=== 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: 845a3ae33c0b6c5fc8bf8a28d823ca3539bb7d75 ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, # Cray-FFTW 3.3.8.10, COSMA 2.6.6, ELPA 2023.05.001, # HDF5 1.14.2, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.2.2, LIBVORI 220621, LIBXSMM 1.17, # PLUMED 2.9.0, SIRIUS 7.4.3, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Last update: 04.10.2023 # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2023.05.001 USE_HDF5 := 1.14.2 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.9.0 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) 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__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 += $(LIBVDWXC_LIB)/libvdwxc.a endif ifneq ($(USE_HDF5),) USE_HDF5 := $(strip $(USE_HDF5)) HDF5_INC := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/include HDF5_LIB := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/lib CFLAGS += -I$(HDF5_INC) DFLAGS += -D__HDF5 LIBS += $(HDF5_LIB)/libhdf5_fortran.a LIBS += $(HDF5_LIB)/libhdf5_hl.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/845a3ae33c0b6c5fc8bf8a28d823ca3539bb7d75_performance_tests/01 job id: 50261425 --- 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/845a3ae33c0b6c5fc8bf8a28d823ca3539bb7d75_performance_tests/02 job id: 50261426 --- 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/845a3ae33c0b6c5fc8bf8a28d823ca3539bb7d75_performance_tests/03 job id: 50261427 --- 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/845a3ae33c0b6c5fc8bf8a28d823ca3539bb7d75_performance_tests/04 job id: 50261428 --- 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/845a3ae33c0b6c5fc8bf8a28d823ca3539bb7d75_performance_tests/05 job id: 50261429 --- 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/845a3ae33c0b6c5fc8bf8a28d823ca3539bb7d75_performance_tests/06 job id: 50261430 --- 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/845a3ae33c0b6c5fc8bf8a28d823ca3539bb7d75_performance_tests/07 job id: 50261431 --- 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/845a3ae33c0b6c5fc8bf8a28d823ca3539bb7d75_performance_tests/08 job id: 50261432 --- 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/845a3ae33c0b6c5fc8bf8a28d823ca3539bb7d75_performance_tests/09 job id: 50261433 --- 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/845a3ae33c0b6c5fc8bf8a28d823ca3539bb7d75_performance_tests/10 job id: 50261435 --- 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/845a3ae33c0b6c5fc8bf8a28d823ca3539bb7d75_performance_tests/11 job id: 50261436 --- 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/845a3ae33c0b6c5fc8bf8a28d823ca3539bb7d75_performance_tests/12 job id: 50261438 --- 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/845a3ae33c0b6c5fc8bf8a28d823ca3539bb7d75_performance_tests/13 job id: 50261439 --- 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/845a3ae33c0b6c5fc8bf8a28d823ca3539bb7d75_performance_tests/14 job id: 50261441 --- 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/845a3ae33c0b6c5fc8bf8a28d823ca3539bb7d75_performance_tests/15 job id: 50261442 --- 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/845a3ae33c0b6c5fc8bf8a28d823ca3539bb7d75_performance_tests/16 job id: 50261443 --- 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/845a3ae33c0b6c5fc8bf8a28d823ca3539bb7d75_performance_tests/17 job id: 50261445 --- 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/845a3ae33c0b6c5fc8bf8a28d823ca3539bb7d75_performance_tests/18 job id: 50261446 --- 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/845a3ae33c0b6c5fc8bf8a28d823ca3539bb7d75_performance_tests/19 job id: 50261448 --- 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/845a3ae33c0b6c5fc8bf8a28d823ca3539bb7d75_performance_tests/20 job id: 50261449 --- 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/845a3ae33c0b6c5fc8bf8a28d823ca3539bb7d75_performance_tests/21 job id: 50261450 --- 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/845a3ae33c0b6c5fc8bf8a28d823ca3539bb7d75_performance_tests/22 job id: 50261451 --- 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/845a3ae33c0b6c5fc8bf8a28d823ca3539bb7d75_performance_tests/23 job id: 50261452 --- 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/845a3ae33c0b6c5fc8bf8a28d823ca3539bb7d75_performance_tests/24 job id: 50261454 --- 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/845a3ae33c0b6c5fc8bf8a28d823ca3539bb7d75_performance_tests/25 job id: 50261456 --- 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/845a3ae33c0b6c5fc8bf8a28d823ca3539bb7d75_performance_tests/26 job id: 50261457 --- Point --- name: 510 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/1r/12t) --- Point --- name: 511 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: 512 H2O (4 NVE MD steps on 64 nodes) input file: benchmarks/QS/00512_H2O/H2O-512_md.inp required files: [] output file: result.log # nodes = 64 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/845a3ae33c0b6c5fc8bf8a28d823ca3539bb7d75_performance_tests/27 job id: 50261459 --- Point --- name: 601 plot: h2o_512_md regex: CP2K label: (64n/12r/1t) --- Point --- name: 602 plot: h2o_512_md_mem regex: Estimated peak process memory label: (64n/12r/1t) ~~~~~~~ END TEST ~~~~~~~ === END TESTS (description) === ===== PLOTS (description) ===== ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2_mem", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md_mem", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md_mem", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md_mem", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls_mem", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_512_md", title="512 H2O (4 NVE MD steps on 64 nodes)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_512_md_mem", title="512 H2O (4 NVE MD steps on 64 nodes)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" === END PLOTS (description) === ============ RESULTS ============ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/845a3ae33c0b6c5fc8bf8a28d823ca3539bb7d75_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 424 8. 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 0.019 0.033 138.170 138.171 farming_run 1 2.0 137.608 137.609 138.121 138.124 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.490993E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 230 1134128. MP_Allreduce 571 1938539. MP_Sync 25 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 10 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.007 0.026 116.744 116.744 qs_energies 1 2.0 0.000 0.000 116.490 116.492 mp2_main 1 3.0 0.000 0.000 113.872 113.873 mp2_gpw_main 1 4.0 0.055 0.064 112.563 112.565 mp2_ri_gpw_compute_in 1 5.0 0.171 0.173 93.036 93.386 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 54.636 54.981 mp2_eri_3c_integrate_gpw 272 7.0 0.152 0.162 41.053 46.127 get_2c_integrals 1 6.0 0.018 0.025 37.297 38.234 integrate_v_rspace 273 8.0 0.438 0.452 24.672 29.516 pw_transfer 6555 10.6 0.372 0.387 26.888 27.580 fft_wrap_pw1pw2 5465 11.4 0.045 0.047 25.523 26.109 grid_integrate_task_list 273 9.0 20.570 25.882 20.570 25.882 fft_wrap_pw1pw2_100 2178 12.4 0.027 0.030 23.050 23.621 rpa_ri_compute_en 1 5.0 0.019 0.020 19.393 19.631 compute_2c_integrals 1 7.0 0.002 0.002 19.517 19.518 compute_2c_integrals_loop_lm 1 8.0 0.002 0.003 18.629 19.068 mp2_eri_2c_integrate_gpw 1 9.0 2.369 2.415 18.627 19.067 cp_fm_cholesky_decompose 12 8.2 17.895 18.970 17.895 18.970 cholesky_decomp 1 7.0 0.000 0.000 16.618 17.578 fft3d_s 5443 13.4 16.154 16.570 16.175 16.591 ao_to_mo_and_store_B_mult_1 272 7.0 10.761 15.323 10.761 15.323 calculate_wavefunction 272 8.0 5.372 5.524 12.310 13.027 rpa_num_int 1 6.0 0.000 0.001 10.902 10.911 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.775 10.891 calc_mat_Q 8 8.0 0.000 0.000 9.482 9.566 contract_S_to_Q 8 9.0 0.000 0.000 8.904 8.986 calc_potential_gpw 544 9.5 0.004 0.004 8.210 8.676 parallel_gemm_fm 14 9.1 0.000 0.000 8.484 8.582 parallel_gemm_fm_cosma 14 10.1 8.483 8.582 8.483 8.582 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 8.095 8.407 potential_pw2rs 545 10.0 0.106 0.107 7.522 8.225 create_integ_mat 1 6.0 0.015 0.028 8.054 8.064 collocate_single_gaussian 272 10.0 0.039 0.041 7.354 7.592 array2fm 1 7.0 0.000 0.000 6.693 7.105 pw_scatter_s 2720 13.7 4.348 4.586 4.348 4.586 pw_gather_s 2722 13.2 3.432 3.779 3.432 3.779 array2fm_buffer_send 1 8.0 2.936 3.105 2.936 3.105 pw_poisson_solve 545 10.5 1.121 1.198 2.195 2.353 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.562874, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2808.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/845a3ae33c0b6c5fc8bf8a28d823ca3539bb7d75_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 424 9. 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 0.023 0.036 422.114 422.115 farming_run 1 2.0 421.247 421.250 422.079 422.080 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.248174E+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 757 478553. MP_Allreduce 2021 21391. MP_Sync 37 MP_Alltoall 77 14575236. MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.032 216.287 216.288 qs_energies 1 2.0 0.000 0.000 216.057 216.062 scf_env_do_scf 1 3.0 0.000 0.000 114.056 114.056 qs_ks_update_qs_env 5 5.0 0.000 0.000 113.066 113.073 rebuild_ks_matrix 4 6.0 0.000 0.000 113.064 113.072 qs_ks_build_kohn_sham_matrix 4 7.0 0.055 0.061 113.064 113.072 hfx_ks_matrix 4 8.0 0.001 0.001 112.668 112.673 integrate_four_center 4 9.0 0.152 0.456 112.667 112.672 integrate_four_center_main 4 10.0 0.122 0.659 101.821 104.195 integrate_four_center_bin 263 11.0 101.699 104.136 101.699 104.136 mp2_main 1 3.0 0.000 0.000 101.678 101.683 mp2_gpw_main 1 4.0 0.047 0.071 100.750 100.759 init_scf_loop 1 4.0 0.000 0.000 96.220 96.221 mp2_ri_gpw_compute_in 1 5.0 0.064 0.064 74.015 75.046 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 53.768 54.800 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.158 41.674 46.465 integrate_v_rspace 95 8.0 0.397 0.569 28.198 32.826 pw_transfer 2240 10.6 0.145 0.179 29.512 29.938 fft_wrap_pw1pw2 1868 11.4 0.018 0.023 28.510 28.952 ao_to_mo_and_store_B_mult_1 91 7.0 10.419 28.903 10.419 28.903 mp2_ri_gpw_compute_en 1 5.0 0.056 0.062 26.584 28.712 grid_integrate_task_list 95 9.0 23.569 28.398 23.569 28.398 fft_wrap_pw1pw2_100 730 12.4 0.012 0.014 26.246 26.707 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.843 1.916 24.907 24.918 get_2c_integrals 1 6.0 0.000 0.000 20.150 20.183 compute_2c_integrals 1 7.0 0.002 0.003 19.135 19.140 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 18.669 18.974 mp2_eri_2c_integrate_gpw 1 9.0 1.728 1.887 18.668 18.972 fft3d_s 1823 13.4 18.459 18.862 18.473 18.876 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 17.834 17.834 calculate_wavefunction 91 8.0 2.009 2.047 9.617 9.839 mp2_ri_gpw_compute_en_expansio 172 7.0 0.557 0.588 8.809 9.402 potential_pw2rs 186 10.0 0.033 0.035 8.443 9.028 local_gemm 172 8.0 8.251 8.814 8.251 8.814 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.181 8.487 mp2_ri_gpw_compute_en_comm 22 7.0 0.505 0.523 7.884 8.355 calc_potential_gpw 182 9.5 0.002 0.002 7.955 8.181 collocate_single_gaussian 91 10.0 0.016 0.022 7.771 7.968 integrate_four_center_load 4 10.0 0.000 0.000 6.818 6.822 hfx_load_balance 1 11.0 0.000 0.000 6.818 6.822 mp2_ri_gpw_compute_en_ener 172 7.0 6.327 6.386 6.327 6.386 mp_sendrecv_dm3 2068 8.0 5.905 6.359 5.905 6.359 mp_sync 37 10.5 2.747 6.109 2.747 6.109 pw_gather_s 912 13.2 4.454 4.970 4.454 4.970 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=100.751852, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1507.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/845a3ae33c0b6c5fc8bf8a28d823ca3539bb7d75_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 452.538368E+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 62379. MP_Allreduce 10329 270. MP_Sync 530 MP_Alltoall 2083 1604322. MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.009 0.032 54.905 54.906 qs_mol_dyn_low 1 2.0 0.003 0.007 54.472 54.602 qs_forces 11 3.9 0.005 0.024 54.387 54.387 qs_energies 11 4.9 0.002 0.002 52.877 52.897 scf_env_do_scf 11 5.9 0.000 0.001 46.642 46.643 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 44.517 44.517 dbcsr_multiply_generic 2286 12.5 0.093 0.098 34.236 34.731 qs_scf_new_mos 108 7.5 0.000 0.001 34.432 34.709 qs_scf_loop_do_ot 108 8.5 0.000 0.001 34.432 34.709 ot_scf_mini 108 9.5 0.002 0.002 32.734 32.916 multiply_cannon 2286 13.5 0.189 0.196 26.678 28.404 multiply_cannon_loop 2286 14.5 1.813 1.917 26.010 27.758 velocity_verlet 10 3.0 0.001 0.002 27.234 27.235 ot_mini 108 10.5 0.001 0.001 19.563 19.796 qs_ot_get_derivative 108 11.5 0.001 0.001 16.508 16.686 mp_waitall_1 245248 16.5 8.656 15.140 8.656 15.140 multiply_cannon_metrocomm3 54864 15.5 0.072 0.078 6.211 13.382 multiply_cannon_multrec 54864 15.5 3.644 5.713 7.914 11.406 qs_ot_get_p 119 10.4 0.001 0.001 8.374 8.691 rebuild_ks_matrix 119 8.3 0.000 0.000 7.852 7.989 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.016 7.852 7.989 mp_sum_l 7287 12.8 5.572 7.336 5.572 7.336 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.913 7.032 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 6.064 6.528 multiply_cannon_sync_h2d 54864 15.5 5.117 6.484 5.117 6.484 dbcsr_mm_accdrv_process 76910 16.1 1.837 2.896 4.184 5.964 qs_ot_p2m_diag 50 11.0 0.005 0.010 5.336 5.384 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.059 5.159 init_scf_run 11 5.9 0.000 0.001 4.958 4.959 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.958 4.958 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.524 4.525 sum_up_and_integrate 119 10.3 0.001 0.003 4.439 4.446 integrate_v_rspace 119 11.3 0.002 0.003 4.429 4.437 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.180 4.375 calculate_rho_elec 119 8.7 0.011 0.016 4.179 4.375 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.309 4.309 cp_fm_redistribute_end 50 14.0 2.203 4.283 2.212 4.287 cp_fm_diag_elpa_base 50 14.0 2.069 4.166 2.073 4.176 calculate_dm_sparse 119 9.5 0.000 0.000 3.039 3.198 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.946 3.139 apply_single 119 13.6 0.000 0.000 2.946 3.139 jit_kernel_multiply 13 15.8 2.285 2.981 2.285 2.981 multiply_cannon_metrocomm1 54864 15.5 0.055 0.060 1.786 2.929 acc_transpose_blocks 54864 15.5 0.225 0.248 2.265 2.818 ot_diis_step 108 11.5 0.006 0.006 2.793 2.794 calculate_first_density_matrix 1 7.0 0.001 0.010 2.762 2.771 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.615 2.700 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.403 2.405 density_rs2pw 119 9.7 0.004 0.004 2.193 2.296 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.133 2.193 grid_integrate_task_list 119 12.3 2.013 2.126 2.013 2.126 wfi_extrapolate 11 7.9 0.001 0.001 2.122 2.122 init_scf_loop 11 6.9 0.000 0.001 2.108 2.108 mp_sum_d 4137 12.0 1.431 1.983 1.431 1.983 potential_pw2rs 119 12.3 0.004 0.004 1.813 1.823 pw_transfer 1439 11.6 0.051 0.055 1.720 1.795 make_m2s 4572 13.5 0.053 0.055 1.675 1.726 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.645 1.723 make_images 4572 14.5 0.132 0.139 1.594 1.644 mp_alltoall_d11v 2130 13.8 1.321 1.592 1.321 1.592 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.505 1.554 mp_waitany 12084 13.8 1.263 1.506 1.263 1.506 transfer_rs2pw 487 10.6 0.005 0.006 1.379 1.496 acc_transpose_blocks_sync 164592 16.5 1.199 1.446 1.199 1.446 fft3d_ps 1201 14.6 0.375 0.480 1.359 1.431 grid_collocate_task_list 119 9.7 1.350 1.430 1.350 1.430 fft_wrap_pw1pw2_140 487 13.2 0.107 0.114 1.274 1.350 transfer_pw2rs 487 13.2 0.005 0.006 1.315 1.325 dbcsr_dot_sd 1205 11.9 0.048 0.059 0.780 1.175 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=54.906000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=431.727273, yerr=0.616575 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/845a3ae33c0b6c5fc8bf8a28d823ca3539bb7d75_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 488.927232E+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 62658. MP_Allreduce 10306 303. MP_Sync 54 MP_Alltoall 2060 1567776. MP_SendRecv 16779 37093. MP_ISendRecv 16779 37093. MP_Wait 23539 MP_comm_split 50 MP_ISend 5720 128509. MP_IRecv 5720 128509. