=== 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: 44ddd7182aebe9205a531925afe90506688fd6a0 ################# 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/44ddd7182aebe9205a531925afe90506688fd6a0_performance_tests/01 job id: 49932182 --- 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/44ddd7182aebe9205a531925afe90506688fd6a0_performance_tests/02 job id: 49932184 --- 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/44ddd7182aebe9205a531925afe90506688fd6a0_performance_tests/03 job id: 49932185 --- 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/44ddd7182aebe9205a531925afe90506688fd6a0_performance_tests/04 job id: 49932186 --- 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/44ddd7182aebe9205a531925afe90506688fd6a0_performance_tests/05 job id: 49932187 --- 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/44ddd7182aebe9205a531925afe90506688fd6a0_performance_tests/06 job id: 49932189 --- 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/44ddd7182aebe9205a531925afe90506688fd6a0_performance_tests/07 job id: 49932190 --- 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/44ddd7182aebe9205a531925afe90506688fd6a0_performance_tests/08 job id: 49932191 --- 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/44ddd7182aebe9205a531925afe90506688fd6a0_performance_tests/09 job id: 49932192 --- 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/44ddd7182aebe9205a531925afe90506688fd6a0_performance_tests/10 job id: 49932194 --- 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/44ddd7182aebe9205a531925afe90506688fd6a0_performance_tests/11 job id: 49932195 --- 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/44ddd7182aebe9205a531925afe90506688fd6a0_performance_tests/12 job id: 49932197 --- 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/44ddd7182aebe9205a531925afe90506688fd6a0_performance_tests/13 job id: 49932199 --- 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/44ddd7182aebe9205a531925afe90506688fd6a0_performance_tests/14 job id: 49932200 --- 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/44ddd7182aebe9205a531925afe90506688fd6a0_performance_tests/15 job id: 49932202 --- 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/44ddd7182aebe9205a531925afe90506688fd6a0_performance_tests/16 job id: 49932204 --- 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/44ddd7182aebe9205a531925afe90506688fd6a0_performance_tests/17 job id: 49932205 --- 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/44ddd7182aebe9205a531925afe90506688fd6a0_performance_tests/18 job id: 49932206 --- 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/44ddd7182aebe9205a531925afe90506688fd6a0_performance_tests/19 job id: 49932208 --- 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/44ddd7182aebe9205a531925afe90506688fd6a0_performance_tests/20 job id: 49932209 --- 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/44ddd7182aebe9205a531925afe90506688fd6a0_performance_tests/21 job id: 49932210 --- 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/44ddd7182aebe9205a531925afe90506688fd6a0_performance_tests/22 job id: 49932211 --- 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/44ddd7182aebe9205a531925afe90506688fd6a0_performance_tests/23 job id: 49932212 --- 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/44ddd7182aebe9205a531925afe90506688fd6a0_performance_tests/24 job id: 49932215 --- 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/44ddd7182aebe9205a531925afe90506688fd6a0_performance_tests/25 job id: 49932216 --- 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/44ddd7182aebe9205a531925afe90506688fd6a0_performance_tests/26 job id: 49932218 --- 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/44ddd7182aebe9205a531925afe90506688fd6a0_performance_tests/27 job id: 49932219 --- 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/44ddd7182aebe9205a531925afe90506688fd6a0_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 0.022 0.034 133.303 133.304 farming_run 1 2.0 132.651 132.652 133.261 133.264 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.492435E+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 0.006 0.027 114.726 114.728 qs_energies 1 2.0 0.000 0.000 114.539 114.541 mp2_main 1 3.0 0.000 0.000 112.471 112.473 mp2_gpw_main 1 4.0 0.027 0.034 111.447 111.449 mp2_ri_gpw_compute_in 1 5.0 0.171 0.173 92.773 93.110 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 54.694 55.033 mp2_eri_3c_integrate_gpw 272 7.0 0.152 0.165 41.087 46.141 get_2c_integrals 1 6.0 0.008 0.009 37.184 37.906 integrate_v_rspace 273 8.0 0.441 0.456 24.625 29.442 pw_transfer 6555 10.6 0.377 0.397 26.864 27.774 fft_wrap_pw1pw2 5465 11.4 0.045 0.047 25.476 26.384 grid_integrate_task_list 273 9.0 20.529 25.824 20.529 25.824 fft_wrap_pw1pw2_100 2178 12.4 1.154 1.486 23.043 23.933 compute_2c_integrals 1 7.0 0.002 0.002 19.474 19.475 compute_2c_integrals_loop_lm 1 8.0 0.001 0.004 18.667 19.208 mp2_eri_2c_integrate_gpw 1 9.0 2.363 2.425 18.665 19.207 rpa_ri_compute_en 1 5.0 0.019 0.022 18.570 18.713 cp_fm_cholesky_decompose 12 8.2 17.634 18.346 17.634 18.346 cholesky_decomp 1 7.0 0.000 0.000 16.544 17.256 fft3d_s 5443 13.4 