258 51.975 multiply_cannon_loop 111 8.7 0.067 0.069 45.835 46.700 ls_scf_init_scf 1 4.0 0.000 0.000 33.915 33.931 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.563 32.577 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.746 29.758 mp_waitall_1 4569 11.1 21.352 24.805 21.352 24.805 make_m2s 222 7.7 0.005 0.005 23.632 24.570 make_images 222 8.7 4.576 4.966 23.526 24.463 multiply_cannon_multrec 444 9.7 17.889 18.444 22.482 23.032 hybrid_alltoall_any 227 10.6 1.660 3.625 12.731 15.553 make_images_data 222 9.7 0.003 0.003 12.896 15.289 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 9.820 10.908 multiply_cannon_sync_h2d 444 9.7 8.843 8.896 8.843 8.896 arnoldi_extremal 4 6.8 0.000 0.000 7.487 7.503 arnoldi_normal_ev 4 7.8 0.004 0.010 7.487 7.503 build_subspace 16 8.4 0.026 0.036 6.932 6.944 dbcsr_matrix_vector_mult 304 9.0 0.009 0.025 5.489 5.665 dbcsr_matrix_vector_mult_local 304 10.0 5.043 5.345 5.046 5.348 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.983 5.218 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.818 4.909 dbcsr_mm_accdrv_process 1814 10.4 0.207 0.317 4.418 4.556 dbcsr_mm_accdrv_process_sort 1814 11.4 4.120 4.261 4.120 4.261 ls_scf_post 1 4.0 0.000 0.000 3.648 3.659 make_images_sizes 222 9.7 0.000 0.000 1.421 3.505 mp_alltoall_i44 222 10.7 1.421 3.505 1.421 3.505 ls_scf_store_result 1 5.0 0.000 0.000 3.399 3.409 mp_allgather_i34 111 8.7 1.070 3.298 1.070 3.298 calculate_norms 792 9.8 3.246 3.285 3.246 3.285 dbcsr_finalize 304 7.8 0.082 0.090 3.112 3.161 dbcsr_merge_all 275 8.9 0.893 0.920 2.877 2.920 dbcsr_data_release 12724 10.6 2.360 2.915 2.360 2.915 dbcsr_complete_redistribute 5 7.6 1.440 1.477 2.789 2.892 qs_energies_init_hamiltonians 1 3.0 0.002 0.002 2.862 2.863 matrix_ls_to_qs 2 6.0 0.000 0.000 2.434 2.548 dbcsr_sort_data 325 11.1 2.444 2.508 2.444 2.508 mp_sum_l 807 5.4 1.593 2.328 1.593 2.328 dbcsr_new_transposed 4 7.5 0.243 0.252 2.293 2.303 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.216 2.233 dbcsr_add_d 103 6.2 0.000 0.000 2.165 2.233 dbcsr_add_anytype 103 7.2 0.859 0.891 2.165 2.233 dbcsr_frobenius_norm 74 6.6 2.054 2.129 2.180 2.214 rebuild_ks_matrix 3 7.3 0.000 0.000 2.150 2.167 qs_ks_build_kohn_sham_matrix 3 8.3 0.011 0.027 2.150 2.167 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=107.111000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6873.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: d60029441ac03bb22e0b17392546c4f4f2d728f9 Summary: empty Status: OK