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.008 0.029 45.849 45.850 qs_mol_dyn_low 1 2.0 0.003 0.004 45.653 45.660 qs_forces 11 3.9 0.002 0.003 45.589 45.591 qs_energies 11 4.9 0.001 0.002 43.774 43.780 scf_env_do_scf 11 5.9 0.000 0.001 38.184 38.185 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 34.828 34.829 dbcsr_multiply_generic 2286 12.5 0.103 0.108 24.065 24.418 qs_scf_new_mos 108 7.5 0.001 0.001 23.802 24.026 qs_scf_loop_do_ot 108 8.5 0.001 0.001 23.801 24.025 ot_scf_mini 108 9.5 0.002 0.003 22.784 22.946 velocity_verlet 10 3.0 0.001 0.002 22.154 22.156 multiply_cannon 2286 13.5 0.210 0.218 17.478 19.748 multiply_cannon_loop 2286 14.5 1.180 1.238 16.010 17.858 ot_mini 108 10.5 0.001 0.001 13.824 14.062 mp_waitall_1 200699 16.5 7.123 11.721 7.123 11.721 qs_ot_get_derivative 108 11.5 0.001 0.001 11.283 11.447 multiply_cannon_metrocomm3 27432 15.5 0.071 0.074 4.376 9.677 multiply_cannon_multrec 27432 15.5 1.804 4.071 6.139 8.940 rebuild_ks_matrix 119 8.3 0.000 0.000 8.103 8.315 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 8.103 8.314 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.181 7.374 dbcsr_mm_accdrv_process 47894 16.0 3.498 5.773 4.257 6.509 qs_ot_get_p 119 10.4 0.001 0.001 5.777 5.998 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 4.280 5.242 qs_rho_update_rho_low 119 7.7 0.001 0.001 5.003 5.028 calculate_rho_elec 119 8.7 0.021 0.024 5.002 5.028 sum_up_and_integrate 119 10.3 0.001 0.001 4.894 4.899 integrate_v_rspace 119 11.3 0.002 0.003 4.880 4.885 mp_sum_l 7287 12.8 2.614 4.691 2.614 4.691 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.190 4.369 apply_single 119 13.6 0.000 0.000 3.190 4.368 init_scf_run 11 5.9 0.000 0.001 4.291 4.291 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.290 4.291 make_m2s 4572 13.5 0.051 0.053 3.526 3.803 qs_ot_p2m_diag 50 11.0 0.009 0.012 3.713 3.733 make_images 4572 14.5 0.204 0.243 3.436 3.716 density_rs2pw 119 9.7 0.004 0.004 3.275 3.350 init_scf_loop 11 6.9 0.000 0.001 3.319 3.320 cp_dbcsr_syevd 50 12.0 0.003 0.003 3.257 3.258 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.745 2.831 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.712 2.714 cp_fm_redistribute_end 50 14.0 1.369 2.659 1.377 2.664 potential_pw2rs 119 12.3 0.006 0.006 2.593 2.623 cp_fm_diag_elpa_base 50 14.0 1.244 2.516 1.277 2.571 pw_transfer 1439 11.6 0.064 0.067 2.465 2.522 ot_diis_step 108 11.5 0.011 0.012 2.444 2.445 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.374 2.433 calculate_first_density_matrix 1 7.0 0.000 0.000 2.331 2.333 prepare_preconditioner 11 7.9 0.000 0.000 2.281 2.308 make_preconditioner 11 8.9 0.000 0.000 2.281 2.308 calculate_dm_sparse 119 9.5 0.000 0.001 2.188 2.282 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.154 2.223 transfer_rs2pw 487 10.6 0.005 0.005 2.113 2.198 multiply_cannon_sync_h2d 27432 15.5 1.637 2.174 1.637 2.174 fft3d_ps 1201 14.6 0.529 0.585 2.008 2.067 make_images_data 4572 15.5 0.047 0.053 1.518 2.067 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.035 2.037 acc_transpose_blocks 27432 15.5 0.116 0.122 1.606 1.945 grid_integrate_task_list 119 12.3 1.826 1.930 1.826 1.930 wfi_extrapolate 11 7.9 0.001 0.001 1.872 1.873 transfer_pw2rs 487 13.2 0.004 0.005 1.831 1.857 jit_kernel_multiply 9 16.4 0.700 1.801 0.700 1.801 hybrid_alltoall_any 4725 16.4 0.053 0.115 1.331 1.776 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.693 1.751 mp_alltoall_d11v 2130 13.8 1.595 1.736 1.595 1.736 fft_wrap_pw1pw2_140 487 13.2 0.126 0.132 1.621 1.682 make_images_sizes 4572 15.5 0.005 0.005 1.150 1.564 mp_alltoall_i44 4572 16.5 1.145 1.559 1.145 1.559 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.375 1.445 mp_allgather_i34 2286 14.5 0.892 1.385 0.892 1.385 grid_collocate_task_list 119 9.7 1.290 1.359 1.290 1.359 mp_sendrecv_dv 16779 12.7 1.314 1.349 1.314 1.349 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.300 1.312 mp_sum_d 4137 12.0 0.695 1.206 0.695 1.206 mp_alltoall_z22v 1201 16.6 1.092 1.184 1.092 1.184 cp_fm_cholesky_invert 11 10.9 1.092 1.099 1.092 1.099 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.984 1.002 dbcsr_complete_redistribute 329 12.2 0.077 0.095 0.908 0.995 transfer_rs2pw_50 119 11.7 0.182 0.189 0.877 0.975 qs_energies_init_hamiltonians 11 5.9 0.000 0.002 0.963 0.964 acc_transpose_blocks_sync 82296 16.5 0.808 0.934 0.808 0.934 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=45.850000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=465.727273, yerr=1.420045 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/845a3ae33c0b6c5fc8bf8a28d823ca3539bb7d75_performance_tests/05/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 59.051995E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3143552 0.0% 0.0% 100.0% average stack size 0.0 0.0 46.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 519.520256E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63490. MP_Allreduce 10155 305. MP_Sync 54 MP_Alltoall 1821 1607811. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.009 0.028 39.052 39.053 qs_mol_dyn_low 1 2.0 0.003 0.004 38.789 38.797 qs_forces 11 3.9 0.002 0.002 38.477 38.479 qs_energies 11 4.9 0.001 0.002 36.765 36.774 scf_env_do_scf 11 5.9 0.000 0.001 31.582 31.582 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 28.277 28.278 dbcsr_multiply_generic 2286 12.5 0.100 0.102 18.963 19.318 velocity_verlet 10 3.0 0.001 0.002 19.139 19.148 qs_scf_new_mos 108 7.5 0.001 0.001 18.502 18.516 qs_scf_loop_do_ot 108 8.5 0.001 0.001 18.502 18.516 ot_scf_mini 108 9.5 0.002 0.002 17.621 17.644 multiply_cannon 2286 13.5 0.195 0.201 14.214 15.258 multiply_cannon_loop 2286 14.5 0.860 0.908 13.145 14.271 ot_mini 108 10.5 0.001 0.001 10.593 10.618 qs_ot_get_derivative 108 11.5 0.001 0.001 8.874 8.905 multiply_cannon_multrec 18288 15.5 1.846 2.854 6.998 7.345 rebuild_ks_matrix 119 8.3 0.000 0.000 7.214 7.314 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.213 7.313 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.374 6.466 dbcsr_mm_accdrv_process 38222 16.0 5.002 5.791 5.061 5.858 mp_waitall_1 158411 16.6 4.012 5.646 4.012 5.646 sum_up_and_integrate 119 10.3 0.001 0.001 4.654 4.665 integrate_v_rspace 119 11.3 0.003 0.003 4.641 4.653 qs_ot_get_p 119 10.4 0.001 0.001 4.482 4.523 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.425 4.434 calculate_rho_elec 119 8.7 0.031 0.032 4.425 4.434 init_scf_run 11 5.9 0.000 0.001 3.918 3.918 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.918 3.918 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.278 3.893 init_scf_loop 11 6.9 0.000 0.001 3.278 3.279 multiply_cannon_metrocomm3 18288 15.5 0.048 0.049 1.655 3.196 density_rs2pw 119 9.7 0.004 0.004 2.927 3.035 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.956 2.963 make_m2s 4572 13.5 0.044 0.045 2.556 2.753 make_images 4572 14.5 0.192 0.203 2.469 2.666 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.613 2.614 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.089 2.463 apply_single 119 13.6 0.000 0.000 2.089 2.463 potential_pw2rs 119 12.3 0.007 0.008 2.378 2.415 prepare_preconditioner 11 7.9 0.000 0.000 2.373 2.377 make_preconditioner 11 8.9 0.000 0.000 2.373 2.377 pw_transfer 1439 11.6 0.064 0.068 2.314 2.355 calculate_first_density_matrix 1 7.0 0.000 0.000 2.345 2.347 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.205 2.300 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.222 2.263 mp_sum_l 7287 12.8 1.799 2.245 1.799 2.245 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.227 2.238 cp_fm_diag_elpa_base 50 14.0 2.195 2.206 2.225 2.236 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.159 2.174 calculate_dm_sparse 119 9.5 0.000 0.001 2.022 2.036 transfer_rs2pw 487 10.6 0.005 0.005 1.858 1.987 grid_integrate_task_list 119 12.3 1.796 1.910 1.796 1.910 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.829 1.831 fft3d_ps 1201 14.6 0.552 0.574 1.780 1.823 acc_transpose_blocks 18288 15.5 0.080 0.083 1.617 1.693 transfer_pw2rs 487 13.2 0.004 0.004 1.649 1.689 fft_wrap_pw1pw2_140 487 13.2 0.178 0.184 1.588 1.627 ot_diis_step 108 11.5 0.012 0.012 1.620 1.620 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.586 1.619 wfi_extrapolate 11 7.9 0.001 0.001 1.500 1.500 make_images_data 4572 15.5 0.047 0.051 1.062 1.386 grid_collocate_task_list 119 9.7 1.232 1.300 1.232 1.300 hybrid_alltoall_any 4725 16.4 0.058 0.118 0.890 1.272 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.253 1.267 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.188 1.246 cp_fm_cholesky_invert 11 10.9 1.163 1.170 1.163 1.170 multiply_cannon_sync_h2d 18288 15.5 0.994 1.167 0.994 1.167 mp_alltoall_z22v 1201 16.6 1.027 1.126 1.027 1.126 mp_sendrecv_dv 11067 12.7 1.082 1.121 1.082 1.121 mp_alltoall_d11v 2130 13.8 1.011 1.115 1.011 1.115 make_images_sizes 4572 15.5 0.005 0.005 0.747 0.999 mp_alltoall_i44 4572 16.5 0.742 0.994 0.742 0.994 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.974 0.975 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.915 0.924 mp_allgather_i34 2286 14.5 0.520 0.921 0.520 0.921 acc_transpose_blocks_kernels 18288 16.5 0.219 0.226 0.785 0.805 acc_transpose_blocks_sync 54864 16.5 0.733 0.802 0.733 0.802 dbcsr_complete_redistribute 329 12.2 0.092 0.101 0.676 0.786 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=39.053000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=495.272727, yerr=0.749656 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/845a3ae33c0b6c5fc8bf8a28d823ca3539bb7d75_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 558.555136E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63489. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.008 0.027 37.334 37.335 qs_mol_dyn_low 1 2.0 0.003 0.004 37.139 37.148 qs_forces 11 3.9 0.002 0.002 36.643 36.643 qs_energies 11 4.9 0.001 0.001 34.939 34.945 scf_env_do_scf 11 5.9 0.000 0.001 29.302 29.303 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 25.951 25.951 dbcsr_multiply_generic 2286 12.5 0.102 0.105 20.396 20.517 velocity_verlet 10 3.0 0.001 0.002 18.849 18.869 qs_scf_new_mos 108 7.5 0.001 0.001 17.854 17.911 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.853 17.911 multiply_cannon 2286 13.5 0.219 0.226 16.722 17.459 ot_scf_mini 108 9.5 0.002 0.003 16.817 16.876 multiply_cannon_loop 2286 14.5 1.545 1.607 15.784 16.336 ot_mini 108 10.5 0.001 0.001 10.299 10.361 multiply_cannon_multrec 27432 15.5 2.442 3.128 9.567 9.890 qs_ot_get_derivative 108 11.5 0.001 0.001 8.434 8.496 dbcsr_mm_accdrv_process 47916 15.9 6.289 8.121 7.023 8.426 rebuild_ks_matrix 119 8.3 0.000 0.000 6.288 6.348 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.288 6.348 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.569 5.623 init_scf_run 11 5.9 0.000 0.002 4.293 4.293 scf_env_initial_rho_setup 11 6.9 0.002 0.003 4.293 4.293 qs_ot_get_p 119 10.4 0.001 0.001 3.670 3.742 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.129 3.526 sum_up_and_integrate 119 10.3 0.001 0.001 3.469 3.477 integrate_v_rspace 119 11.3 0.002 0.003 3.458 3.465 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.418 3.457 calculate_rho_elec 119 8.7 0.040 0.046 3.418 3.457 init_scf_loop 11 6.9 0.001 0.001 3.332 3.333 calculate_first_density_matrix 1 7.0 0.000 0.000 2.889 2.892 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.113 2.661 apply_single 119 13.6 0.000 0.000 2.113 2.660 acc_transpose_blocks 27432 15.5 0.121 0.124 2.432 2.566 prepare_preconditioner 11 7.9 0.000 0.000 2.515 2.524 make_preconditioner 11 8.9 0.000 0.000 2.515 2.524 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.125 2.449 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.395 2.396 make_m2s 4572 13.5 0.053 0.056 2.267 2.371 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.329 2.361 make_images 4572 14.5 0.273 0.335 2.160 2.262 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.253 2.262 mp_waitall_1 137007 16.6 1.789 2.260 1.789 2.260 calculate_dm_sparse 119 9.5 0.000 0.001 2.179 2.233 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.220 2.231 pw_transfer 1439 11.6 0.064 0.068 2.000 2.034 density_rs2pw 119 9.7 0.003 0.004 1.839 1.964 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.908 1.946 grid_integrate_task_list 119 12.3 1.824 1.928 1.824 1.928 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.849 1.850 ot_diis_step 108 11.5 0.012 0.012 1.827 1.828 jit_kernel_multiply 6 16.0 0.668 1.728 0.668 1.728 fft_wrap_pw1pw2_140 487 13.2 0.227 0.234 1.571 1.612 mp_sum_l 7287 12.8 0.977 1.576 0.977 1.576 acc_transpose_blocks_sync 82296 16.5 1.455 1.576 1.455 1.576 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.566 1.575 cp_fm_diag_elpa_base 50 14.0 1.533 1.549 1.564 1.574 fft3d_ps 1201 14.6 0.592 0.655 1.396 1.422 wfi_extrapolate 11 7.9 0.001 0.001 1.350 1.350 multiply_cannon_metrocomm3 27432 15.5 0.039 0.040 0.810 1.347 grid_collocate_task_list 119 9.7 1.248 1.344 1.248 1.344 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.291 1.311 potential_pw2rs 119 12.3 0.008 0.009 1.278 1.281 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.268 1.279 cp_fm_upper_to_full 72 14.2 0.815 1.158 0.815 1.158 qs_energies_init_hamiltonians 11 5.9 0.006 0.020 1.087 1.088 transfer_rs2pw 487 10.6 0.004 0.005 0.899 1.028 dbcsr_complete_redistribute 329 12.2 0.121 0.148 0.750 1.026 mp_alltoall_d11v 2130 13.8 0.810 0.961 0.810 0.961 make_images_data 4572 15.5 0.048 0.052 0.844 0.960 hybrid_alltoall_any 4725 16.4 0.066 0.156 0.765 0.930 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.799 0.874 acc_transpose_blocks_kernels 27432 16.5 0.273 0.279 0.828 0.857 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.560 0.828 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.807 0.812 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=37.335000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=528.909091, yerr=2.937363 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/845a3ae33c0b6c5fc8bf8a28d823ca3539bb7d75_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 620.613632E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63488. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.032 35.628 35.630 qs_mol_dyn_low 1 2.0 0.003 0.004 35.428 35.436 qs_forces 11 3.9 0.002 0.002 35.363 35.364 qs_energies 11 4.9 0.002 0.002 33.480 33.487 scf_env_do_scf 11 5.9 0.000 0.001 27.921 27.921 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 24.374 24.375 velocity_verlet 10 3.0 0.002 0.002 18.322 18.325 dbcsr_multiply_generic 2286 12.5 0.097 0.098 15.421 15.555 qs_scf_new_mos 108 7.5 0.001 0.001 15.067 15.106 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.066 15.105 ot_scf_mini 108 9.5 0.002 0.002 14.230 14.266 multiply_cannon 2286 13.5 0.227 0.236 11.583 12.459 multiply_cannon_loop 2286 14.5 0.645 0.663 10.394 10.694 ot_mini 108 10.5 0.001 0.001 7.719 7.772 rebuild_ks_matrix 119 8.3 0.000 0.000 6.836 6.876 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.835 6.876 qs_ot_get_derivative 108 11.5 0.001 0.001 6.247 6.281 multiply_cannon_multrec 9144 15.5 1.665 1.856 5.978 6.193 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.078 6.116 qs_ot_get_p 119 10.4 0.001 0.001 4.298 4.352 sum_up_and_integrate 119 10.3 0.001 0.001 4.294 4.299 integrate_v_rspace 119 11.3 0.003 0.003 4.284 4.289 dbcsr_mm_accdrv_process 12550 15.8 3.515 4.236 4.203 4.277 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.263 4.271 calculate_rho_elec 119 8.7 0.060 0.061 4.262 4.270 init_scf_run 11 5.9 0.000 0.002 4.040 4.040 scf_env_initial_rho_setup 11 6.9 0.001 0.002 4.039 4.040 mp_waitall_1 115863 16.7 2.860 3.730 2.860 3.730 init_scf_loop 11 6.9 0.001 0.001 3.520 3.522 qs_ot_p2m_diag 50 11.0 0.022 0.023 2.904 2.907 make_m2s 4572 13.5 0.033 0.034 2.622 2.887 make_images 4572 14.5 0.268 0.299 2.532 2.798 density_rs2pw 119 9.7 0.003 0.003 2.551 2.698 prepare_preconditioner 11 7.9 0.000 0.000 2.653 2.660 make_preconditioner 11 8.9 0.000 0.000 2.653 2.660 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.592 2.592 pw_transfer 1439 11.6 0.065 0.067 2.541 2.563 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.518 2.560 calculate_first_density_matrix 1 7.0 0.000 0.000 2.481 2.482 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.448 2.472 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.151 2.158 cp_fm_diag_elpa_base 50 14.0 2.114 2.136 2.148 2.155 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.127 2.152 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.035 2.036 grid_integrate_task_list 119 12.3 1.907 2.010 1.907 2.010 potential_pw2rs 119 12.3 0.010 0.011 1.972 1.983 calculate_dm_sparse 119 9.5 0.000 0.000 1.887 1.913 fft_wrap_pw1pw2_140 487 13.2 0.324 0.332 1.875 1.899 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.852 1.871 fft3d_ps 1201 14.6 0.653 0.666 1.801 1.822 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.786 1.797 acc_transpose_blocks 9144 15.5 0.043 0.044 1.759 1.787 make_images_data 4572 15.5 0.042 0.046 1.200 1.573 transfer_rs2pw 487 10.6 0.004 0.004 1.373 1.509 wfi_extrapolate 11 7.9 0.001 0.001 1.495 1.495 jit_kernel_multiply 8 15.5 0.648 1.477 0.648 1.477 hybrid_alltoall_any 4725 16.4 0.064 0.175 1.130 1.476 cp_fm_cholesky_invert 11 10.9 1.461 1.465 1.461 1.465 grid_collocate_task_list 119 9.7 1.338 1.445 1.338 1.445 ot_diis_step 108 11.5 0.013 0.013 1.417 1.418 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.329 1.376 apply_single 119 13.6 0.000 0.000 1.329 1.376 mp_alltoall_d11v 2130 13.8 1.115 1.315 1.115 1.315 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.241 1.242 transfer_pw2rs 487 13.2 0.003 0.004 1.186 1.197 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.173 1.179 mp_allgather_i34 2286 14.5 0.503 1.111 0.503 1.111 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.047 1.063 mp_alltoall_z22v 1201 16.6 1.021 1.060 1.021 1.060 multiply_cannon_metrocomm3 9144 15.5 0.019 0.020 0.483 1.038 mp_sum_l 7287 12.8 0.770 0.997 0.770 0.997 acc_transpose_blocks_kernels 9144 16.5 0.118 0.121 0.980 0.984 make_images_sizes 4572 15.5 0.005 0.005 0.645 0.963 mp_alltoall_i44 4572 16.5 0.640 0.958 0.640 0.958 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.874 0.923 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.882 0.887 jit_kernel_transpose 5 15.6 0.862 0.867 0.862 0.867 multiply_cannon_metrocomm1 9144 15.5 0.023 0.024 0.682 0.854 dbcsr_complete_redistribute 329 12.2 0.157 0.166 0.744 0.787 cp_fm_cholesky_decompose 22 10.9 0.753 0.757 0.753 0.757 acc_transpose_blocks_sync 27432 16.5 0.721 0.749 0.721 0.749 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.690 0.737 mp_sendrecv_dv 5355 12.7 0.719 0.734 0.719 0.734 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=35.630000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=589.363636, yerr=5.788275 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/845a3ae33c0b6c5fc8bf8a28d823ca3539bb7d75_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 780.775424E+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 63723. MP_Allreduce 10154 429. MP_Sync 54 MP_Alltoall 1582 7383731. MP_SendRecv 2499 189067. MP_ISendRecv 2499 189067. MP_Wait 6399 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.020 0.051 49.567 49.570 qs_mol_dyn_low 1 2.0 0.003 0.004 49.293 49.312 qs_forces 11 3.9 0.002 0.002 48.646 48.647 qs_energies 11 4.9 0.001 0.002 46.501 46.506 scf_env_do_scf 11 5.9 0.001 0.001 40.240 40.240 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 31.326 31.327 velocity_verlet 10 3.0 0.016 0.019 27.979 28.057 dbcsr_multiply_generic 2286 12.5 0.108 0.110 20.999 21.313 qs_scf_new_mos 108 7.5 0.001 0.001 20.138 20.256 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.138 20.255 ot_scf_mini 108 9.5 0.002 0.002 18.866 18.976 multiply_cannon 2286 13.5 0.299 0.303 15.711 17.052 multiply_cannon_loop 2286 14.5 0.865 0.888 14.323 15.289 ot_mini 108 10.5 0.001 0.001 11.067 11.211 multiply_cannon_multrec 9144 15.5 3.487 4.822 8.897 8.983 qs_ot_get_derivative 108 11.5 0.001 0.001 8.813 8.922 init_scf_loop 11 6.9 0.001 0.001 8.880 8.882 rebuild_ks_matrix 119 8.3 0.000 0.000 8.212 8.383 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 8.211 8.383 prepare_preconditioner 11 7.9 0.000 0.000 7.795 7.810 make_preconditioner 11 8.9 0.000 0.000 7.795 7.810 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.294 7.679 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.374 7.531 dbcsr_mm_accdrv_process 12550 15.8 4.433 6.543 5.269 6.600 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.994 4.999 calculate_rho_elec 119 8.7 0.117 0.120 4.993 4.999 qs_ot_get_p 119 10.4 0.001 0.001 4.547 4.721 mp_waitall_1 94719 16.7 3.590 4.691 3.590 4.691 cp_fm_upper_to_full 72 14.2 3.208 4.650 3.208 4.650 sum_up_and_integrate 119 10.3 0.001 0.001 4.563 4.569 integrate_v_rspace 119 11.3 0.003 0.003 4.553 4.559 init_scf_run 11 5.9 0.000 0.001 4.112 4.112 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.111 4.112 pw_transfer 1439 11.6 0.068 0.069 3.593 3.606 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.172 3.558 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 3.495 3.508 make_m2s 4572 13.5 0.037 0.037 3.296 3.506 make_images 4572 14.5 0.351 0.383 3.176 3.387 dbcsr_complete_redistribute 329 12.2 0.274 0.280 2.205 3.053 density_rs2pw 119 9.7 0.003 0.003 2.876 2.888 fft_wrap_pw1pw2_140 487 13.2 0.648 0.652 2.863 2.880 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.789 2.792 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.809 2.659 multiply_cannon_metrocomm3 9144 15.5 0.020 0.021 1.672 2.603 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.310 2.579 apply_single 119 13.6 0.000 0.000 2.310 2.578 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.440 2.496 fft3d_ps 1201 14.6 0.871 0.890 2.417 2.424 calculate_dm_sparse 119 9.5 0.000 0.000 2.409 2.419 mp_alltoall_i22 627 13.8 1.565 2.416 1.565 2.416 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.385 2.386 mp_sum_l 7287 12.8 1.449 2.385 1.449 2.385 calculate_first_density_matrix 1 7.0 0.000 0.000 2.337 2.340 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.496 2.332 ot_diis_step 108 11.5 0.014 0.015 2.190 2.191 grid_integrate_task_list 119 12.3 2.171 2.185 2.171 2.185 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.106 2.107 cp_fm_cholesky_invert 11 10.9 2.054 2.057 2.054 2.057 hybrid_alltoall_any 4725 16.4 0.090 0.149 1.586 2.040 make_images_data 4572 15.5 0.046 0.049 1.619 2.016 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.978 1.978 cp_fm_diag_elpa_base 50 14.0 1.819 1.878 1.976 1.976 potential_pw2rs 119 12.3 0.014 0.014 1.931 1.942 acc_transpose_blocks 9144 15.5 0.045 0.046 1.888 1.938 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.805 1.806 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.742 1.749 wfi_extrapolate 11 7.9 0.001 0.001 1.699 1.699 grid_collocate_task_list 119 9.7 1.585 1.594 1.585 1.594 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.499 1.548 mp_alltoall_d11v 2130 13.8 1.478 1.497 1.478 1.497 mp_alltoall_z22v 1201 16.6 1.408 1.440 1.408 1.440 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.324 1.331 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.130 1.147 acc_transpose_blocks_sync 27432 16.5 1.105 1.136 1.105 1.136 transfer_rs2pw 487 10.6 0.005 0.005 1.072 1.086 jit_kernel_multiply 6 15.8 0.807 1.075 0.807 1.075 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.054 1.069 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.989 1.017 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=49.570000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=731.181818, yerr=15.665934 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/845a3ae33c0b6c5fc8bf8a28d823ca3539bb7d75_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 502.599680E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66212. MP_Allreduce 9776 488. MP_Sync 52 MP_Alltoall 1938 1383689. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.038 82.335 82.338 qs_mol_dyn_low 1 2.0 0.004 0.010 81.913 81.922 qs_forces 11 3.9 0.003 0.003 81.677 81.679 qs_energies 11 4.9 0.002 0.003 78.818 78.837 scf_env_do_scf 11 5.9 0.000 0.001 69.848 69.850 scf_env_do_scf_inner_loop 99 6.5 0.002 0.008 64.385 64.385 dbcsr_multiply_generic 2055 12.4 0.111 0.133 51.249 51.500 qs_scf_new_mos 99 7.5 0.000 0.001 47.387 47.507 qs_scf_loop_do_ot 99 8.5 0.001 0.001 47.386 47.507 ot_scf_mini 99 9.5 0.002 0.002 45.036 45.136 multiply_cannon 2055 13.4 0.184 0.190 42.900 43.620 velocity_verlet 10 3.0 0.002 0.004 43.366 43.369 multiply_cannon_loop 2055 14.4 1.780 1.822 41.939 42.709 ot_mini 99 10.5 0.001 0.001 26.591 26.684 qs_ot_get_derivative 99 11.5 0.001 0.001 19.771 19.867 multiply_cannon_multrec 49320 15.4 11.429 12.087 17.585 18.249 rebuild_ks_matrix 110 8.3 0.000 0.001 14.510 14.624 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.016 14.509 14.623 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.723 12.830 mp_waitall_1 220248 16.4 10.978 11.707 10.978 11.707 multiply_cannon_sync_h2d 49320 15.4 9.646 10.366 9.646 10.366 qs_ot_get_p 110 10.4 0.001 0.002 9.909 10.017 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.570 8.153 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.281 7.815 apply_single 110 13.6 0.000 0.001 7.281 7.815 multiply_cannon_metrocomm3 49320 15.4 0.081 0.084 6.671 7.627 sum_up_and_integrate 110 10.3 0.002 0.003 6.984 6.996 integrate_v_rspace 110 11.3 0.003 0.003 6.959 6.977 init_scf_run 11 5.9 0.000 0.001 6.899 6.899 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.899 6.899 ot_diis_step 99 11.5 0.006 0.006 6.622 6.622 qs_ot_p2m_diag 48 11.0 0.013 0.020 6.570 6.586 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.281 6.442 calculate_rho_elec 110 8.6 0.020 0.024 6.281 6.441 dbcsr_mm_accdrv_process 87628 16.1 3.060 3.135 6.025 6.284 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.776 5.776 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.424 5.481 init_scf_loop 11 6.9 0.000 0.001 5.436 5.437 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.192 5.217 cp_fm_diag_elpa_base 48 14.0 5.179 5.203 5.190 5.216 mp_sum_l 6594 12.7 3.921 4.624 3.921 4.624 make_m2s 4110 13.4 0.060 0.065 3.978 4.081 wfi_extrapolate 11 7.9 0.001 0.001 4.040 4.040 make_images 4110 14.4 0.177 0.189 3.883 3.989 calculate_dm_sparse 110 9.5 0.001 0.001 3.732 3.838 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.702 3.706 density_rs2pw 110 9.6 0.004 0.004 3.306 3.518 multiply_cannon_metrocomm1 49320 15.4 0.065 0.067 2.414 3.487 grid_integrate_task_list 110 12.3 3.218 3.430 3.218 3.430 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.354 3.413 prepare_preconditioner 11 7.9 0.000 0.000 3.260 3.276 make_preconditioner 11 8.9 0.000 0.000 3.260 3.276 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.192 3.236 pw_transfer 1331 11.6 0.054 0.066 3.064 3.141 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.046 3.087 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 2.976 3.056 calculate_first_density_matrix 1 7.0 0.000 0.000 2.760 2.765 jit_kernel_multiply 13 15.9 2.688 2.719 2.688 2.719 fft_wrap_pw1pw2_140 451 13.1 0.310 0.332 2.544 2.632 acc_transpose_blocks 49320 15.4 0.212 0.219 2.557 2.624 potential_pw2rs 110 12.3 0.005 0.006 2.579 2.598 mp_alltoall_d11v 2046 13.8 2.114 2.489 2.114 2.489 fft3d_ps 1111 14.6 0.799 0.889 2.270 2.330 grid_collocate_task_list 110 9.6 2.156 2.273 2.156 2.273 mp_waitany 14300 13.8 1.847 2.154 1.847 2.154 transfer_rs2pw 451 10.6 0.005 0.006 1.875 2.071 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.917 1.939 make_images_data 4110 15.4 0.043 0.047 1.810 1.928 mp_sum_d 3891 11.9 1.348 1.856 1.348 1.856 hybrid_alltoall_any 4261 16.3 0.083 0.483 1.559 1.837 cp_fm_cholesky_invert 11 10.9 1.791 1.794 1.791 1.794 transfer_pw2rs 451 13.1 0.005 0.006 1.746 1.756 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.669 1.702 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=82.338000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=477.363636, yerr=2.568081 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/845a3ae33c0b6c5fc8bf8a28d823ca3539bb7d75_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 593.469440E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1972800 MPI messages size (bytes): total size 1.077520E+12 min size 0.000000E+00 max size 4.537280E+06 average size 546.188250E+03 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 222984 1826684928 8192 < size <= 32768 520356 13399818240 32768 < size <= 131072 372336 35386294272 131072 < size <= 4194304 787758 788321309808 4194304 < size <= 16777216 54450 238588003280 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66430. MP_Allreduce 9775 566. MP_Sync 52 MP_Alltoall 1717 2309006. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.008 0.027 69.426 69.428 qs_mol_dyn_low 1 2.0 0.003 0.004 69.209 69.219 qs_forces 11 3.9 0.003 0.003 69.141 69.142 qs_energies 11 4.9 0.001 0.002 65.820 65.823 scf_env_do_scf 11 5.9 0.000 0.001 57.146 57.150 scf_env_do_scf_inner_loop 99 6.5 0.002 0.008 49.431 49.432 dbcsr_multiply_generic 2055 12.4 0.115 0.119 38.200 38.362 velocity_verlet 10 3.0 0.001 0.002 36.321 36.322 qs_scf_new_mos 99 7.5 0.001 0.001 33.397 33.512 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.396 33.512 multiply_cannon 2055 13.4 0.224 0.245 31.428 32.650 ot_scf_mini 99 9.5 0.003 0.003 31.725 31.825 multiply_cannon_loop 2055 14.4 1.157 1.185 30.143 31.059 ot_mini 99 10.5 0.001 0.001 18.484 18.593 multiply_cannon_multrec 24660 15.4 7.000 8.664 14.137 15.742 rebuild_ks_matrix 110 8.3 0.000 0.000 13.626 13.707 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.015 13.625 13.707 qs_ot_get_derivative 99 11.5 0.001 0.001 12.698 12.806 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.994 12.065 mp_waitall_1 176588 16.5 7.610 10.008 7.610 10.008 multiply_cannon_metrocomm3 24660 15.4 0.073 0.074 5.101 7.936 init_scf_loop 11 6.9 0.000 0.001 7.680 7.680 multiply_cannon_sync_h2d 24660 15.4 6.420 7.379 6.420 7.379 dbcsr_mm_accdrv_process 52282 16.1 5.606 6.528 6.968 7.254 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.462 7.137 apply_single 110 13.6 0.000 0.001 6.462 7.137 qs_ot_get_p 110 10.4 0.001 0.001 6.570 6.708 sum_up_and_integrate 110 10.3 0.001 0.002 6.443 6.456 integrate_v_rspace 110 11.3 0.002 0.003 6.416 6.428 init_scf_run 11 5.9 0.000 0.002 6.290 6.291 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.290 6.291 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.024 6.033 calculate_rho_elec 110 8.6 0.039 0.047 6.024 6.032 ot_diis_step 99 11.5 0.010 0.011 5.738 5.738 prepare_preconditioner 11 7.9 0.000 0.000 5.639 5.663 make_preconditioner 11 8.9 0.000 0.000 5.639 5.663 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.734 5.462 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.221 5.375 make_m2s 4110 13.4 0.056 0.058 4.379 4.854 make_images 4110 14.4 0.404 0.466 4.267 4.737 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.639 4.660 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.176 4.177 pw_transfer 1331 11.6 0.065 0.073 3.647 3.792 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.541 3.687 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.598 3.611 cp_fm_diag_elpa_base 48 14.0 3.547 3.568 3.595 3.608 wfi_extrapolate 11 7.9 0.001 0.001 3.559 3.559 density_rs2pw 110 9.6 0.004 0.004 3.251 3.450 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.399 3.401 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.318 3.372 grid_integrate_task_list 110 12.3 3.159 3.303 3.159 3.303 fft_wrap_pw1pw2_140 451 13.1 0.358 0.378 2.989 3.132 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.050 3.079 calculate_dm_sparse 110 9.5 0.001 0.001 2.983 3.010 make_images_data 4110 15.4 0.048 0.052 2.376 2.868 fft3d_ps 1111 14.6 1.116 1.350 2.634 2.794 hybrid_alltoall_any 4261 16.3 0.104 0.448 2.067 2.741 calculate_first_density_matrix 1 7.0 0.000 0.000 2.634 2.636 cp_fm_cholesky_invert 11 10.9 2.603 2.610 2.603 2.610 grid_collocate_task_list 110 9.6 2.170 2.328 2.170 2.328 potential_pw2rs 110 12.3 0.008 0.009 2.286 2.297 mp_sum_l 6594 12.7 1.761 2.293 1.761 2.293 mp_alltoall_d11v 2046 13.8 1.813 2.007 1.813 2.007 acc_transpose_blocks 24660 15.4 0.118 0.121 1.969 1.990 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.969 1.988 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.872 1.873 jit_kernel_multiply 9 16.2 0.996 1.849 0.996 1.849 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.787 1.798 transfer_rs2pw 451 10.6 0.006 0.007 1.485 1.667 mp_allgather_i34 2055 14.4 0.596 1.640 0.596 1.640 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.545 1.556 multiply_cannon_metrocomm4 22605 15.4 0.073 0.076 0.758 1.487 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.385 1.486 dbcsr_complete_redistribute 325 12.2 0.238 0.298 1.171 1.434 cp_fm_cholesky_decompose 22 10.9 1.390 1.397 1.390 1.397 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=69.428000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=561.181818, yerr=6.779892 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/845a3ae33c0b6c5fc8bf8a28d823ca3539bb7d75_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 668.676096E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66421. MP_Allreduce 9774 562. MP_Sync 52 MP_Alltoall 1496 4511006. MP_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.069 60.770 60.771 qs_mol_dyn_low 1 2.0 0.003 0.005 60.386 60.397 qs_forces 11 3.9 0.003 0.007 60.301 60.302 qs_energies 11 4.9 0.019 0.097 56.998 57.030 scf_env_do_scf 11 5.9 0.001 0.002 48.889 48.890 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 40.174 40.175 velocity_verlet 10 3.0 0.004 0.022 32.758 32.761 dbcsr_multiply_generic 2055 12.4 0.109 0.112 28.810 29.053 qs_scf_new_mos 99 7.5 0.001 0.001 25.377 25.463 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.376 25.462 ot_scf_mini 99 9.5 0.003 0.004 24.124 24.237 multiply_cannon 2055 13.4 0.213 0.222 22.582 23.729 multiply_cannon_loop 2055 14.4 0.814 0.834 21.374 22.444 ot_mini 99 10.5 0.001 0.002 13.774 13.886 rebuild_ks_matrix 110 8.3 0.000 0.000 12.311 12.456 qs_ks_build_kohn_sham_matrix 110 9.3 0.018 0.041 12.310 12.456 multiply_cannon_multrec 16440 15.4 3.639 4.798 9.821 11.016 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.758 10.890 mp_waitall_1 139946 16.5 6.955 9.995 6.955 9.995 qs_ot_get_derivative 99 11.5 0.001 0.009 9.326 9.435 init_scf_loop 11 6.9 0.001 0.005 8.679 8.683 multiply_cannon_metrocomm3 16440 15.4 0.045 0.047 4.395 7.224 prepare_preconditioner 11 7.9 0.000 0.000 6.861 6.879 make_preconditioner 11 8.9 0.000 0.002 6.861 6.879 make_full_inverse_cholesky 11 9.9 0.000 0.001 6.204 6.561 sum_up_and_integrate 110 10.3 0.002 0.003 6.394 6.416 integrate_v_rspace 110 11.3 0.003 0.003 6.367 6.390 dbcsr_mm_accdrv_process 34862 16.1 5.382 5.720 6.027 6.202 qs_rho_update_rho_low 110 7.6 0.001 0.002 5.829 5.841 calculate_rho_elec 110 8.6 0.059 0.060 5.828 5.840 qs_ot_get_p 110 10.4 0.001 0.002 5.539 5.685 init_scf_run 11 5.9 0.000 0.001 5.565 5.565 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.565 5.565 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.888 5.398 apply_single 110 13.6 0.000 0.000 4.888 5.397 make_m2s 4110 13.4 0.049 0.051 4.165 4.568 make_images 4110 14.4 0.396 0.518 4.048 4.449 ot_diis_step 99 11.5 0.011 0.013 4.420 4.420 multiply_cannon_sync_h2d 16440 15.4 3.221 3.918 3.221 3.918 qs_ot_p2m_diag 48 11.0 0.042 0.045 3.848 3.852 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.234 3.846 pw_transfer 1331 11.6 0.065 0.072 3.734 3.752 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.628 3.648 cp_dbcsr_syevd 48 12.0 0.003 0.005 3.474 3.474 grid_integrate_task_list 110 12.3 3.203 3.413 3.203 3.413 density_rs2pw 110 9.6 0.004 0.004 3.035 3.211 fft_wrap_pw1pw2_140 451 13.1 0.460 0.470 3.124 3.148 wfi_extrapolate 11 7.9 0.001 0.004 2.984 2.984 cp_fm_diag_elpa 48 13.0 0.000 0.001 2.920 2.932 cp_fm_diag_elpa_base 48 14.0 2.852 2.888 2.916 2.929 hybrid_alltoall_any 4261 16.3 0.107 0.378 2.062 2.828 make_images_data 4110 15.4 0.046 0.051 2.283 2.823 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.776 2.777 cp_fm_cholesky_invert 11 10.9 2.696 2.703 2.696 2.703 calculate_dm_sparse 110 9.5 0.001 0.001 2.589 2.615 fft3d_ps 1111 14.6 1.120 1.132 2.546 2.565 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.463 2.520 calculate_first_density_matrix 1 7.0 0.001 0.005 2.488 2.490 multiply_cannon_metrocomm4 14385 15.4 0.047 0.050 0.903 2.484 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.433 2.477 grid_collocate_task_list 110 9.6 2.220 2.393 2.220 2.393 mp_irecv_dv 48980 15.7 0.829 2.352 0.829 2.352 potential_pw2rs 110 12.3 0.010 0.011 2.211 2.225 mp_alltoall_d11v 2046 13.8 1.873 2.147 1.873 2.147 mp_sum_l 6594 12.7 1.443 2.114 1.443 2.114 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.982 1.984 dbcsr_complete_redistribute 325 12.2 0.318 0.345 1.439 1.894 cp_fm_upper_to_full 70 14.2 1.397 1.763 1.397 1.763 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.747 1.761 acc_transpose_blocks 16440 15.4 0.079 0.082 1.584 1.647 cp_fm_cholesky_decompose 22 10.9 1.619 1.638 1.619 1.638 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.525 1.537 mp_allgather_i34 2055 14.4 0.524 1.494 0.524 1.494 transfer_rs2pw 451 10.6 0.005 0.006 1.306 1.493 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.007 1.364 1.461 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.972 1.416 mp_waitany 17072 13.8 1.150 1.372 1.150 1.372 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.283 1.290 mp_alltoall_z22v 1111 16.6 1.213 1.242 1.213 1.242 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=60.771000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=633.818182, yerr=7.553665 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/845a3ae33c0b6c5fc8bf8a28d823ca3539bb7d75_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 740.151296E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66419. MP_Allreduce 9774 603. MP_Sync 52 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.009 0.029 65.478 65.480 qs_mol_dyn_low 1 2.0 0.003 0.004 65.255 65.265 qs_forces 11 3.9 0.003 0.003 65.105 65.105 qs_energies 11 4.9 0.001 0.002 61.722 61.725 scf_env_do_scf 11 5.9 0.000 0.001 53.520 53.523 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 41.956 41.957 velocity_verlet 10 3.0 0.002 0.002 37.251 37.257 dbcsr_multiply_generic 2055 12.4 0.117 0.122 30.529 30.738 qs_scf_new_mos 99 7.5 0.001 0.001 27.359 27.459 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.358 27.459 ot_scf_mini 99 9.5 0.002 0.003 25.667 25.768 multiply_cannon 2055 13.4 0.273 0.293 23.488 24.713 multiply_cannon_loop 2055 14.4 1.424 1.490 22.117 22.829 ot_mini 99 10.5 0.001 0.001 14.800 14.926 multiply_cannon_multrec 24660 15.4 4.082 6.762 12.999 14.176 rebuild_ks_matrix 110 8.3 0.000 0.000 11.929 12.029 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.928 12.029 init_scf_loop 11 6.9 0.001 0.001 11.524 11.524 qs_ot_get_derivative 99 11.5 0.001 0.001 10.630 10.735 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.536 10.627 prepare_preconditioner 11 7.9 0.000 0.000 9.802 9.815 make_preconditioner 11 8.9 0.000 0.000 9.802 9.815 dbcsr_mm_accdrv_process 52304 16.0 7.699 8.936 8.759 9.665 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.056 9.501 sum_up_and_integrate 110 10.3 0.001 0.002 6.228 6.242 integrate_v_rspace 110 11.3 0.003 0.003 6.202 6.216 qs_ot_get_p 110 10.4 0.001 0.001 5.776 5.929 mp_waitall_1 121746 16.5 4.154 5.892 4.154 5.892 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.861 5.871 calculate_rho_elec 110 8.6 0.077 0.081 5.861 5.871 make_m2s 4110 13.4 0.059 0.061 5.407 5.672 make_images 4110 14.4 0.572 0.690 5.266 5.527 init_scf_run 11 5.9 0.000 0.001 5.472 5.472 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.471 5.472 cp_fm_upper_to_full 70 14.2 3.331 4.817 3.331 4.817 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.051 4.128 apply_single 110 13.6 0.000 0.000 4.051 4.127 ot_diis_step 99 11.5 0.011 0.011 4.127 4.127 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.936 3.955 pw_transfer 1331 11.6 0.065 0.072 3.780 3.813 dbcsr_complete_redistribute 325 12.2 0.413 0.463 2.610 3.739 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.674 3.710 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.513 3.564 grid_integrate_task_list 110 12.3 3.263 3.438 3.263 3.438 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.369 3.423 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.366 3.367 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.090 3.207 fft_wrap_pw1pw2_140 451 13.1 0.530 0.545 3.154 3.191 density_rs2pw 110 9.6 0.004 0.004 2.990 3.158 multiply_cannon_metrocomm3 24660 15.4 0.037 0.038 1.416 3.073 calculate_dm_sparse 110 9.5 0.001 0.001 2.990 3.030 wfi_extrapolate 11 7.9 0.001 0.001 2.972 2.972 make_images_data 4110 15.4 0.048 0.052 2.670 2.964 hybrid_alltoall_any 4261 16.3 0.122 0.457 2.302 2.864 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.737 2.839 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.829 2.838 cp_fm_diag_elpa_base 48 14.0 2.682 2.739 2.827 2.836 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.796 2.798 mp_alltoall_i22 605 13.7 1.632 2.788 1.632 2.788 cp_fm_cholesky_invert 11 10.9 2.683 2.691 2.683 2.691 acc_transpose_blocks 24660 15.4 0.113 0.117 2.423 2.576 multiply_cannon_sync_h2d 24660 15.4 2.369 2.560 2.369 2.560 fft3d_ps 1111 14.6 1.153 1.192 2.501 2.516 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.449 2.486 grid_collocate_task_list 110 9.6 2.263 2.459 2.263 2.459 calculate_first_density_matrix 1 7.0 0.000 0.000 2.404 2.408 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.236 2.236 potential_pw2rs 110 12.3 0.012 0.013 2.038 2.044 mp_alltoall_d11v 2046 13.8 1.774 2.034 1.774 2.034 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.810 1.847 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.609 1.706 cp_fm_cholesky_decompose 22 10.9 1.630 1.672 1.630 1.672 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.632 1.643 mp_sum_l 6594 12.7 0.945 1.608 0.945 1.608 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.541 1.559 acc_transpose_blocks_sync 73980 16.4 1.408 1.550 1.408 1.550 mp_allgather_i34 2055 14.4 0.427 1.516 0.427 1.516 multiply_cannon_metrocomm4 20550 15.4 0.059 0.061 0.856 1.496 mp_irecv_dv 62702 16.1 0.754 1.418 0.754 1.418 transfer_rs2pw 451 10.6 0.005 0.005 1.174 1.366 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=65.480000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=700.545455, yerr=10.084765 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/845a3ae33c0b6c5fc8bf8a28d823ca3539bb7d75_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 858.255360E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66417. MP_Allreduce 9774 644. MP_Sync 52 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.032 62.117 62.118 qs_mol_dyn_low 1 2.0 0.003 0.004 61.842 61.855 qs_forces 11 3.9 0.003 0.003 61.754 61.755 qs_energies 11 4.9 0.001 0.002 58.002 58.007 scf_env_do_scf 11 5.9 0.000 0.001 49.322 49.322 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 39.655 39.656 velocity_verlet 10 3.0 0.023 0.028 34.848 34.855 dbcsr_multiply_generic 2055 12.4 0.108 0.110 24.748 24.952 qs_scf_new_mos 99 7.5 0.001 0.001 23.955 24.035 qs_scf_loop_do_ot 99 8.5 0.001 0.001 23.955 24.034 ot_scf_mini 99 9.5 0.002 0.002 22.603 22.642 multiply_cannon 2055 13.4 0.242 0.256 18.201 20.020 multiply_cannon_loop 2055 14.4 0.603 0.629 16.843 17.466 rebuild_ks_matrix 110 8.3 0.000 0.000 12.331 12.393 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.020 12.331 12.393 ot_mini 99 10.5 0.001 0.001 11.846 11.878 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.970 11.029 init_scf_loop 11 6.9 0.001 0.001 9.609 9.611 mp_waitall_1 103326 16.6 6.978 8.541 6.978 8.541 multiply_cannon_multrec 8220 15.4 3.177 4.319 7.570 8.392 