16.178 16.573 16.201 16.595 ao_to_mo_and_store_B_mult_1 272 7.0 10.771 15.310 10.771 15.310 calculate_wavefunction 272 8.0 5.446 5.581 12.366 13.065 rpa_num_int 1 6.0 0.001 0.010 10.645 10.654 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.630 10.649 calc_mat_Q 8 8.0 0.000 0.000 9.497 9.585 contract_S_to_Q 8 9.0 0.000 0.000 8.922 9.011 calc_potential_gpw 544 9.5 0.005 0.006 8.279 8.755 parallel_gemm_fm 14 9.1 0.000 0.000 8.506 8.611 parallel_gemm_fm_cosma 14 10.1 8.506 8.611 8.506 8.611 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.125 8.459 potential_pw2rs 545 10.0 0.108 0.110 7.492 8.210 collocate_single_gaussian 272 10.0 0.039 0.042 7.368 7.694 create_integ_mat 1 6.0 0.014 0.029 7.494 7.504 array2fm 1 7.0 0.000 0.000 6.570 6.939 pw_scatter_s 2720 13.7 4.349 4.469 4.349 4.469 pw_gather_s 2722 13.2 3.399 3.668 3.399 3.668 array2fm_buffer_send 1 8.0 2.914 3.091 2.914 3.091 pw_poisson_solve 545 10.5 1.121 1.179 2.214 2.474 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.448759, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2811.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/44ddd7182aebe9205a531925afe90506688fd6a0_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 0.025 0.037 418.898 418.899 farming_run 1 2.0 417.996 418.002 418.858 418.863 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.246757E+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 12975697. 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 0.011 0.032 216.297 216.298 qs_energies 1 2.0 0.002 0.016 215.912 215.983 scf_env_do_scf 1 3.0 0.000 0.000 113.631 113.631 qs_ks_update_qs_env 5 5.0 0.000 0.000 112.744 112.750 rebuild_ks_matrix 4 6.0 0.000 0.000 112.742 112.749 qs_ks_build_kohn_sham_matrix 4 7.0 0.054 0.060 112.742 112.749 hfx_ks_matrix 4 8.0 0.001 0.001 112.395 112.400 integrate_four_center 4 9.0 0.154 0.465 112.394 112.399 integrate_four_center_main 4 10.0 0.132 0.524 101.691 104.065 integrate_four_center_bin 264 11.0 101.559 104.040 101.559 104.040 mp2_main 1 3.0 0.005 0.050 101.992 102.049 mp2_gpw_main 1 4.0 0.059 0.164 101.124 101.163 init_scf_loop 1 4.0 0.000 0.000 95.894 95.894 mp2_ri_gpw_compute_in 1 5.0 0.064 0.064 73.976 75.270 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 53.887 55.177 mp2_eri_3c_integrate_gpw 91 7.0 0.145 0.160 41.525 46.421 integrate_v_rspace 95 8.0 0.398 0.559 27.976 32.710 pw_transfer 2240 10.6 0.146 0.177 29.472 29.899 mp2_ri_gpw_compute_en 1 5.0 0.060 0.109 26.926 29.154 fft_wrap_pw1pw2 1868 11.4 0.018 0.024 28.493 28.921 grid_integrate_task_list 95 9.0 23.394 28.336 23.394 28.336 ao_to_mo_and_store_B_mult_1 91 7.0 10.690 27.996 10.690 27.996 fft_wrap_pw1pw2_100 730 12.4 1.262 1.425 26.250 26.697 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.839 1.888 24.986 24.997 get_2c_integrals 1 6.0 0.000 0.000 19.993 20.029 compute_2c_integrals 1 7.0 0.002 0.003 18.970 18.980 fft3d_s 1823 13.4 18.496 18.954 18.510 18.968 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 18.569 18.852 mp2_eri_2c_integrate_gpw 1 9.0 1.745 1.846 18.568 18.851 scf_env_do_scf_inner_loop 4 4.0 0.000 0.001 17.735 17.735 calculate_wavefunction 91 8.0 2.025 2.061 9.635 9.875 mp2_ri_gpw_compute_en_expansio 172 7.0 0.558 0.586 8.743 9.166 potential_pw2rs 186 10.0 0.033 0.035 8.393 8.862 local_gemm 172 8.0 8.186 8.580 8.186 8.580 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.109 8.444 mp2_ri_gpw_compute_en_comm 22 7.0 0.498 0.515 8.013 8.368 calc_potential_gpw 182 9.5 0.002 0.002 7.890 8.070 collocate_single_gaussian 91 10.0 0.017 0.021 7.731 7.970 mp_sync 37 10.5 2.681 7.467 2.681 7.467 integrate_four_center_load 4 10.0 0.000 0.000 6.764 6.768 hfx_load_balance 1 11.0 0.000 0.000 6.764 6.768 mp2_ri_gpw_compute_en_ener 172 7.0 6.346 6.411 6.346 6.411 mp_sendrecv_dm3 2068 8.0 6.051 6.407 6.051 6.407 pw_gather_s 912 13.2 4.471 4.948 4.471 4.948 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.121688, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1504.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/44ddd7182aebe9205a531925afe90506688fd6a0_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 452.657152E+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 43626. 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 0.031 56.018 56.019 qs_mol_dyn_low 1 2.0 0.003 0.003 55.775 55.783 qs_forces 11 3.9 0.003 0.004 55.714 55.715 qs_energies 11 4.9 0.001 0.002 54.196 54.208 scf_env_do_scf 11 5.9 0.000 0.001 47.841 47.841 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 45.620 45.621 qs_scf_new_mos 108 7.5 0.000 0.001 35.196 35.537 qs_scf_loop_do_ot 108 8.5 0.000 0.001 35.196 35.537 dbcsr_multiply_generic 2286 12.5 0.102 0.106 34.869 35.445 ot_scf_mini 108 9.5 0.002 0.002 33.492 33.704 multiply_cannon 2286 13.5 0.188 0.200 26.788 28.797 multiply_cannon_loop 2286 14.5 1.824 1.952 26.036 28.053 velocity_verlet 10 3.0 0.002 0.002 27.916 27.917 ot_mini 108 10.5 0.001 0.001 20.314 20.609 qs_ot_get_derivative 108 11.5 0.001 0.001 17.230 17.446 mp_waitall_1 245248 16.5 8.843 15.067 8.843 15.067 multiply_cannon_metrocomm3 54864 15.5 0.074 0.082 6.243 13.358 multiply_cannon_multrec 54864 15.5 3.633 5.798 7.793 11.148 qs_ot_get_p 119 10.4 0.001 0.001 8.401 8.700 rebuild_ks_matrix 119 8.3 0.000 0.000 8.096 8.291 qs_ks_build_kohn_sham_matrix 119 9.3 0.011 0.012 8.096 8.291 mp_sum_l 7287 12.8 5.876 7.708 5.876 7.708 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.140 7.320 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 6.133 6.620 multiply_cannon_sync_h2d 54864 15.5 5.150 6.388 5.150 6.388 dbcsr_mm_accdrv_process 76910 16.1 1.839 2.916 4.073 5.834 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.442 5.562 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.348 5.403 init_scf_run 11 5.9 0.000 0.001 5.022 5.022 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.022 5.022 sum_up_and_integrate 119 10.3 0.001 0.002 4.616 4.622 integrate_v_rspace 119 11.3 0.002 0.002 4.604 4.612 cp_dbcsr_syevd 50 12.0 0.002 0.003 4.572 4.572 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.349 4.484 calculate_rho_elec 119 8.7 0.011 0.017 4.348 4.484 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.331 4.331 cp_fm_redistribute_end 50 14.0 2.209 4.298 2.217 4.302 cp_fm_diag_elpa_base 50 14.0 2.078 4.176 2.083 4.181 calculate_dm_sparse 119 9.5 0.000 0.000 3.046 3.204 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.984 3.201 apply_single 119 13.6 0.000 0.000 2.983 3.201 multiply_cannon_metrocomm1 54864 15.5 0.056 0.062 1.829 3.131 acc_transpose_blocks 54864 15.5 0.227 0.255 2.272 2.833 jit_kernel_multiply 13 15.8 2.171 2.824 2.171 2.824 ot_diis_step 108 11.5 0.006 0.006 2.808 2.809 calculate_first_density_matrix 1 7.0 0.000 0.001 2.674 2.683 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.595 2.677 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.466 2.468 density_rs2pw 119 9.7 0.004 0.004 2.334 2.465 wfi_extrapolate 11 7.9 0.001 0.001 2.269 2.269 init_scf_loop 11 6.9 0.000 0.000 2.195 2.195 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.109 2.184 grid_integrate_task_list 119 12.3 2.017 2.120 2.017 2.120 mp_sum_d 4135 12.0 1.447 2.040 1.447 2.040 potential_pw2rs 119 12.3 0.004 0.004 1.975 1.996 pw_transfer 1439 11.6 0.052 0.058 1.883 1.965 make_m2s 4572 13.5 0.053 0.056 1.885 1.927 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.808 1.891 make_images 4572 14.5 0.133 0.140 1.803 1.845 mp_alltoall_d11v 2130 13.8 1.435 1.638 1.435 1.638 fft3d_ps 1201 14.6 0.374 0.480 1.523 1.603 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.511 1.540 transfer_rs2pw 487 10.6 0.005 0.006 1.428 1.529 fft_wrap_pw1pw2_140 487 13.2 0.140 0.155 1.382 1.465 acc_transpose_blocks_sync 164592 16.5 1.199 1.450 1.199 1.450 mp_waitany 12084 13.8 1.250 1.435 1.250 1.435 transfer_pw2rs 487 13.2 0.006 0.006 1.413 1.435 grid_collocate_task_list 119 9.7 1.349 1.430 1.349 1.430 dbcsr_dot_sd 1205 11.9 0.049 0.060 0.789 1.203 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=56.019000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=431.272727, yerr=1.135454 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/44ddd7182aebe9205a531925afe90506688fd6a0_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 488.161280E+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 1577285. 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 0.010 0.027 39.727 39.728 qs_mol_dyn_low 1 2.0 0.003 0.004 39.512 39.521 qs_forces 11 3.9 0.003 0.003 39.218 39.219 qs_energies 11 4.9 0.003 0.018 37.526 37.531 scf_env_do_scf 11 5.9 0.001 0.002 32.106 32.106 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 29.617 29.617 dbcsr_multiply_generic 2286 12.5 0.106 0.108 22.269 22.597 qs_scf_new_mos 108 7.5 0.001 0.001 20.886 21.117 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.885 21.116 ot_scf_mini 108 9.5 0.002 0.003 19.947 20.113 velocity_verlet 10 3.0 0.002 0.002 18.797 18.803 multiply_cannon 2286 13.5 0.209 0.218 16.869 18.278 multiply_cannon_loop 2286 14.5 1.192 1.247 15.641 17.158 ot_mini 108 10.5 0.001 0.001 12.208 12.441 mp_waitall_1 200699 16.5 5.598 10.689 5.598 10.689 qs_ot_get_derivative 108 11.5 0.001 0.001 9.839 10.007 multiply_cannon_metrocomm3 27432 15.5 0.073 0.076 3.966 9.352 multiply_cannon_multrec 27432 15.5 1.833 4.145 6.143 8.965 rebuild_ks_matrix 119 8.3 0.000 0.000 6.840 6.984 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.840 6.983 dbcsr_mm_accdrv_process 47894 16.0 3.508 6.053 4.231 6.311 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.012 6.141 qs_ot_get_p 119 10.4 0.001 0.001 4.795 5.016 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.662 4.574 init_scf_run 11 5.9 0.001 0.003 4.167 4.168 scf_env_initial_rho_setup 11 6.9 0.009 0.012 4.166 4.168 mp_sum_l 7287 12.8 2.120 4.119 2.120 4.119 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.073 4.048 apply_single 119 13.6 0.000 0.000 3.073 4.048 sum_up_and_integrate 119 10.3 0.001 0.001 3.751 3.757 integrate_v_rspace 119 11.3 0.002 0.003 3.737 3.742 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.692 3.723 calculate_rho_elec 119 8.7 0.021 0.024 3.691 3.722 make_m2s 4572 13.5 0.051 0.053 2.825 3.200 qs_ot_p2m_diag 50 11.0 0.008 0.012 3.111 3.129 make_images 4572 14.5 0.206 0.246 2.737 3.115 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.711 2.712 calculate_first_density_matrix 1 7.0 0.000 0.001 2.590 