qs_ot_get_derivative 99 11.5 0.001 0.001 8.071 8.107 prepare_preconditioner 11 7.9 0.000 0.000 7.870 7.876 make_preconditioner 11 8.9 0.000 0.000 7.870 7.876 make_full_inverse_cholesky 11 9.9 0.000 0.000 7.408 7.515 sum_up_and_integrate 110 10.3 0.001 0.002 6.741 6.754 integrate_v_rspace 110 11.3 0.003 0.003 6.714 6.727 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.627 6.637 calculate_rho_elec 110 8.6 0.112 0.113 6.626 6.636 qs_ot_get_p 110 10.4 0.001 0.001 6.582 6.620 init_scf_run 11 5.9 0.000 0.001 5.439 5.439 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.439 5.439 dbcsr_mm_accdrv_process 17442 15.9 3.112 4.297 4.254 5.175 make_m2s 4110 13.4 0.037 0.038 4.884 5.173 make_images 4110 14.4 0.643 0.705 4.753 5.042 qs_ot_p2m_diag 48 11.0 0.081 0.085 4.885 4.893 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 3.050 4.711 pw_transfer 1331 11.6 0.065 0.069 4.515 4.534 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.491 4.493 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.407 4.431 cp_fm_cholesky_invert 11 10.9 4.182 4.188 4.182 4.188 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.767 3.863 apply_single 110 13.6 0.000 0.000 3.767 3.863 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.764 3.772 cp_fm_diag_elpa_base 48 14.0 3.690 3.720 3.761 3.769 fft_wrap_pw1pw2_140 451 13.1 0.716 0.729 3.710 3.735 density_rs2pw 110 9.6 0.004 0.004 3.565 3.733 ot_diis_step 99 11.5 0.012 0.012 3.728 3.728 grid_integrate_task_list 110 12.3 3.383 3.496 3.383 3.496 hybrid_alltoall_any 4261 16.3 0.198 0.841 2.573 3.135 make_images_data 4110 15.4 0.040 0.046 2.631 3.111 wfi_extrapolate 11 7.9 0.001 0.001 3.087 3.087 fft3d_ps 1111 14.6 1.295 1.302 2.957 2.977 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.681 2.681 calculate_dm_sparse 110 9.5 0.001 0.001 2.586 2.634 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.567 2.568 grid_collocate_task_list 110 9.6 2.363 2.554 2.363 2.554 potential_pw2rs 110 12.3 0.015 0.015 2.455 2.467 multiply_cannon_sync_h2d 8220 15.4 2.361 2.466 2.361 2.466 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.351 2.380 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.308 2.343 calculate_first_density_matrix 1 7.0 0.000 0.000 2.244 2.246 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.154 2.183 cp_fm_cholesky_decompose 22 10.9 2.146 2.164 2.146 2.164 mp_alltoall_d11v 2046 13.8 1.992 2.154 1.992 2.154 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.786 1.997 dbcsr_complete_redistribute 325 12.2 0.538 0.559 1.641 1.759 mp_allgather_i34 2055 14.4 0.529 1.740 0.529 1.740 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.689 1.695 qs_env_update_s_mstruct 11 6.9 0.002 0.011 1.514 1.631 transfer_rs2pw 451 10.6 0.005 0.005 1.440 1.612 jit_kernel_multiply 9 15.5 0.826 1.522 0.826 1.522 multiply_cannon_metrocomm1 8220 15.4 0.022 0.023 1.210 1.522 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.454 1.467 mp_alltoall_z22v 1111 16.6 1.425 1.449 1.425 1.449 acc_transpose_blocks 8220 15.4 0.041 0.042 1.365 1.397 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.325 1.352 mp_waitany 9240 13.8 1.155 1.344 1.155 1.344 qs_create_task_list 11 7.9 0.000 0.000 1.216 1.314 generate_qs_task_list 11 8.9 0.374 0.442 1.216 1.314 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=62.118000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=809.181818, yerr=12.156691 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/845a3ae33c0b6c5fc8bf8a28d823ca3539bb7d75_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 1.399128E+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 67098. MP_Allreduce 9752 812. MP_Sync 52 MP_Alltoall 1474 16505187. MP_SendRecv 2310 360267. MP_ISendRecv 2310 360267. MP_Wait 5214 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.033 0.040 98.127 98.128 qs_mol_dyn_low 1 2.0 0.003 0.005 97.698 97.710 qs_forces 11 3.9 0.003 0.003 97.622 97.623 qs_energies 11 4.9 0.002 0.002 93.242 93.243 scf_env_do_scf 11 5.9 0.000 0.001 82.683 82.684 velocity_verlet 10 3.0 0.002 0.002 61.953 61.960 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 51.132 51.133 dbcsr_multiply_generic 2055 12.4 0.123 0.127 32.370 32.532 init_scf_loop 11 6.9 0.001 0.001 31.466 31.468 qs_scf_new_mos 99 7.5 0.001 0.001 31.417 31.452 qs_scf_loop_do_ot 99 8.5 0.001 0.001 31.416 31.452 ot_scf_mini 99 9.5 0.002 0.002 29.454 29.488 prepare_preconditioner 11 7.9 0.000 0.000 29.280 29.290 make_preconditioner 11 8.9 0.000 0.000 29.280 29.290 make_full_inverse_cholesky 11 9.9 0.000 0.000 23.419 28.753 multiply_cannon 2055 13.4 0.340 0.378 23.889 25.043 multiply_cannon_loop 2055 14.4 0.835 0.848 21.782 22.181 cp_fm_upper_to_full 70 14.2 12.748 18.326 12.748 18.326 ot_mini 99 10.5 0.001 0.001 16.087 16.122 rebuild_ks_matrix 110 8.3 0.001 0.001 15.151 15.216 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 15.151 15.215 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.725 13.785 qs_ot_get_derivative 99 11.5 0.001 0.001 11.109 11.144 dbcsr_complete_redistribute 325 12.2 1.004 1.041 7.641 10.862 mp_waitall_1 84994 16.7 9.614 10.661 9.614 10.661 multiply_cannon_multrec 8220 15.4 4.076 4.313 9.762 9.866 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.507 9.732 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.845 9.040 mp_alltoall_i22 605 13.7 5.494 8.734 5.494 8.734 qs_rho_update_rho_low 110 7.6 0.001 0.001 8.382 8.407 calculate_rho_elec 110 8.6 0.222 0.223 8.381 8.407 sum_up_and_integrate 110 10.3 0.002 0.002 7.747 7.761 integrate_v_rspace 110 11.3 0.003 0.003 7.719 7.733 qs_ot_get_p 110 10.4 0.001 0.001 7.580 7.617 cp_fm_cholesky_invert 11 10.9 7.387 7.393 7.387 7.393 make_m2s 4110 13.4 0.042 0.043 6.307 6.982 make_images 4110 14.4 0.877 0.935 6.115 6.787 pw_transfer 1331 11.6 0.074 0.074 6.588 6.593 fft_wrap_pw1pw2 1111 12.6 0.009 0.010 6.472 6.477 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 5.922 6.365 init_scf_run 11 5.9 0.000 0.001 6.160 6.160 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.160 6.160 dbcsr_mm_accdrv_process 11614 15.7 3.958 4.625 5.535 5.743 qs_ot_p2m_diag 48 11.0 0.151 0.156 5.558 5.567 fft_wrap_pw1pw2_140 451 13.1 1.363 1.367 5.527 5.533 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.086 5.498 apply_single 110 13.6 0.000 0.000 5.086 5.498 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.958 4.959 ot_diis_step 99 11.5 0.015 0.016 4.934 4.934 density_rs2pw 110 9.6 0.004 0.004 4.703 4.722 make_images_data 4110 15.4 0.045 0.048 3.485 4.474 hybrid_alltoall_any 4261 16.3 0.264 0.566 3.387 4.389 fft3d_ps 1111 14.6 1.887 1.894 4.180 4.186 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.158 4.158 cp_fm_diag_elpa_base 48 14.0 3.593 3.798 4.155 4.155 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.454 3.931 grid_integrate_task_list 110 12.3 3.703 3.755 3.703 3.755 wfi_extrapolate 11 7.9 0.001 0.001 3.710 3.710 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.702 3.703 calculate_dm_sparse 110 9.5 0.001 0.001 3.398 3.431 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.148 3.171 multiply_cannon_sync_h2d 8220 15.4 3.115 3.134 3.115 3.134 potential_pw2rs 110 12.3 0.021 0.021 3.084 3.093 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.999 3.003 cp_fm_cholesky_decompose 22 10.9 2.790 2.809 2.790 2.809 grid_collocate_task_list 110 9.6 2.689 2.718 2.689 2.718 mp_alltoall_d11v 2046 13.8 2.554 2.598 2.554 2.598 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.428 2.452 calculate_first_density_matrix 1 7.0 0.000 0.000 2.316 2.318 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.222 2.292 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.108 2.202 mp_allgather_i34 2055 14.4 0.864 2.138 0.864 2.138 mp_alltoall_z22v 1111 16.6 2.041 2.047 2.041 2.047 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.028 2.032 qs_ot_get_orbitals 99 10.5 0.000 0.000 1.949 1.963 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=98.128000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1248.090909, yerr=61.126478 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/845a3ae33c0b6c5fc8bf8a28d823ca3539bb7d75_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 631.050240E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175954870160 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 57905. MP_Allreduce 11059 797. MP_Sync 87 MP_Alltoall 2226 1873266. MP_SendRecv 24320 18752. MP_ISendRecv 24320 18752. MP_Wait 42476 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.043 0.044 221.428 221.429 qs_mol_dyn_low 1 2.0 0.003 0.005 220.946 220.962 qs_forces 11 3.9 0.005 0.005 220.844 220.845 qs_energies 11 4.9 0.002 0.002 215.030 215.053 scf_env_do_scf 11 5.9 0.001 0.001 198.059 198.063 scf_env_do_scf_inner_loop 117 6.6 0.003 0.007 174.165 174.167 qs_scf_new_mos 117 7.6 0.001 0.001 132.415 132.621 qs_scf_loop_do_ot 117 8.6 0.001 0.001 132.414 132.620 velocity_verlet 10 3.0 0.001 0.002 131.856 131.859 dbcsr_multiply_generic 2507 12.6 0.177 0.182 126.832 127.325 ot_scf_mini 117 9.6 0.003 0.003 125.696 125.849 multiply_cannon 2507 13.6 0.238 0.249 101.983 103.400 multiply_cannon_loop 2507 14.6 2.392 2.464 99.537 100.943 ot_mini 117 10.6 0.001 0.001 67.706 67.861 multiply_cannon_multrec 60168 15.6 31.629 33.486 41.390 43.271 qs_ot_get_derivative 117 11.6 0.001 0.001 43.013 43.156 qs_ot_get_p 128 10.4 0.001 0.001 35.773 36.010 rebuild_ks_matrix 128 8.3 0.001 0.001 34.720 34.967 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 34.719 34.966 mp_waitall_1 267128 16.5 30.335 33.206 30.335 33.206 qs_ks_update_qs_env 128 7.6 0.001 0.001 31.181 31.402 qs_ot_p2m_diag 83 11.4 0.079 0.092 28.697 28.754 multiply_cannon_sync_h2d 60168 15.6 26.251 27.890 26.251 27.890 cp_dbcsr_syevd 83 12.4 0.005 0.005 25.792 25.794 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.252 24.970 apply_single 128 13.6 0.001 0.001 24.252 24.970 ot_diis_step 117 11.6 0.008 0.008 24.445 24.446 init_scf_loop 11 6.9 0.000 0.001 23.817 23.819 cp_fm_diag_elpa 83 13.4 0.000 0.000 22.081 22.110 cp_fm_diag_elpa_base 83 14.4 21.961 22.024 22.075 22.105 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 20.777 20.937 prepare_preconditioner 11 7.9 0.000 0.000 19.163 19.211 make_preconditioner 11 8.9 0.000 0.000 19.163 19.211 make_full_inverse_cholesky 11 9.9 0.000 0.000 18.395 18.596 multiply_cannon_metrocomm3 60168 15.6 0.114 0.122 16.188 17.745 sum_up_and_integrate 128 10.3 0.002 0.005 15.203 15.220 integrate_v_rspace 128 11.3 0.003 0.004 15.145 15.165 make_m2s 5014 13.6 0.104 0.112 14.569 14.961 make_images 5014 14.6 0.399 0.416 14.383 14.781 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.599 14.717 calculate_rho_elec 128 8.7 0.045 0.064 14.598 14.717 init_scf_run 11 5.9 0.000 0.001 12.714 12.714 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.713 12.714 cp_fm_cholesky_invert 11 10.9 11.228 11.238 11.228 11.238 mp_sum_l 7950 12.9 9.023 10.235 9.023 10.235 dbcsr_mm_accdrv_process 124484 16.2 4.727 4.929 9.330 9.895 wfi_extrapolate 11 7.9 0.001 0.001 9.407 9.408 density_rs2pw 128 9.7 0.006 0.007 8.276 8.785 pw_transfer 1547 11.6 0.074 0.090 8.362 8.646 calculate_dm_sparse 128 9.5 0.001 0.001 8.532 8.613 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.446 8.574 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 8.160 8.440 multiply_cannon_metrocomm1 60168 15.6 0.094 0.098 6.593 8.375 qs_ot_get_orbitals 117 10.6 0.001 0.001 8.153 8.230 make_images_data 5014 15.6 0.066 0.071 6.951 7.894 grid_integrate_task_list 128 12.3 7.063 7.634 7.063 7.634 hybrid_alltoall_any 5200 16.5 0.293 2.271 6.011 7.234 fft_wrap_pw1pw2_140 523 13.2 0.851 0.897 6.553 6.820 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.652 6.663 fft3d_ps 1291 14.7 2.199 2.844 6.182 6.611 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.765 5.889 mp_alltoall_d11v 2415 14.1 4.380 5.792 4.380 5.792 cp_fm_cholesky_decompose 22 10.9 5.752 5.765 5.752 5.765 potential_pw2rs 128 12.3 0.009 0.010 5.676 5.744 grid_collocate_task_list 128 9.7 4.837 5.308 4.837 5.308 transfer_rs2pw 523 10.6 0.007 0.009 4.127 4.666 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=221.429000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=598.090909, yerr=6.006883 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/845a3ae33c0b6c5fc8bf8a28d823ca3539bb7d75_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 848.560128E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100942E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834332048 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58357. MP_Allreduce 11058 960. MP_Sync 87 MP_Alltoall 1969 5849724. MP_SendRecv 12032 47072. MP_ISendRecv 12032 47072. MP_Wait 25916 MP_ISend 11748 212467. MP_IRecv 11748 212467. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.031 194.265 194.265 qs_mol_dyn_low 1 2.0 0.003 0.004 193.877 193.892 qs_forces 11 3.9 0.004 0.005 193.795 193.796 qs_energies 11 4.9 0.002 0.002 187.005 187.019 scf_env_do_scf 11 5.9 0.001 0.001 168.871 168.881 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 135.651 135.653 velocity_verlet 10 3.0 0.001 0.002 120.611 120.613 dbcsr_multiply_generic 2507 12.6 0.187 0.194 99.541 100.937 qs_scf_new_mos 117 7.6 0.001 0.001 96.517 97.013 qs_scf_loop_do_ot 117 8.6 0.001 0.001 96.516 97.012 ot_scf_mini 117 9.6 0.004 0.004 91.690 92.261 multiply_cannon 2507 13.6 0.503 0.561 79.091 83.367 multiply_cannon_loop 2507 14.6 1.557 1.625 75.601 78.659 ot_mini 117 10.6 0.001 0.001 50.490 50.997 mp_waitall_1 214728 16.6 24.400 37.753 24.400 37.753 multiply_cannon_multrec 30084 15.6 21.134 25.018 33.041 37.669 rebuild_ks_matrix 128 8.3 0.001 0.001 32.609 33.365 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.020 32.609 33.364 init_scf_loop 11 6.9 0.000 0.001 33.129 33.130 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.337 30.004 qs_ot_get_derivative 117 11.6 0.001 0.002 28.547 29.121 prepare_preconditioner 11 7.9 0.000 0.000 28.761 28.836 make_preconditioner 11 8.9 0.000 0.000 28.761 28.836 multiply_cannon_metrocomm3 30084 15.6 0.098 0.103 15.728 28.776 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.449 27.987 qs_ot_get_p 128 10.4 0.001 0.001 23.136 23.751 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.112 23.233 apply_single 128 13.6 0.001 0.001 22.111 23.233 ot_diis_step 117 11.6 0.015 0.015 21.763 21.765 multiply_cannon_sync_h2d 30084 15.6 18.112 20.403 18.112 20.403 qs_ot_p2m_diag 83 11.4 0.189 0.217 18.088 18.124 cp_dbcsr_syevd 83 12.4 0.005 0.006 16.849 16.850 cp_fm_cholesky_invert 11 10.9 16.756 16.768 16.756 16.768 make_m2s 5014 13.6 0.086 0.094 14.383 16.116 make_images 5014 14.6 1.157 1.368 14.166 15.897 sum_up_and_integrate 128 10.3 0.002 0.004 14.217 14.256 integrate_v_rspace 128 11.3 0.003 0.003 14.157 14.200 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.707 13.752 calculate_rho_elec 128 8.7 0.087 0.103 13.707 13.751 cp_fm_diag_elpa 83 13.4 0.000 0.000 13.564 13.592 cp_fm_diag_elpa_base 83 14.4 13.301 13.405 13.558 13.581 init_scf_run 11 5.9 0.000 0.002 13.292 13.293 scf_env_initial_rho_setup 11 6.9 0.001 0.001 13.291 13.293 dbcsr_mm_accdrv_process 62242 16.2 5.561 7.199 11.356 12.056 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 11.330 11.699 multiply_cannon_metrocomm4 27577 15.6 0.099 0.113 3.808 10.815 make_images_data 5014 15.6 0.066 0.075 8.436 10.445 mp_irecv_dv 69486 16.3 3.610 10.416 3.610 10.416 hybrid_alltoall_any 5200 16.5 0.349 1.566 7.211 9.929 pw_transfer 1547 11.6 0.084 0.095 8.831 8.893 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.609 8.677 wfi_extrapolate 11 7.9 0.001 0.001 8.471 8.471 density_rs2pw 128 9.7 0.006 0.006 7.343 7.941 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 7.769 7.778 fft_wrap_pw1pw2_140 523 13.2 0.933 0.955 7.534 7.622 grid_integrate_task_list 128 12.3 7.154 7.568 7.154 7.568 cp_fm_cholesky_decompose 22 10.9 7.079 7.141 7.079 7.141 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.387 7.127 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 6.840 7.061 calculate_dm_sparse 128 9.5 0.001 0.001 6.571 6.725 mp_sum_l 7950 12.9 4.271 6.269 4.271 6.269 fft3d_ps 1291 14.7 2.828 3.015 6.169 6.223 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.392 5.466 grid_collocate_task_list 128 9.7 5.008 5.440 5.008 5.440 mp_allgather_i34 2507 14.6 2.019 5.172 2.019 5.172 mp_alltoall_d11v 2415 14.1 4.323 4.888 4.323 4.888 potential_pw2rs 128 12.3 0.015 0.017 4.683 4.707 calculate_first_density_matrix 1 7.0 0.000 0.000 4.632 4.634 dbcsr_complete_redistribute 395 12.7 0.761 0.845 3.256 4.124 mp_sum_d 4467 12.1 2.633 4.038 2.633 4.038 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=194.265000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=806.636364, yerr=4.205467 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/845a3ae33c0b6c5fc8bf8a28d823ca3539bb7d75_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 2.928533E+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 3984192 0.0% 0.0% 100.0% average stack size 0.0 0.0 1695.