2.603 init_scf_loop 11 6.9 0.000 0.000 2.466 2.466 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.345 2.431 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.369 2.369 cp_fm_redistribute_end 50 14.0 1.198 2.338 1.202 2.341 ot_diis_step 108 11.5 0.011 0.011 2.321 2.322 cp_fm_diag_elpa_base 50 14.0 1.105 2.238 1.134 2.284 multiply_cannon_sync_h2d 27432 15.5 1.709 2.274 1.709 2.274 calculate_dm_sparse 119 9.5 0.000 0.001 2.082 2.170 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.107 2.110 density_rs2pw 119 9.7 0.004 0.004 1.987 2.067 acc_transpose_blocks 27432 15.5 0.114 0.119 1.616 1.952 grid_integrate_task_list 119 12.3 1.842 1.948 1.842 1.948 pw_transfer 1439 11.6 0.065 0.069 1.894 1.929 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.855 1.895 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.802 1.839 make_images_data 4572 15.5 0.047 0.054 1.342 1.755 jit_kernel_multiply 9 16.4 0.665 1.738 0.665 1.738 hybrid_alltoall_any 4725 16.4 0.054 0.116 1.184 1.657 prepare_preconditioner 11 7.9 0.000 0.000 1.561 1.586 make_preconditioner 11 8.9 0.000 0.001 1.561 1.586 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.456 1.515 wfi_extrapolate 11 7.9 0.001 0.001 1.506 1.507 potential_pw2rs 119 12.3 0.006 0.007 1.493 1.501 fft3d_ps 1201 14.6 0.519 0.577 1.438 1.471 fft_wrap_pw1pw2_140 487 13.2 0.160 0.170 1.410 1.448 mp_alltoall_d11v 2130 13.8 1.252 1.397 1.252 1.397 grid_collocate_task_list 119 9.7 1.287 1.341 1.287 1.341 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.253 1.265 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.214 1.262 transfer_rs2pw 487 10.6 0.005 0.005 1.161 1.218 mp_allgather_i34 2286 14.5 0.649 1.190 0.649 1.190 mp_sum_d 4135 12.0 0.558 1.044 0.558 1.044 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 0.973 0.975 transfer_pw2rs 487 13.2 0.004 0.006 0.940 0.945 acc_transpose_blocks_sync 82296 16.5 0.818 0.945 0.818 0.945 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.886 0.901 acc_transpose_blocks_kernels 27432 16.5 0.188 0.275 0.658 0.872 make_images_sizes 4572 15.5 0.005 0.005 0.563 0.810 mp_alltoall_i44 4572 16.5 0.558 0.805 0.558 0.805 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=39.728000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=465.000000, yerr=1.705606 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/44ddd7182aebe9205a531925afe90506688fd6a0_performance_tests/05/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 59.051995E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3143552 0.0% 0.0% 100.0% average stack size 0.0 0.0 46.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 522.141696E+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 0.027 31.895 31.896 qs_mol_dyn_low 1 2.0 0.003 0.003 31.614 31.629 qs_forces 11 3.9 0.004 0.008 31.554 31.554 qs_energies 11 4.9 0.002 0.002 30.012 30.014 scf_env_do_scf 11 5.9 0.000 0.001 25.422 25.423 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 22.997 22.997 dbcsr_multiply_generic 2286 12.5 0.095 0.096 16.895 16.955 qs_scf_new_mos 108 7.5 0.001 0.001 15.405 15.420 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.404 15.419 velocity_verlet 10 3.0 0.002 0.002 15.013 15.015 ot_scf_mini 108 9.5 0.002 0.002 14.651 14.663 multiply_cannon 2286 13.5 0.194 0.198 13.748 14.373 multiply_cannon_loop 2286 14.5 0.859 0.890 12.973 13.626 ot_mini 108 10.5 0.001 0.001 9.080 9.093 qs_ot_get_derivative 108 11.5 0.001 0.001 7.602 7.614 multiply_cannon_multrec 18288 15.5 1.907 2.901 6.969 7.351 rebuild_ks_matrix 119 8.3 0.000 0.000 5.928 5.938 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.928 5.938 dbcsr_mm_accdrv_process 38222 16.0 4.910 5.711 4.968 5.774 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.231 5.246 mp_waitall_1 158411 16.6 2.684 3.743 2.684 3.743 sum_up_and_integrate 119 10.3 0.001 0.002 3.520 3.524 integrate_v_rspace 119 11.3 0.003 0.003 3.507 3.513 init_scf_run 11 5.9 0.000 0.001 3.428 3.428 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.428 3.428 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.898 3.424 qs_ot_get_p 119 10.4 0.001 0.001 3.335 3.368 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.214 3.217 calculate_rho_elec 119 8.7 0.031 0.031 3.214 3.216 init_scf_loop 11 6.9 0.000 0.000 2.409 2.409 multiply_cannon_metrocomm3 18288 15.5 0.047 0.049 1.483 2.363 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.924 2.337 apply_single 119 13.6 0.000 0.000 1.924 2.337 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.233 2.240 calculate_first_density_matrix 1 7.0 0.000 0.001 2.215 2.216 make_m2s 4572 13.5 0.043 0.045 1.785 1.942 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.934 1.935 density_rs2pw 119 9.7 0.004 0.004 1.788 1.897 grid_integrate_task_list 119 12.3 1.798 1.892 1.798 1.892 make_images 4572 14.5 0.193 0.205 1.701 1.857 calculate_dm_sparse 119 9.5 0.000 0.000 1.833 1.842 pw_transfer 1439 11.6 0.065 0.069 1.797 1.806 acc_transpose_blocks 18288 15.5 0.080 0.082 1.632 1.748 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.720 1.727 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.704 1.715 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.689 1.700 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.697 1.699 cp_fm_diag_elpa_base 50 14.0 1.666 1.680 1.688 1.698 prepare_preconditioner 11 7.9 0.000 0.000 1.643 1.646 make_preconditioner 11 8.9 0.000 0.000 1.643 1.646 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.503 1.586 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.529 1.536 ot_diis_step 108 11.5 0.011 0.012 1.462 1.462 mp_sum_l 7287 12.8 0.977 1.423 0.977 1.423 fft_wrap_pw1pw2_140 487 13.2 0.212 0.217 1.376 1.386 potential_pw2rs 119 12.3 0.007 0.008 1.321 1.324 grid_collocate_task_list 119 9.7 1.233 1.319 1.233 1.319 fft3d_ps 1201 14.6 0.525 0.544 1.272 1.281 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.175 1.180 wfi_extrapolate 11 7.9 0.001 0.001 1.161 1.161 multiply_cannon_sync_h2d 18288 15.5 1.033 1.141 1.033 1.141 transfer_rs2pw 487 10.6 0.005 0.005 0.953 1.066 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.939 0.957 make_images_data 4572 15.5 0.048 0.052 0.768 0.945 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 0.940 0.941 acc_transpose_blocks_sync 54864 16.5 0.745 0.858 0.745 0.858 hybrid_alltoall_any 4725 16.4 0.058 0.117 0.670 0.852 transfer_pw2rs 487 13.2 0.004 0.004 0.806 0.810 acc_transpose_blocks_kernels 18288 16.5 0.217 0.226 0.788 0.799 mp_alltoall_d11v 2130 13.8 0.653 0.784 0.653 0.784 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.780 0.782 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.678 0.739 multiply_cannon_metrocomm1 18288 15.5 0.030 0.031 0.475 0.701 cp_fm_cholesky_invert 11 10.9 0.692 0.696 0.692 0.696 mp_alltoall_z22v 1201 16.6 0.598 0.663 0.598 0.663 mp_waitany 9880 13.7 0.525 0.640 0.525 0.640 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=31.896000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.090909, yerr=2.193152 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/44ddd7182aebe9205a531925afe90506688fd6a0_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 559.009792E+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 0.012 0.032 36.423 36.425 qs_mol_dyn_low 1 2.0 0.003 0.003 36.234 36.242 qs_forces 11 3.9 0.002 0.003 35.995 35.995 qs_energies 11 4.9 0.001 0.001 34.298 34.305 scf_env_do_scf 11 5.9 0.000 0.001 29.285 29.286 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 25.942 25.942 dbcsr_multiply_generic 2286 12.5 0.098 0.102 19.907 20.058 velocity_verlet 10 3.0 0.002 0.003 18.563 18.572 qs_scf_new_mos 108 7.5 0.001 0.001 17.964 18.015 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.963 18.014 multiply_cannon 2286 13.5 0.218 0.228 16.226 17.103 ot_scf_mini 108 9.5 0.002 0.003 16.931 16.977 multiply_cannon_loop 2286 14.5 1.533 1.603 15.293 16.006 ot_mini 108 10.5 0.001 0.001 10.483 10.548 multiply_cannon_multrec 27432 15.5 2.477 3.147 9.025 9.361 qs_ot_get_derivative 108 11.5 0.001 0.001 8.615 8.663 dbcsr_mm_accdrv_process 47916 15.9 5.977 7.675 6.444 7.759 rebuild_ks_matrix 119 8.3 0.000 0.000 6.271 6.324 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.270 6.323 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.557 5.603 qs_ot_get_p 119 10.4 0.001 0.001 3.608 3.673 init_scf_run 11 5.9 0.000 0.001 3.623 3.623 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.622 3.623 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.158 3.523 sum_up_and_integrate 119 10.3 0.001 0.001 3.475 3.482 integrate_v_rspace 119 11.3 0.003 0.003 3.463 3.471 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.302 3.337 calculate_rho_elec 119 8.7 0.040 0.046 3.301 3.337 init_scf_loop 11 6.9 0.000 0.000 3.325 3.325 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.093 2.598 apply_single 119 13.6 0.000 0.000 2.093 2.597 acc_transpose_blocks 27432 15.5 0.119 0.123 2.411 2.530 prepare_preconditioner 11 7.9 0.000 0.000 2.501 2.509 make_preconditioner 11 8.9 0.000 0.000 2.501 2.509 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.113 2.434 mp_waitall_1 137007 16.6 1.857 2.383 1.857 2.383 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.342 2.367 make_m2s 4572 13.5 0.054 0.056 2.243 2.341 make_images 4572 14.5 0.274 0.335 2.137 2.233 calculate_first_density_matrix 1 7.0 0.000 0.000 2.216 2.218 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.201 2.213 calculate_dm_sparse 119 9.5 0.000 0.000 2.131 2.182 pw_transfer 1439 11.6 0.065 0.069 1.907 1.943 grid_integrate_task_list 119 12.3 1.830 1.928 1.830 1.928 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.814 1.853 density_rs2pw 119 9.7 0.004 0.004 1.773 1.836 ot_diis_step 108 11.5 0.012 0.012 1.825 1.826 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.805 1.806 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.801 1.802 mp_sum_l 7287 12.8 1.014 1.782 1.014 1.782 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.632 1.644 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.559 1.567 acc_transpose_blocks_sync 82296 16.5 1.444 1.566 1.444 1.566 cp_fm_diag_elpa_base 50 14.0 1.525 1.541 1.556 1.564 fft_wrap_pw1pw2_140 487 13.2 0.246 0.259 1.495 1.536 multiply_cannon_metrocomm3 27432 15.5 0.041 0.042 0.867 1.533 fft3d_ps 1201 14.6 0.553 0.601 1.334 1.356 wfi_extrapolate 11 7.9 0.001 0.001 1.354 1.354 grid_collocate_task_list 119 9.7 1.249 