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 930.193408E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1042912 MPI messages size (bytes): total size 2.716210E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.604448E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 281856 36943429632 131072 < size <= 4194304 660064 996105256960 4194304 < size <= 16777216 65632 931531265168 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 58208. MP_Allreduce 11082 999. MP_Sync 87 MP_Alltoall 1712 9388896. MP_SendRecv 7936 75008. MP_ISendRecv 7936 75008. MP_Wait 21820 MP_ISend 11748 275205. MP_IRecv 11748 275205. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.029 177.035 177.038 qs_mol_dyn_low 1 2.0 0.003 0.004 176.646 176.662 qs_forces 11 3.9 0.004 0.004 176.548 176.551 qs_energies 11 4.9 0.001 0.002 170.012 170.022 scf_env_do_scf 11 5.9 0.001 0.001 153.745 153.748 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 118.358 118.359 velocity_verlet 10 3.0 0.002 0.004 113.077 113.080 dbcsr_multiply_generic 2507 12.6 0.184 0.188 82.880 83.908 qs_scf_new_mos 117 7.6 0.001 0.001 81.212 81.621 qs_scf_loop_do_ot 117 8.6 0.001 0.001 81.211 81.620 ot_scf_mini 117 9.6 0.003 0.004 76.954 77.408 multiply_cannon 2507 13.6 0.507 0.529 62.942 67.258 multiply_cannon_loop 2507 14.6 1.127 1.199 59.977 62.623 ot_mini 117 10.6 0.001 0.001 42.802 43.280 init_scf_loop 11 6.9 0.000 0.001 35.282 35.284 mp_waitall_1 170520 16.6 25.174 34.746 25.174 34.746 prepare_preconditioner 11 7.9 0.000 0.000 31.289 31.347 make_preconditioner 11 8.9 0.000 0.000 31.289 31.347 rebuild_ks_matrix 128 8.3 0.001 0.001 30.255 30.803 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.017 30.254 30.802 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.935 30.301 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.262 27.768 multiply_cannon_multrec 20056 15.6 13.072 15.810 22.549 25.339 multiply_cannon_metrocomm3 20056 15.6 0.061 0.065 15.581 25.013 qs_ot_get_derivative 117 11.6 0.001 0.002 22.995 23.444 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.846 20.979 apply_single 128 13.6 0.001 0.001 19.845 20.978 qs_ot_get_p 128 10.4 0.001 0.001 19.489 20.064 ot_diis_step 117 11.6 0.018 0.019 19.694 19.694 make_m2s 5014 13.6 0.078 0.083 14.882 16.063 make_images 5014 14.6 1.140 1.236 14.646 15.820 qs_ot_p2m_diag 83 11.4 0.265 0.272 15.217 15.228 multiply_cannon_sync_h2d 20056 15.6 13.563 15.198 13.563 15.198 cp_fm_cholesky_invert 11 10.9 14.692 14.703 14.692 14.703 cp_dbcsr_syevd 83 12.4 0.005 0.006 14.145 14.146 sum_up_and_integrate 128 10.3 0.002 0.003 13.974 14.000 integrate_v_rspace 128 11.3 0.003 0.003 13.914 13.942 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.590 13.629 calculate_rho_elec 128 8.7 0.128 0.143 13.589 13.628 init_scf_run 11 5.9 0.000 0.001 11.233 11.233 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.233 11.233 cp_fm_diag_elpa 83 13.4 0.000 0.000 10.919 10.942 cp_fm_diag_elpa_base 83 14.4 10.506 10.683 10.916 10.939 make_images_data 5014 15.6 0.063 0.073 8.978 10.471 hybrid_alltoall_any 5200 16.5 0.448 2.040 7.717 9.772 multiply_cannon_metrocomm4 17549 15.6 0.065 0.076 3.507 9.651 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.058 9.390 mp_irecv_dv 50230 16.2 3.378 9.384 3.378 9.384 dbcsr_mm_accdrv_process 41502 16.2 5.745 6.192 8.945 9.174 pw_transfer 1547 11.6 0.084 0.100 8.881 8.984 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 8.657 8.767 fft_wrap_pw1pw2_140 523 13.2 1.058 1.090 7.663 7.786 grid_integrate_task_list 128 12.3 7.306 7.658 7.306 7.658 wfi_extrapolate 11 7.9 0.001 0.001 7.483 7.483 density_rs2pw 128 9.7 0.005 0.006 6.983 7.434 cp_fm_cholesky_decompose 22 10.9 7.412 7.431 7.412 7.431 cp_fm_upper_to_full 105 14.8 5.755 7.237 5.755 7.237 dbcsr_complete_redistribute 395 12.7 1.152 1.180 4.616 6.371 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.232 6.235 fft3d_ps 1291 14.7 2.822 3.068 5.975 6.069 calculate_dm_sparse 128 9.5 0.001 0.001 5.856 5.955 grid_collocate_task_list 128 9.7 5.149 5.495 5.149 5.495 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.298 5.439 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.648 5.263 mp_alltoall_d11v 2415 14.1 4.359 5.208 4.359 5.208 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.435 5.191 mp_sum_l 7950 12.9 3.223 4.836 3.223 4.836 mp_allgather_i34 2507 14.6 1.469 4.769 1.469 4.769 potential_pw2rs 128 12.3 0.020 0.021 4.400 4.410 transfer_fm_to_dbcsr 11 9.9 0.019 0.023 2.334 4.065 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.975 4.005 mp_alltoall_i22 716 14.1 1.932 3.828 1.932 3.828 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.731 3.732 calculate_first_density_matrix 1 7.0 0.000 0.000 3.562 3.565 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=177.038000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=885.090909, yerr=1.443137 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/845a3ae33c0b6c5fc8bf8a28d823ca3539bb7d75_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 4.320339E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 5927808 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.2 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.165660E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1133160 MPI messages size (bytes): total size 2.008142E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.772161E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 315952 35695099904 131072 < size <= 4194304 709496 778939400192 4194304 < size <= 16777216 69840 660837692480 16777216 < size 30480 532676608000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4003 58189. MP_Allreduce 11085 1083. MP_Sync 86 MP_Alltoall 1700 12496381. MP_SendRecv 5842 75008. MP_ISendRecv 5842 75008. MP_Wait 22272 MP_ISend 14840 244848. MP_IRecv 14840 244848. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.031 197.844 197.846 qs_mol_dyn_low 1 2.0 0.003 0.004 197.473 197.487 qs_forces 11 3.9 0.004 0.004 197.361 197.370 qs_energies 11 4.9 0.002 0.002 190.118 190.128 scf_env_do_scf 11 5.9 0.001 0.001 173.051 173.064 velocity_verlet 10 3.0 0.002 0.002 129.632 129.635 scf_env_do_scf_inner_loop 116 6.6 0.003 0.009 123.939 123.941 qs_scf_new_mos 116 7.6 0.001 0.001 86.648 86.958 qs_scf_loop_do_ot 116 8.6 0.001 0.001 86.647 86.957 dbcsr_multiply_generic 2485 12.5 0.190 0.196 81.597 82.483 ot_scf_mini 116 9.6 0.003 0.003 81.977 82.306 multiply_cannon 2485 13.5 0.549 0.587 56.200 59.391 multiply_cannon_loop 2485 14.5 1.825 1.918 51.619 53.708 init_scf_loop 11 6.9 0.001 0.001 48.983 48.987 prepare_preconditioner 11 7.9 0.000 0.000 44.825 44.853 make_preconditioner 11 8.9 0.000 0.000 44.824 44.853 ot_mini 116 10.6 0.001 0.001 43.779 44.118 make_full_inverse_cholesky 11 9.9 0.011 0.021 38.305 43.376 multiply_cannon_multrec 29820 15.5 13.600 18.661 26.424 31.103 rebuild_ks_matrix 127 8.3 0.001 0.001 29.815 30.148 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.019 29.814 30.148 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.856 27.158 mp_waitall_1 146592 16.7 17.718 26.928 17.718 26.928 qs_ot_get_derivative 116 11.6 0.001 0.002 24.217 24.548 qs_ot_get_p 127 10.4 0.001 0.001 24.071 24.419 make_m2s 4970 13.5 0.091 0.097 20.942 21.977 make_images 4970 14.5 1.937 2.429 20.633 21.673 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 18.928 19.606 apply_single 127 13.6 0.001 0.001 18.928 19.606 qs_ot_p2m_diag 82 11.4 0.339 0.386 19.465 19.518 ot_diis_step 116 11.6 0.017 0.018 19.438 19.439 cp_dbcsr_syevd 82 12.4 0.005 0.005 18.010 18.014 cp_fm_cholesky_invert 11 10.9 17.733 17.742 17.733 17.742 cp_fm_upper_to_full 104 14.8 11.524 17.032 11.524 17.032 sum_up_and_integrate 127 10.3 0.002 0.003 14.523 14.550 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.477 14.525 calculate_rho_elec 127 8.7 0.168 0.183 14.476 14.524 multiply_cannon_metrocomm3 29820 15.5 0.048 0.050 6.054 14.497 integrate_v_rspace 127 11.3 0.003 0.004 14.464 14.493 cp_fm_diag_elpa 82 13.4 0.000 0.000 14.448 14.470 cp_fm_diag_elpa_base 82 14.4 13.434 13.758 14.441 14.462 make_images_data 4970 15.5 0.064 0.070 11.282 13.295 dbcsr_complete_redistribute 393 12.7 1.484 1.602 9.401 13.288 dbcsr_mm_accdrv_process 61748 16.2 8.426 9.126 12.394 12.876 copy_fm_to_dbcsr 208 11.6 0.001 0.002 7.940 11.827 hybrid_alltoall_any 5155 16.4 0.521 2.054 9.929 11.681 multiply_cannon_sync_h2d 29820 15.5 10.347 11.493 10.347 11.493 init_scf_run 11 5.9 0.000 0.001 11.130 11.131 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.130 11.131 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 10.334 10.576 transfer_fm_to_dbcsr 11 9.9 0.001 0.009 6.497 10.274 mp_alltoall_i22 712 14.1 5.725 9.667 5.725 9.667 pw_transfer 1535 11.6 0.085 0.100 9.348 9.461 fft_wrap_pw1pw2 1281 12.7 0.010 0.012 9.124 9.244 cp_fm_cholesky_decompose 22 10.9 8.366 8.440 8.366 8.440 fft_wrap_pw1pw2_140 519 13.2 1.239 1.279 7.972 8.102 wfi_extrapolate 11 7.9 0.001 0.001 8.017 8.017 grid_integrate_task_list 127 12.3 7.457 7.934 7.457 7.934 density_rs2pw 127 9.7 0.005 0.005 7.497 7.864 multiply_cannon_metrocomm4 24850 15.5 0.078 0.087 2.773 6.723 mp_irecv_dv 75445 16.2 2.625 6.453 2.625 6.453 calculate_dm_sparse 127 9.5 0.001 0.001 6.364 6.447 fft3d_ps 1281 14.7 2.968 3.025 6.203 6.295 mp_alltoall_d11v 2401 14.1 5.374 6.068 5.374 6.068 grid_collocate_task_list 127 9.7 5.246 5.806 5.246 5.806 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.598 5.668 potential_pw2rs 127 12.3 0.023 0.023 4.863 4.905 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.661 4.754 mp_allgather_i34 2485 14.5 2.691 4.718 2.691 4.718 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.522 4.583 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.426 4.565 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 4.413 4.414 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=197.846000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1104.454545, yerr=16.805646 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/845a3ae33c0b6c5fc8bf8a28d823ca3539bb7d75_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019187724288 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019187724288 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 92.796577E+12 0.0% 0.0% 100.0% flops max/rank 5.820059E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500544 0.0% 0.0% 100.0% number of processed stacks 1944496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3448.5 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.552663E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 238560 MPI messages size (bytes): total size 1.321104E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.537828E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 112800 59139686400 4194304 < size <= 16777216 104112 545846722560 16777216 < size 20064 716108613552 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8852 52. MP_Alltoall 9584 804353. MP_ISend 39716 2104723. MP_IRecv 39716 2103824. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3967 58706. MP_Allreduce 10995 1169. MP_Sync 86 MP_Alltoall 1700 18828162. MP_SendRecv 3810 122880. MP_ISendRecv 3810 122880. MP_Wait 16000 MP_ISend 10600 423612. MP_IRecv 10600 423612. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.031 184.942 184.944 qs_mol_dyn_low 1 2.0 0.003 0.004 184.531 184.546 qs_forces 11 3.9 0.004 0.004 184.423 184.425 qs_energies 11 4.9 0.002 0.002 176.819 176.832 scf_env_do_scf 11 5.9 0.001 0.001 158.998 159.005 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 119.769 119.771 velocity_verlet 10 3.0 0.002 0.002 119.567 119.571 qs_scf_new_mos 116 7.6 0.001 0.001 81.198 81.270 qs_scf_loop_do_ot 116 8.6 0.001 0.001 81.198 81.270 ot_scf_mini 116 9.6 0.003 0.004 76.714 76.786 dbcsr_multiply_generic 2485 12.5 0.184 0.189 74.178 74.871 multiply_cannon 2485 13.5 0.556 0.578 54.154 57.586 multiply_cannon_loop 2485 14.5 0.801 0.834 50.637 52.279 ot_mini 116 10.6 0.001 0.001 39.331 39.413 init_scf_loop 11 6.9 0.001 0.001 39.067 39.070 prepare_preconditioner 11 7.9 0.000 0.000 35.085 35.107 make_preconditioner 11 8.9 0.000 0.000 35.085 35.107 mp_waitall_1 124680 16.7 26.286 33.422 26.286 33.422 make_full_inverse_cholesky 11 9.9 0.017 0.026 32.980 33.296 rebuild_ks_matrix 127 8.3 0.001 0.001 30.145 30.274 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.017 30.144 30.273 qs_ks_update_qs_env 127 7.6 0.001 0.001 27.451 27.564 qs_ot_get_p 127 10.4 0.001 0.001 24.382 24.497 qs_ot_get_derivative 116 11.6 0.002 0.002 22.231 22.303 multiply_cannon_multrec 9940 15.5 10.263 14.372 17.798 21.122 cp_fm_cholesky_invert 11 10.9 20.517 20.526 20.517 20.526 multiply_cannon_metrocomm3 9940 15.5 0.024 0.025 12.614 20.449 qs_ot_p2m_diag 82 11.4 0.489 0.495 20.176 20.193 cp_dbcsr_syevd 82 12.4 0.005 0.005 18.833 18.835 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 17.156 17.399 apply_single 127 13.6 0.001 0.001 17.156 17.399 make_m2s 4970 13.5 0.064 0.070 15.689 17.012 ot_diis_step 116 11.6 0.020 0.020 16.987 16.988 make_images 4970 14.5 2.127 2.559 15.383 16.709 cp_fm_diag_elpa 82 13.4 0.000 0.000 15.342 15.353 cp_fm_diag_elpa_base 82 14.4 15.066 15.174 15.337 15.348 qs_rho_update_rho_low 127 7.7 0.001 0.001 15.051 15.106 calculate_rho_elec 127 8.7 0.251 0.262 15.051 15.106 sum_up_and_integrate 127 10.3 0.002 0.002 15.017 15.069 integrate_v_rspace 127 11.3 0.004 0.004 14.957 15.010 multiply_cannon_sync_h2d 9940 15.5 10.700 11.216 10.700 11.216 init_scf_run 11 5.9 0.000 0.001 10.870 10.871 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.870 10.870 pw_transfer 1535 11.6 0.084 0.093 10.432 10.481 make_images_data 4970 15.5 0.054 0.064 9.024 10.430 fft_wrap_pw1pw2 1281 12.7 0.010 0.010 10.211 10.266 hybrid_alltoall_any 5155 16.4 0.836 3.797 8.678 10.200 cp_fm_cholesky_decompose 22 10.9 8.900 9.055 8.900 9.055 qs_ot_get_derivative_diag 76 12.4 0.002 0.003 8.854 8.918 fft_wrap_pw1pw2_140 519 13.2 1.663 1.708 8.812 8.869 multiply_cannon_metrocomm1 9940 15.5 0.030 0.030 4.975 8.237 grid_integrate_task_list 127 12.3 7.644 8.098 7.644 8.098 density_rs2pw 127 9.7 0.005 0.005 7.691 8.022 dbcsr_mm_accdrv_process 20590 16.1 3.165 4.130 7.187 7.896 wfi_extrapolate 11 7.9 0.001 0.001 7.830 7.830 fft3d_ps 1281 14.7 3.119 3.224 6.699 6.724 calculate_dm_sparse 127 9.5 0.001 0.001 6.139 6.222 dbcsr_complete_redistribute 393 12.7 2.079 2.129 5.443 5.894 mp_alltoall_d11v 2401 14.1 5.127 5.816 5.127 5.816 grid_collocate_task_list 127 9.7 5.529 5.774 5.529 5.774 mp_allgather_i34 2485 14.5 1.736 5.395 1.736 5.395 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.278 5.280 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.253 5.264 potential_pw2rs 127 12.3 0.026 0.026 5.084 5.110 multiply_cannon_metrocomm4 7455 15.5 0.025 0.028 1.793 4.249 mp_irecv_dv 28618 15.9 1.754 4.170 1.754 4.170 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.084 4.135 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.950 4.023 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.581 3.910 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.828 3.892 copy_fm_to_dbcsr 208 11.6 0.001 0.002 3.467 3.816 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.637 3.800 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=184.944000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1475.272727, yerr=18.111227 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/845a3ae33c0b6c5fc8bf8a28d823ca3539bb7d75_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 3.167281E+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 3997 59207. MP_Allreduce 11070 1504. MP_Sync 86 MP_Alltoall 1700 36954382. MP_SendRecv 1778 218624. MP_ISendRecv 1778 218624. MP_Wait 9728 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 0.043 0.078 305.520 305.530 qs_mol_dyn_low 1 2.0 0.003 0.004 304.820 304.833 qs_forces 11 3.9 0.004 0.004 304.723 304.728 qs_energies 11 4.9 0.001 0.002 295.662 295.669 scf_env_do_scf 11 5.9 0.001 0.001 271.290 271.301 velocity_verlet 10 3.0 0.002 0.002 218.543 218.550 scf_env_do_scf_inner_loop 116 6.6 0.004 0.008 142.729 142.732 init_scf_loop 11 6.9 0.001 0.001 128.267 128.271 prepare_preconditioner 11 7.9 0.000 0.000 123.203 123.229 make_preconditioner 11 8.9 0.000 0.000 123.203 123.229 make_full_inverse_cholesky 11 9.9 0.038 0.039 99.226 120.236 qs_scf_new_mos 116 7.6 0.001 0.001 96.237 96.299 qs_scf_loop_do_ot 116 8.6 0.001 0.001 96.236 96.298 ot_scf_mini 116 9.6 0.004 0.004 91.287 91.325 dbcsr_multiply_generic 2485 12.5 0.214 0.225 84.519 85.126 cp_fm_upper_to_full 104 14.8 53.132 76.006 53.132 76.006 multiply_cannon 2485 13.5 0.686 0.743 60.319 61.468 multiply_cannon_loop 2485 14.5 1.038 1.066 56.170 57.667 ot_mini 116 10.6 0.001 0.001 45.613 45.659 dbcsr_complete_redistribute 393 12.7 3.966 4.040 30.302 43.255 copy_fm_to_dbcsr 208 11.6 0.001 0.002 26.579 39.526 transfer_fm_to_dbcsr 11 9.9 0.030 0.030 23.934 36.704 rebuild_ks_matrix 127 8.3 0.001 0.001 35.785 35.844 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.017 35.784 35.843 cp_fm_cholesky_invert 11 10.9 35.646 35.653 35.646 35.653 mp_alltoall_i22 712 14.1 21.701 34.897 21.701 34.897 mp_waitall_1 102768 16.8 29.460 34.638 29.460 34.638 qs_ks_update_qs_env 127 7.6 0.001 0.001 33.094 33.158 qs_ot_get_p 127 10.4 0.001 0.001 30.551 30.587 qs_ot_p2m_diag 82 11.4 0.868 0.873 26.027 26.057 qs_ot_get_derivative 116 11.6 0.002 0.002 25.879 25.918 cp_dbcsr_syevd 82 12.4 0.006 0.006 24.148 24.151 make_m2s 4970 13.5 0.072 0.076 19.391 20.936 multiply_cannon_metrocomm3 9940 15.5 0.024 0.025 19.423 20.932 cp_fm_diag_elpa 82 13.4 0.000 0.000 20.591 20.592 cp_fm_diag_elpa_base 82 14.4 16.145 17.797 20.586 20.586 make_images 4970 14.5 3.032 3.220 18.913 20.460 ot_diis_step 116 11.6 0.022 0.022 19.694 19.695 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 18.981 19.124 apply_single 127 13.6 0.001 0.001 18.981 19.124 qs_rho_update_rho_low 127 7.7 0.001 0.001 18.342 18.354 calculate_rho_elec 127 8.7 0.474 0.475 18.341 18.354 multiply_cannon_multrec 9940 15.5 10.170 11.976 18.050 18.171 sum_up_and_integrate 127 10.3 0.002 0.002 17.318 17.405 integrate_v_rspace 127 11.3 0.004 0.004 17.256 17.342 multiply_cannon_sync_h2d 9940 15.5 14.195 14.244 14.195 14.244 pw_transfer 1535 11.6 0.094 0.095 14.208 14.214 fft_wrap_pw1pw2 1281 12.7 0.012 0.012 13.972 13.976 init_scf_run 11 5.9 0.000 0.001 13.896 13.897 scf_env_initial_rho_setup 11 6.9 0.001 0.001 13.896 13.897 hybrid_alltoall_any 5155 16.4 1.302 3.048 10.997 13.543 make_images_data 4970 15.5 0.062 0.068 10.962 13.492 fft_wrap_pw1pw2_140 519 13.2 3.204 3.253 12.288 12.295 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 10.269 10.300 cp_fm_cholesky_decompose 22 10.9 10.236 10.254 10.236 10.254 wfi_extrapolate 11 7.9 0.001 0.001 9.507 9.507 dbcsr_mm_accdrv_process 20590 16.0 4.237 5.977 7.631 9.469 density_rs2pw 127 9.7 0.005 0.005 9.224 9.267 grid_integrate_task_list 127 12.3 8.506 8.694 8.506 8.694 fft3d_ps 1281 14.7 3.944 3.965 8.470 8.543 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 8.182 8.183 mp_alltoall_d11v 2401 14.1 7.849 7.969 7.849 7.969 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 7.641 7.674 calculate_dm_sparse 127 9.5 0.001 0.001 6.842 6.892 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.506 6.686 grid_collocate_task_list 127 9.7 6.373 6.451 6.373 6.451 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=305.530000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2862.181818, yerr=150.917086 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/845a3ae33c0b6c5fc8bf8a28d823ca3539bb7d75_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 1.260941E+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 255646. MP_Allreduce 3139 6114. MP_Sync 4 MP_Alltoall 54 MP_SendRecv 285 19200. MP_ISendRecv 285 19200. MP_Wait 1017 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.027 0.114 85.488 85.490 qs_energies 1 2.0 0.002 0.017 84.443 84.542 ls_scf 1 3.0 0.001 0.008 83.523 83.598 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.389 72.550 multiply_cannon 111 7.7 0.018 0.021 55.820 57.163 multiply_cannon_loop 111 8.7 0.229 0.247 52.413 54.016 ls_scf_main 1 4.0 0.000 0.005 52.276 52.277 density_matrix_trs4 2 5.0 0.002 0.004 46.705 46.803 ls_scf_init_scf 1 4.0 0.000 0.003 28.158 28.159 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.016 27.070 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 24.906 24.927 mp_waitall_1 11031 10.9 21.956 24.306 21.956 24.306 multiply_cannon_multrec 2664 9.7 8.182 8.958 15.572 17.340 multiply_cannon_sync_h2d 2664 9.7 13.792 15.880 13.792 15.880 make_m2s 222 7.7 0.009 0.011 12.921 13.467 make_images 222 8.7 0.098 0.108 12.899 13.447 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.460 13.277 make_images_data 222 9.7 0.004 0.005 7.475 8.038 dbcsr_mm_accdrv_process 4760 10.4 0.586 0.700 7.004 7.997 hybrid_alltoall_any 227 10.6 0.215 1.832 6.397 7.542 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.341 7.390 dbcsr_mm_accdrv_process_sort 4760 11.4 6.218 7.135 6.218 7.135 calculate_norms 4752 9.8 5.552 6.266 5.552 6.266 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.030 5.151 mp_sum_l 887 5.1 3.156 4.651 3.156 4.651 make_images_sizes 222 9.7 0.000 0.000 0.757 3.458 mp_alltoall_i44 222 10.7 0.757 3.457 0.757 3.457 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.285 3.379 arnoldi_extremal 4 6.8 0.000 0.000 3.304 3.336 arnoldi_normal_ev 4 7.8 0.001 0.003 3.304 3.336 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.034 3.242 mp_irecv_dv 6231 10.9 2.017 3.211 2.017 3.211 build_subspace 16 8.4 0.009 0.012 3.201 3.202 ls_scf_post 1 4.0 0.011 0.119 3.088 3.158 ls_scf_store_result 1 5.0 0.000 0.000 2.855 2.892 dbcsr_special_finalize 555 9.7 0.005 0.006 2.335 2.819 dbcsr_merge_single_wm 555 10.7 0.456 0.579 2.326 2.810 make_images_pack 222 9.7 2.208 2.633 2.210 2.634 dbcsr_matrix_vector_mult 304 9.0 0.006 0.014 2.346 2.587 dbcsr_sort_data 658 11.4 2.130 2.540 2.130 2.540 dbcsr_matrix_vector_mult_local 304 10.0 2.062 2.456 2.064 2.458 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.319 2.441 buffer_matrices_ensure_size 222 8.7 1.740 2.046 1.740 2.046 mp_allgather_i34 111 8.7 0.734 1.827 0.734 1.827 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.779 1.781 rebuild_ks_matrix 3 7.3 0.000 0.000 1.770 1.771 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.005 1.770 1.771 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.490000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1132.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/845a3ae33c0b6c5fc8bf8a28d823ca3539bb7d75_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 2.173817E+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 266673. MP_Allreduce 3138 10075. MP_Sync 4 MP_Alltoall 47 15335933. MP_SendRecv 141 57600. MP_ISendRecv 141 57600. MP_Wait 687 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.033 0.085 89.844 89.845 qs_energies 1 2.0 0.000 0.001 89.286 89.292 ls_scf 1 3.0 0.000 0.001 87.941 87.946 dbcsr_multiply_generic 111 6.7 0.015 0.016 73.715 74.053 multiply_cannon 111 7.7 0.027 0.044 52.776 57.123 ls_scf_main 1 4.0 0.000 0.004 54.315 54.320 multiply_cannon_loop 111 8.7 0.135 0.147 50.055 53.374 density_matrix_trs4 2 5.0 0.002 0.004 48.657 48.837 ls_scf_init_scf 1 4.0 0.000 0.002 30.010 30.012 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.770 28.882 mp_waitall_1 9105 10.9 20.575 28.609 20.575 28.609 multiply_cannon_multrec 1332 9.7 13.289 16.615 22.587 27.077 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.398 26.408 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.703 20.540 make_m2s 222 7.7 0.006 0.008 14.751 15.213 make_images 222 8.7 1.373 1.681 14.720 15.183 dbcsr_mm_accdrv_process 4041 10.4 0.323 0.534 8.896 10.443 dbcsr_mm_accdrv_process_sort 4041 11.4 8.408 9.918 8.408 9.918 make_images_data 222 9.7 0.004 0.005 8.336 9.041 hybrid_alltoall_any 227 10.6 0.544 2.563 7.725 8.918 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.233 7.683 mp_irecv_dv 3311 11.0 3.212 7.652 3.212 7.652 mp_sum_l 887 5.1 4.762 7.519 4.762 7.519 calculate_norms 2376 9.8 6.043 6.772 6.043 6.772 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.604 6.165 multiply_cannon_sync_h2d 1332 9.7 4.806 5.911 4.806 5.911 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.910 5.119 arnoldi_extremal 4 6.8 0.000 0.000 4.992 5.005 arnoldi_normal_ev 4 7.8 0.001 0.005 4.992 5.005 build_subspace 16 8.4 0.014 0.021 4.741 4.744 ls_scf_post 1 4.0 0.000 0.001 3.616 3.622 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.264 3.541 ls_scf_store_result 1 5.0 0.000 0.000 3.365 3.441 dbcsr_matrix_vector_mult_local 304 10.0 2.737 3.211 2.739 3.212 mp_allgather_i34 111 8.7 0.709 2.708 0.709 2.708 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.603 2.682 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.176 2.619 dbcsr_data_new 4174 10.1 2.116 2.407 2.116 2.407 make_images_pack 222 9.7 1.827 2.134 1.830 2.136 dbcsr_sort_data 436 11.2 1.839 2.085 1.839 2.085 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.889 1.891 rebuild_ks_matrix 3 7.3 0.000 0.000 1.875 1.877 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.875 1.877 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=89.845000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1775.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/845a3ae33c0b6c5fc8bf8a28d823ca3539bb7d75_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 2.917323E+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 265448. MP_Allreduce 3138 10896. MP_Sync 4 MP_Alltoall 47 23526250. MP_SendRecv 93 57600. MP_ISendRecv 93 57600. MP_Wait 639 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.025 0.043 91.117 91.117 qs_energies 1 2.0 0.000 0.000 90.562 90.565 ls_scf 1 3.0 0.000 0.000 89.145 89.149 dbcsr_multiply_generic 111 6.7 0.015 0.016 73.789 74.098 ls_scf_main 1 4.0 0.000 0.000 55.501 55.504 multiply_cannon 111 7.7 0.035 0.110 51.448 55.467 multiply_cannon_loop 111 8.7 0.116 0.131 48.711 52.707 density_matrix_trs4 2 5.0 0.002 0.003 49.712 49.865 mp_waitall_1 7281 11.0 22.272 32.547 22.272 32.547 ls_scf_init_scf 1 4.0 0.000 0.000 30.048 30.051 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.862 28.940 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.543 26.554 multiply_cannon_multrec 888 9.7 12.718 15.448 21.288 24.610 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.580 22.766 make_m2s 222 7.7 0.006 0.007 15.864 16.564 make_images 222 8.7 1.589 1.867 15.826 16.526 make_images_data 222 9.7 0.004 0.005 8.913 9.948 hybrid_alltoall_any 227 10.6 0.642 2.941 8.402 9.529 dbcsr_mm_accdrv_process 3754 10.4 0.302 0.503 8.109 9.347 mp_sum_l 887 5.1 4.791 8.893 4.791 8.893 dbcsr_mm_accdrv_process_sort 3754 11.4 7.671 8.844 7.671 8.844 multiply_cannon_sync_h2d 888 9.7 6.137 7.445 6.137 7.445 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.472 7.401 mp_irecv_dv 2335 11.1 2.456 7.326 2.456 7.326 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.723 7.219 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.418 6.624 arnoldi_extremal 4 6.8 0.000 0.000 5.232 5.251 arnoldi_normal_ev 4 7.8 0.001 0.005 5.232 5.251 build_subspace 16 8.4 0.014 0.020 4.927 4.933 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.656 4.844 calculate_norms 1584 9.8 4.348 4.821 4.348 4.821 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.493 3.817 dbcsr_matrix_vector_mult_local 304 10.0 3.016 3.601 3.018 3.603 ls_scf_post 1 4.0 0.000 0.000 3.595 3.601 mp_allgather_i34 111 8.7 0.853 3.466 0.853 3.466 ls_scf_store_result 1 5.0 0.000 0.000 3.365 3.421 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.828 2.909 dbcsr_data_new 4116 9.9 2.109 2.480 2.109 2.480 make_images_sizes 222 9.7 0.000 0.000 1.146 2.248 mp_alltoall_i44 222 10.7 1.145 2.247 1.145 2.247 dbcsr_sort_data 325 11.1 1.827 2.172 1.827 2.172 make_images_pack 222 9.7 1.628 1.885 1.631 1.888 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.879 1.881 dbcsr_finalize 304 7.8 0.026 0.032 1.599 1.877 rebuild_ks_matrix 3 7.3 0.000 0.000 1.861 1.863 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.861 1.863 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=91.117000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2285.