1.349 1.249 1.349 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.291 1.310 potential_pw2rs 119 12.3 0.008 0.010 1.283 1.287 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.226 1.237 cp_fm_upper_to_full 72 14.2 0.812 1.150 0.812 1.150 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.124 1.125 jit_kernel_multiply 6 16.2 0.402 1.072 0.402 1.072 dbcsr_complete_redistribute 329 12.2 0.117 0.140 0.748 1.020 transfer_rs2pw 487 10.6 0.004 0.005 0.889 0.960 make_images_data 4572 15.5 0.047 0.051 0.827 0.933 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.799 0.874 hybrid_alltoall_any 4725 16.4 0.065 0.154 0.718 0.849 mp_alltoall_d11v 2130 13.8 0.713 0.845 0.713 0.845 acc_transpose_blocks_kernels 27432 16.5 0.271 0.279 0.818 0.834 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.555 0.822 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.804 0.808 transfer_pw2rs 487 13.2 0.004 0.004 0.735 0.739 cp_fm_cholesky_invert 11 10.9 0.731 0.734 0.731 0.734 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=36.425000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=531.272727, yerr=2.894852 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/44ddd7182aebe9205a531925afe90506688fd6a0_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 622.878720E+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 0.008 0.029 28.221 28.223 qs_mol_dyn_low 1 2.0 0.003 0.003 28.051 28.060 qs_forces 11 3.9 0.002 0.003 27.978 27.979 qs_energies 11 4.9 0.001 0.001 26.291 26.293 scf_env_do_scf 11 5.9 0.000 0.001 21.674 21.674 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 19.245 19.246 velocity_verlet 10 3.0 0.002 0.002 14.526 14.529 dbcsr_multiply_generic 2286 12.5 0.097 0.103 12.902 12.969 qs_scf_new_mos 108 7.5 0.001 0.001 11.702 11.727 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.702 11.726 ot_scf_mini 108 9.5 0.002 0.002 10.994 11.019 multiply_cannon 2286 13.5 0.223 0.231 10.352 10.813 multiply_cannon_loop 2286 14.5 0.643 0.661 9.461 9.638 ot_mini 108 10.5 0.001 0.001 6.280 6.311 multiply_cannon_multrec 9144 15.5 1.668 1.845 5.921 6.151 rebuild_ks_matrix 119 8.3 0.000 0.000 5.693 5.715 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.693 5.715 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.067 5.086 qs_ot_get_derivative 108 11.5 0.001 0.001 4.994 5.020 dbcsr_mm_accdrv_process 12550 15.8 3.503 4.143 4.141 4.219 sum_up_and_integrate 119 10.3 0.001 0.001 3.398 3.403 integrate_v_rspace 119 11.3 0.003 0.004 3.388 3.393 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.299 3.307 calculate_rho_elec 119 8.7 0.060 0.061 3.299 3.307 init_scf_run 11 5.9 0.000 0.001 3.186 3.186 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.186 3.186 qs_ot_get_p 119 10.4 0.001 0.001 2.841 2.872 init_scf_loop 11 6.9 0.000 0.000 2.409 2.411 calculate_first_density_matrix 1 7.0 0.000 0.000 2.071 2.072 pw_transfer 1439 11.6 0.066 0.068 1.956 1.964 grid_integrate_task_list 119 12.3 1.867 1.953 1.867 1.953 mp_waitall_1 115863 16.7 1.348 1.892 1.348 1.892 make_m2s 4572 13.5 0.034 0.034 1.756 1.881 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.862 1.871 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.864 1.867 make_images 4572 14.5 0.270 0.301 1.668 1.792 density_rs2pw 119 9.7 0.003 0.004 1.693 1.783 calculate_dm_sparse 119 9.5 0.000 0.000 1.728 1.745 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.696 1.718 prepare_preconditioner 11 7.9 0.000 0.000 1.692 1.697 make_preconditioner 11 8.9 0.000 0.000 1.692 1.697 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.622 1.623 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.585 1.613 fft_wrap_pw1pw2_140 487 13.2 0.322 0.331 1.536 1.547 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.512 1.514 acc_transpose_blocks 9144 15.5 0.041 0.043 1.419 1.448 jit_kernel_multiply 8 15.6 0.599 1.416 0.599 1.416 grid_collocate_task_list 119 9.7 1.296 1.385 1.296 1.385 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.360 1.367 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.338 1.350 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.324 1.332 cp_fm_diag_elpa_base 50 14.0 1.297 1.315 1.322 1.330 fft3d_ps 1201 14.6 0.560 0.571 1.276 1.286 ot_diis_step 108 11.5 0.013 0.013 1.272 1.272 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.216 1.246 apply_single 119 13.6 0.000 0.000 1.216 1.245 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.219 1.220 potential_pw2rs 119 12.3 0.010 0.011 1.216 1.219 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.117 1.121 wfi_extrapolate 11 7.9 0.001 0.001 1.064 1.064 hybrid_alltoall_any 4725 16.4 0.065 0.175 0.762 0.944 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.873 0.923 make_images_data 4572 15.5 0.042 0.046 0.748 0.896 transfer_rs2pw 487 10.6 0.004 0.005 0.775 0.851 cp_fm_cholesky_invert 11 10.9 0.843 0.845 0.843 0.845 mp_alltoall_d11v 2130 13.8 0.769 0.845 0.769 0.845 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.815 0.824 acc_transpose_blocks_sync 27432 16.5 0.722 0.753 0.722 0.753 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.748 0.750 qs_env_update_s_mstruct 11 6.9 0.000 0.001 0.666 0.713 multiply_cannon_metrocomm3 9144 15.5 0.020 0.021 0.375 0.672 transfer_pw2rs 487 13.2 0.003 0.004 0.651 0.654 acc_transpose_blocks_kernels 9144 16.5 0.118 0.120 0.639 0.648 mp_alltoall_z22v 1201 16.6 0.590 0.621 0.590 0.621 mp_allgather_i34 2286 14.5 0.208 0.618 0.208 0.618 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=28.223000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=590.545455, yerr=5.805384 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/44ddd7182aebe9205a531925afe90506688fd6a0_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 766.521344E+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 0.013 0.048 42.407 42.421 qs_mol_dyn_low 1 2.0 0.003 0.003 42.191 42.207 qs_forces 11 3.9 0.003 0.003 42.130 42.130 qs_energies 11 4.9 0.001 0.001 40.136 40.138 scf_env_do_scf 11 5.9 0.001 0.001 34.543 34.543 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 26.819 26.821 velocity_verlet 10 3.0 0.002 0.002 23.849 23.855 dbcsr_multiply_generic 2286 12.5 0.100 0.104 19.123 19.358 qs_scf_new_mos 108 7.5 0.001 0.001 17.496 17.598 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.495 17.597 ot_scf_mini 108 9.5 0.002 0.002 16.341 16.444 multiply_cannon 2286 13.5 0.294 0.303 15.067 16.141 multiply_cannon_loop 2286 14.5 0.856 0.862 13.814 14.887 ot_mini 108 10.5 0.001 0.001 10.038 10.156 multiply_cannon_multrec 9144 15.5 3.496 5.041 8.884 9.112 qs_ot_get_derivative 108 11.5 0.001 0.001 7.966 8.067 init_scf_loop 11 6.9 0.000 0.000 7.696 7.699 rebuild_ks_matrix 119 8.3 0.000 0.000 7.057 7.202 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.057 7.201 dbcsr_mm_accdrv_process 12550 15.8 4.575 6.470 5.244 6.754 prepare_preconditioner 11 7.9 0.000 0.000 6.737 6.752 make_preconditioner 11 8.9 0.000 0.000 6.737 6.752 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.310 6.630 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.362 6.493 cp_fm_upper_to_full 72 14.2 3.138 4.508 3.138 4.508 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.983 3.987 calculate_rho_elec 119 8.7 0.118 0.121 3.983 3.987 sum_up_and_integrate 119 10.3 0.001 0.001 3.753 3.760 integrate_v_rspace 119 11.3 0.004 0.005 3.743 3.749 init_scf_run 11 5.9 0.000 0.001 3.533 3.533 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.533 3.533 qs_ot_get_p 119 10.4 0.001 0.001 3.378 3.518 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.988 3.476 mp_waitall_1 94719 16.7 2.310 3.432 2.310 3.432 dbcsr_complete_redistribute 329 12.2 0.287 0.293 1.968 2.784 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.305 2.775 apply_single 119 13.6 0.000 0.000 2.305 2.774 pw_transfer 1439 11.6 0.069 0.069 2.578 2.583 make_m2s 4572 13.5 0.037 0.038 2.426 2.577 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.479 2.484 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.667 2.473 make_images 4572 14.5 0.352 0.384 2.306 2.455 multiply_cannon_metrocomm3 9144 15.5 0.021 0.022 1.312 2.394 mp_alltoall_i22 627 13.8 1.399 2.243 1.399 2.243 calculate_dm_sparse 119 9.5 0.000 0.000 2.222 2.236 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.422 2.221 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.070 2.122 fft_wrap_pw1pw2_140 487 13.2 0.576 0.579 2.094 2.102 grid_integrate_task_list 119 12.3 2.080 2.101 2.080 2.101 calculate_first_density_matrix 1 7.0 0.000 0.000 2.070 2.070 ot_diis_step 108 11.5 0.014 0.014 2.046 2.047 density_rs2pw 119 9.7 0.003 0.003 2.018 2.038 mp_sum_l 7287 12.8 1.087 1.999 1.087 1.999 qs_ot_p2m_diag 50 11.0 0.044 0.044 1.997 1.998 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.801 1.802 acc_transpose_blocks 9144 15.5 0.044 0.045 1.794 1.802 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.778 1.779 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.642 1.643 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.547 1.585 fft3d_ps 1201 14.6 0.596 0.605 1.574 1.577 grid_collocate_task_list 119 9.7 1.523 1.553 1.523 1.553 cp_fm_cholesky_invert 11 10.9 1.424 1.427 1.424 1.427 wfi_extrapolate 11 7.9 0.001 0.001 1.399 1.399 hybrid_alltoall_any 4725 16.4 0.089 0.149 1.135 1.371 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.353 1.353 cp_fm_diag_elpa_base 50 14.0 1.210 1.261 1.351 1.351 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.303 1.321 potential_pw2rs 119 12.3 0.014 0.014 1.264 1.265 make_images_data 4572 15.5 0.046 0.048 1.073 1.265 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.233 1.238 mp_alltoall_d11v 2130 13.8 1.163 1.188 1.163 1.188 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.101 1.118 acc_transpose_blocks_sync 27432 16.5 1.087 1.096 1.087 1.096 jit_kernel_multiply 6 15.6 0.640 1.041 0.640 1.041 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 0.997 1.034 qs_create_task_list 11 7.9 0.001 0.001 0.931 0.944 generate_qs_task_list 11 8.9 0.368 0.387 0.930 0.943 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.906 0.920 mp_alltoall_z22v 1201 16.6 0.841 0.861 0.841 0.861 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=42.421000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=720.181818, yerr=13.217194 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/44ddd7182aebe9205a531925afe90506688fd6a0_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 501.657600E+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 1350805. 