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/845a3ae33c0b6c5fc8bf8a28d823ca3539bb7d75_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 3.325215E+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 266673. MP_Allreduce 3138 13030. MP_Sync 4 MP_Alltoall 47 30278988. MP_SendRecv 69 86400. MP_ISendRecv 69 86400. MP_Wait 531 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.030 0.048 96.985 96.986 qs_energies 1 2.0 0.000 0.000 96.315 96.322 ls_scf 1 3.0 0.000 0.001 94.650 94.656 dbcsr_multiply_generic 111 6.7 0.016 0.018 78.025 78.267 ls_scf_main 1 4.0 0.000 0.003 58.838 58.839 multiply_cannon 111 7.7 0.046 0.091 51.495 56.144 density_matrix_trs4 2 5.0 0.002 0.003 52.646 52.786 multiply_cannon_loop 111 8.7 0.154 0.166 46.396 48.877 ls_scf_init_scf 1 4.0 0.000 0.001 32.568 32.571 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.357 31.416 mp_waitall_1 6369 11.0 22.488 29.046 22.488 29.046 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.907 28.918 multiply_cannon_multrec 1332 9.7 14.141 17.325 22.042 25.135 make_m2s 222 7.7 0.007 0.008 21.068 22.558 make_images 222 8.7 3.152 3.615 21.018 22.510 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.080 16.890 make_images_data 222 9.7 0.004 0.004 11.774 13.388 hybrid_alltoall_any 227 10.6 0.796 3.784 11.134 12.933 dbcsr_mm_accdrv_process 3641 10.4 0.287 0.481 7.549 9.078 dbcsr_mm_accdrv_process_sort 3641 11.4 7.108 8.582 7.108 8.582 mp_sum_l 887 5.1 3.871 7.592 3.871 7.592 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.062 6.174 mp_irecv_dv 3229 10.9 2.037 6.092 2.037 6.092 multiply_cannon_sync_h2d 1332 9.7 5.484 6.036 5.484 6.036 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.899 5.926 arnoldi_extremal 4 6.8 0.000 0.000 5.527 5.543 arnoldi_normal_ev 4 7.8 0.001 0.004 5.527 5.543 build_subspace 16 8.4 0.014 0.021 5.187 5.194 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.458 4.780 mp_allgather_i34 111 8.7 2.217 4.711 2.217 4.711 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.560 4.696 calculate_norms 2376 9.8 4.164 4.520 4.164 4.520 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.691 4.044 dbcsr_matrix_vector_mult_local 304 10.0 3.187 3.676 3.189 3.677 dbcsr_sort_data 658 11.4 3.085 3.475 3.085 3.475 ls_scf_post 1 4.0 0.000 0.001 3.245 3.252 dbcsr_special_finalize 555 9.7 0.006 0.007 2.826 3.192 dbcsr_merge_single_wm 555 10.7 0.542 0.665 2.818 3.184 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.055 3.128 ls_scf_store_result 1 5.0 0.000 0.000 2.981 3.047 dbcsr_data_release 10477 10.7 1.581 2.433 1.581 2.433 dbcsr_finalize 304 7.8 0.049 0.061 1.813 2.038 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.947 1.950 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=96.986000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2726.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/845a3ae33c0b6c5fc8bf8a28d823ca3539bb7d75_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 4.721570E+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 265536. MP_Allreduce 3129 15263. MP_Sync 4 MP_Alltoall 47 46208988. MP_SendRecv 45 115200. MP_ISendRecv 45 115200. MP_Wait 528 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.035 0.056 91.992 91.992 qs_energies 1 2.0 0.000 0.000 91.207 91.212 ls_scf 1 3.0 0.000 0.000 89.246 89.253 dbcsr_multiply_generic 111 6.7 0.017 0.018 70.823 71.006 ls_scf_main 1 4.0 0.000 0.000 56.620 56.620 multiply_cannon 111 7.7 0.102 0.180 52.872 56.212 multiply_cannon_loop 111 8.7 0.088 0.097 50.290 51.609 density_matrix_trs4 2 5.0 0.002 0.003 49.692 49.779 ls_scf_init_scf 1 4.0 0.000 0.000 29.226 29.228 mp_waitall_1 5436 11.0 24.594 28.423 24.594 28.423 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.972 27.998 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.867 25.883 multiply_cannon_multrec 444 9.7 13.664 16.711 20.753 22.229 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.549 15.633 make_m2s 222 7.7 0.005 0.005 13.443 14.369 make_images 222 8.7 2.035 2.473 13.376 14.300 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.668 13.507 hybrid_alltoall_any 227 10.6 0.801 3.804 8.096 9.640 make_images_data 222 9.7 0.003 0.004 8.195 9.621 dbcsr_mm_accdrv_process 3003 10.4 0.353 0.408 6.783 7.960 multiply_cannon_sync_h2d 444 9.7 6.790 7.862 6.790 7.862 dbcsr_mm_accdrv_process_sort 3003 11.4 6.419 7.558 6.419 7.558 arnoldi_extremal 4 6.8 0.000 0.000 5.855 5.871 arnoldi_normal_ev 4 7.8 0.002 0.005 5.854 5.871 build_subspace 16 8.4 0.015 0.020 5.459 5.467 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.381 4.544 dbcsr_matrix_vector_mult 304 9.0 0.011 0.022 4.168 4.381 dbcsr_matrix_vector_mult_local 304 10.0 3.706 4.183 3.708 4.185 mp_sum_l 887 5.1 2.504 4.150 2.504 4.150 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.572 3.863 mp_irecv_dv 1241 11.2 1.559 3.831 1.559 3.831 mp_allgather_i34 111 8.7 1.157 3.725 1.157 3.725 calculate_norms 792 9.8 3.615 3.674 3.615 3.674 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.528 3.584 ls_scf_post 1 4.0 0.000 0.000 3.401 3.405 ls_scf_store_result 1 5.0 0.000 0.000 3.184 3.225 make_images_sizes 222 9.7 0.000 0.000 0.866 3.142 mp_alltoall_i44 222 10.7 0.866 3.141 0.866 3.141 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.703 3.005 dbcsr_finalize 304 7.8 0.062 0.077 2.193 2.258 dbcsr_data_new 4608 9.7 1.771 2.191 1.771 2.191 dbcsr_merge_all 275 8.9 0.480 0.523 2.053 2.123 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.037 2.038 rebuild_ks_matrix 3 7.3 0.000 0.000 2.005 2.006 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.005 2.006 qs_energies_init_hamiltonians 1 3.0 0.001 0.002 1.945 1.945 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=91.992000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3713.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/845a3ae33c0b6c5fc8bf8a28d823ca3539bb7d75_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 8.853684E+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 284089. MP_Allreduce 3123 21388. MP_Sync 4 MP_Alltoall 47 88727262. MP_SendRecv 42 732600. MP_ISendRecv 42 732600. MP_Wait 267 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.073 0.092 105.462 105.463 qs_energies 1 2.0 0.000 0.000 104.103 104.111 ls_scf 1 3.0 0.000 0.000 101.114 101.121 dbcsr_multiply_generic 111 6.7 0.024 0.027 74.688 74.824 ls_scf_main 1 4.0 0.000 0.000 63.716 63.717 density_matrix_trs4 2 5.0 0.002 0.003 54.674 54.735 multiply_cannon 111 7.7 0.148 0.206 48.520 50.857 multiply_cannon_loop 111 8.7 0.097 0.099 45.431 46.050 ls_scf_init_scf 1 4.0 0.000 0.000 33.611 33.612 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.053 32.077 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.327 29.334 mp_waitall_1 4527 11.1 21.846 25.642 21.846 25.642 make_m2s 222 7.7 0.005 0.005 22.563 23.757 make_images 222 8.7 3.579 3.874 22.455 23.648 multiply_cannon_multrec 444 9.7 17.779 18.535 22.414 23.183 hybrid_alltoall_any 227 10.6 1.655 3.634 12.779 15.586 make_images_data 222 9.7 0.003 0.004 12.955 15.171 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.308 10.564 multiply_cannon_sync_h2d 444 9.7 8.784 8.819 8.784 8.819 arnoldi_extremal 4 6.8 0.000 0.000 7.449 7.460 arnoldi_normal_ev 4 7.8 0.003 0.010 7.449 7.460 build_subspace 16 8.4 0.026 0.036 6.901 6.912 dbcsr_matrix_vector_mult 304 9.0 0.017 0.033 5.490 5.653 dbcsr_matrix_vector_mult_local 304 10.0 5.070 5.387 5.072 5.390 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.176 5.272 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.882 5.141 dbcsr_mm_accdrv_process 1814 10.4 0.289 0.353 4.452 4.579 dbcsr_mm_accdrv_process_sort 1814 11.4 4.117 4.253 4.117 4.253 mp_allgather_i34 111 8.7 1.215 3.877 1.215 3.877 ls_scf_post 1 4.0 0.000 0.000 3.787 3.796 make_images_sizes 222 9.7 0.000 0.000 1.474 3.561 mp_alltoall_i44 222 10.7 1.473 3.561 1.473 3.561 ls_scf_store_result 1 5.0 0.000 0.000 3.507 3.546 calculate_norms 792 9.8 3.237 3.273 3.237 3.273 dbcsr_finalize 304 7.8 0.082 0.090 3.074 3.171 qs_energies_init_hamiltonians 1 3.0 0.001 0.002 2.959 2.959 dbcsr_merge_all 275 8.9 0.885 0.918 2.860 2.951 dbcsr_complete_redistribute 5 7.6 1.430 1.477 2.737 2.854 matrix_ls_to_qs 2 6.0 0.000 0.000 2.391 2.519 dbcsr_sort_data 325 11.1 2.440 2.505 2.440 2.505 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.465 2.467 rebuild_ks_matrix 3 7.3 0.000 0.000 2.399 2.401 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.000 2.399 2.401 dbcsr_new_transposed 4 7.5 0.243 0.257 2.260 2.273 dbcsr_data_new 6591 9.6 1.823 2.272 1.823 2.272 dbcsr_frobenius_norm 74 6.6 2.058 2.131 2.201 2.230 dbcsr_add_d 103 6.2 0.000 0.000 2.126 2.206 dbcsr_add_anytype 103 7.2 0.858 0.890 2.125 2.206 dbcsr_data_release 12724 10.6 1.974 2.175 1.974 2.175 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=105.463000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=7057.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/845a3ae33c0b6c5fc8bf8a28d823ca3539bb7d75_performance_tests/27/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 7009386627072 0.0% 0.0% 100.0% flops 9 x 9 x 32 7335108845568 0.0% 0.0% 100.0% flops 9 x 22 x 32 9866241589248 0.0% 0.0% 100.0% flops 22 x 9 x 32 9884108906496 0.0% 0.0% 100.0% flops 22 x 22 x 32 13354440523776 0.0% 0.0% 100.0% flops 32 x 32 x 9 20607185977344 0.0% 0.0% 100.0% flops 32 x 32 x 22 25186560638976 0.0% 0.0% 100.0% flops 9 x 32 x 32 28458319085568 0.0% 0.0% 100.0% flops 22 x 32 x 32 34782389993472 0.0% 0.0% 100.0% flops 9 x 32 x 9 42881542373376 0.0% 0.0% 100.0% flops 22 x 32 x 9 55680402235392 0.0% 0.0% 100.0% flops 9 x 32 x 22 55680402235392 0.0% 0.0% 100.0% flops 22 x 32 x 22 72328573419520 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 383.054662E+12 0.0% 0.0% 100.0% flops max/rank 733.641090E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 26899403712 0.0% 0.0% 100.0% number of processed stacks 118860288 0.0% 0.0% 100.0% average stack size 0.0 0.0 226.3 marketing flops 780.439111E+12 ------------------------------------------------------------------------------- # multiplications 1445 max memory usage/rank 591.835136E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 102097920 MPI messages size (bytes): total size 37.227590E+12 min size 0.000000E+00 max size 4.551360E+06 average size 364.626312E+03 MPI breakdown and total messages size (bytes): size <= 128 731472 0 128 < size <= 8192 11922720 97670922240 8192 < size <= 32768 24718992 614677610496 32768 < size <= 131072 20000256 1970081366016 131072 < size <= 4194304 42515668 24886801223040 4194304 < size <= 16777216 2208812 9656099886720 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4640 78072. MP_Allreduce 13232 2081. MP_Sync 1064 MP_Alltoall 2588 4803079. MP_SendRecv 168740 11136. MP_ISendRecv 92040 11136. MP_Wait 102830 MP_comm_split 40 MP_ISend 26090 85106. MP_IRecv 37890 59644. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.042 0.114 291.127 291.128 qs_mol_dyn_low 1 2.0 0.004 0.049 289.934 290.010 qs_forces 5 3.8 0.036 0.098 289.829 289.844 qs_energies 5 4.8 0.004 0.052 286.292 286.344 scf_env_do_scf 5 5.8 0.002 0.028 270.404 270.407 scf_env_do_scf_inner_loop 105 6.6 0.005 0.055 236.759 236.761 qs_scf_new_mos 105 7.6 0.000 0.001 187.243 187.405 qs_scf_loop_do_ot 105 8.6 0.001 0.001 187.242 187.404 ot_scf_mini 105 9.6 0.003 0.006 177.030 177.187 dbcsr_multiply_generic 1445 12.2 0.127 0.138 138.723 139.330 multiply_cannon 1445 13.2 0.277 0.287 116.785 118.804 velocity_verlet 4 3.0 0.002 0.012 117.729 117.731 multiply_cannon_loop 1445 14.2 2.821 2.984 114.195 115.983 qs_ot_get_p 112 10.4 0.001 0.026 82.459 82.771 qs_ot_p2m_diag 40 11.0 0.020 0.032 69.794 69.872 cp_dbcsr_syevd 40 12.0 0.002 0.004 66.196 66.200 ot_mini 105 10.6 0.001 0.004 63.705 63.818 cp_fm_syevd 40 13.0 0.000 0.002 59.824 59.985 cp_fm_redistribute_end 40 14.0 27.077 54.012 27.086 54.021 cp_fm_syevd_base 40 14.0 26.918 53.863 26.918 53.863 mp_waitall_1 488190 16.1 39.558 48.195 39.558 48.195 multiply_cannon_multrec 69360 15.2 29.488 34.779 39.509 44.454 qs_ot_get_derivative 55 11.6 0.001 0.001 41.779 41.901 multiply_cannon_metrocomm3 69360 15.2 0.198 0.205 27.149 37.454 init_scf_loop 7 6.6 0.000 0.005 33.591 33.595 multiply_cannon_sync_h2d 69360 15.2 28.687 33.494 28.687 33.494 rebuild_ks_matrix 110 8.4 0.000 0.000 32.749 32.923 qs_ks_build_kohn_sham_matrix 110 9.4 0.015 0.045 32.748 32.922 qs_ks_update_qs_env 112 7.6 0.001 0.001 30.056 30.211 prepare_preconditioner 7 7.6 0.000 0.000 28.194 28.225 make_preconditioner 7 8.6 0.000 0.002 28.194 28.225 qs_rho_update_rho_low 110 7.6 0.000 0.002 23.316 23.772 calculate_rho_elec 110 8.6 0.029 0.032 23.316 23.772 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 23.145 23.479 apply_single 62 13.6 0.000 0.000 23.144 23.479 ot_new_cg_direction 55 11.6 0.001 0.003 21.154 21.155 make_full_inverse_cholesky 7 9.6 0.000 0.002 19.589 19.662 density_rs2pw 110 9.6 0.005 0.007 16.995 17.406 qs_ot_get_orbitals 105 10.6 0.001 0.001 15.517 15.773 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 14.888 14.982 mp_sum_l 4764 12.2 13.124 13.938 13.124 13.938 transfer_rs2pw 445 10.6 0.007 0.008 13.278 13.803 pw_transfer 1645 12.4 0.080 0.103 12.926 13.189 cp_fm_cholesky_invert 7 10.6 13.079 13.098 13.079 13.098 fft_wrap_pw1pw2 1425 13.5 0.012 0.015 12.787 13.053 init_scf_run 5 5.8 0.000 0.001 12.839 12.840 scf_env_initial_rho_setup 5 6.8 0.002 0.003 12.839 12.840 calculate_dm_sparse 110 9.5 0.000 0.001 11.857 12.040 sum_up_and_integrate 60 10.3 0.001 0.003 11.551 11.585 integrate_v_rspace 60 11.3 0.002 0.003 11.534 11.569 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 11.430 11.503 check_diag 80 13.5 8.597 8.897 11.246 11.427 fft_wrap_pw1pw2_240 915 15.0 0.831 0.919 11.038 11.270 qs_vxc_create 110 10.4 0.002 0.004 10.973 11.037 dbcsr_mm_accdrv_process 154766 15.8 6.231 6.412 9.891 10.738 fft3d_pb 915 16.0 2.402 2.685 9.216 9.565 make_m2s 2890 13.2 0.077 0.086 8.394 9.034 make_images 2890 14.2 0.237 0.256 8.288 8.927 acc_transpose_blocks 69360 15.2 0.362 0.380 7.826 8.323 make_full_single_inverse 7 9.6 0.001 0.004 8.238 8.269 multiply_cannon_metrocomm1 69360 15.2 0.097 0.102 5.100 7.932 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.018 7.882 7.897 xc_rho_set_and_dset_create 110 12.4 0.075 0.096 7.612 7.866 potential_pw2rs 60 12.3 0.002 0.003 7.492 7.569 calculate_first_density_matrix 1 7.0 0.000 0.007 7.438 7.450 xc_vxc_pw_create 60 11.3 0.039 0.050 7.376 7.438 mp_sendrecv_dv 168740 12.6 7.225 7.409 7.225 7.409 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 7.331 7.380 transfer_rs2pw_30 110 11.6 1.248 1.302 6.846 7.147 xc_pw_derive 510 13.4 0.006 0.007 6.645 6.714 mp_alltoall_z22v 2340 17.7 6.317 6.709 6.317 6.709 transfer_pw2rs 245 13.2 0.003 0.004 6.234 6.304 cp_fm_cholesky_decompose 14 10.2 5.965 5.973 5.965 5.973 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=291.128000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=562.400000, yerr=2.653300 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 845a3ae33c0b6c5fc8bf8a28d823ca3539bb7d75 Summary: empty Status: OK