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 0.008 0.027 82.495 82.496 qs_mol_dyn_low 1 2.0 0.003 0.003 82.244 82.253 qs_forces 11 3.9 0.003 0.003 82.174 82.175 qs_energies 11 4.9 0.001 0.002 79.320 79.333 scf_env_do_scf 11 5.9 0.000 0.001 70.425 70.428 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 64.969 64.969 dbcsr_multiply_generic 2055 12.4 0.115 0.137 51.734 52.010 qs_scf_new_mos 99 7.5 0.000 0.001 48.039 48.160 qs_scf_loop_do_ot 99 8.5 0.000 0.001 48.038 48.160 ot_scf_mini 99 9.5 0.002 0.002 45.644 45.736 multiply_cannon 2055 13.4 0.182 0.186 42.936 43.640 velocity_verlet 10 3.0 0.001 0.002 43.328 43.329 multiply_cannon_loop 2055 14.4 1.794 1.846 41.992 42.697 ot_mini 99 10.5 0.001 0.001 27.100 27.200 qs_ot_get_derivative 99 11.5 0.001 0.001 20.291 20.414 multiply_cannon_multrec 49320 15.4 11.310 11.944 17.313 17.928 rebuild_ks_matrix 110 8.3 0.000 0.000 14.405 14.526 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.405 14.526 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.624 12.722 mp_waitall_1 220248 16.4 11.241 12.032 11.241 12.032 qs_ot_get_p 110 10.4 0.001 0.001 9.958 10.069 multiply_cannon_sync_h2d 49320 15.4 9.543 10.001 9.543 10.001 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.877 8.419 multiply_cannon_metrocomm3 49320 15.4 0.084 0.086 6.860 7.795 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.171 7.619 apply_single 110 13.6 0.000 0.000 7.170 7.619 sum_up_and_integrate 110 10.3 0.002 0.003 6.959 6.972 integrate_v_rspace 110 11.3 0.003 0.003 6.933 6.953 init_scf_run 11 5.9 0.000 0.001 6.868 6.868 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.868 6.868 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.746 6.796 ot_diis_step 99 11.5 0.006 0.006 6.545 6.545 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.290 6.443 calculate_rho_elec 110 8.6 0.020 0.024 6.290 6.443 dbcsr_mm_accdrv_process 87628 16.1 3.053 3.160 5.875 6.138 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.704 5.704 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.566 5.629 init_scf_loop 11 6.9 0.000 0.000 5.428 5.429 mp_sum_l 6594 12.7 4.475 5.227 4.475 5.227 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.135 5.160 cp_fm_diag_elpa_base 48 14.0 5.122 5.146 5.133 5.158 wfi_extrapolate 11 7.9 0.001 0.001 4.002 4.002 calculate_dm_sparse 110 9.5 0.001 0.001 3.893 3.995 multiply_cannon_metrocomm1 49320 15.4 0.065 0.067 2.581 3.972 make_m2s 4110 13.4 0.060 0.065 3.852 3.956 make_images 4110 14.4 0.177 0.189 3.757 3.865 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.553 3.557 density_rs2pw 110 9.6 0.004 0.005 3.297 3.429 grid_integrate_task_list 110 12.3 3.276 3.413 3.276 3.413 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.331 3.383 prepare_preconditioner 11 7.9 0.000 0.000 3.275 3.296 make_preconditioner 11 8.9 0.000 0.000 3.275 3.296 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.205 3.268 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.069 3.121 pw_transfer 1331 11.6 0.055 0.066 3.033 3.113 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 2.945 3.025 calculate_first_density_matrix 1 7.0 0.000 0.000 2.763 2.770 acc_transpose_blocks 49320 15.4 0.212 0.219 2.584 2.740 fft_wrap_pw1pw2_140 451 13.1 0.390 0.427 2.526 2.613 mp_alltoall_d11v 2046 13.8 2.094 2.595 2.094 2.595 jit_kernel_multiply 13 15.9 2.543 2.581 2.543 2.581 potential_pw2rs 110 12.3 0.005 0.006 2.551 2.567 fft3d_ps 1111 14.6 0.794 0.881 2.240 2.300 grid_collocate_task_list 110 9.6 2.162 2.281 2.162 2.281 mp_waitany 14300 13.8 1.810 2.152 1.810 2.152 transfer_rs2pw 451 10.6 0.005 0.006 1.863 1.998 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.911 1.933 mp_sum_d 3889 11.9 1.417 1.899 1.417 1.899 make_images_data 4110 15.4 0.043 0.047 1.737 1.870 hybrid_alltoall_any 4261 16.3 0.085 0.486 1.541 1.831 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.775 1.804 cp_fm_cholesky_invert 11 10.9 1.796 1.800 1.796 1.800 transfer_pw2rs 451 13.1 0.006 0.007 1.732 1.742 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=82.496000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.272727, yerr=2.489648 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/44ddd7182aebe9205a531925afe90506688fd6a0_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 593.575936E+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 1560146. 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 68.500 68.501 qs_mol_dyn_low 1 2.0 0.003 0.003 68.287 68.296 qs_forces 11 3.9 0.003 0.003 68.147 68.148 qs_energies 11 4.9 0.001 0.002 64.788 64.793 scf_env_do_scf 11 5.9 0.000 0.001 56.338 56.342 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 48.821 48.821 dbcsr_multiply_generic 2055 12.4 0.117 0.121 37.742 37.969 velocity_verlet 10 3.0 0.020 0.036 35.932 35.934 qs_scf_new_mos 99 7.5 0.001 0.001 33.058 33.199 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.058 33.198 multiply_cannon 2055 13.4 0.224 0.244 31.173 32.203 ot_scf_mini 99 9.5 0.003 0.003 31.382 31.509 multiply_cannon_loop 2055 14.4 1.164 1.189 30.036 31.010 ot_mini 99 10.5 0.001 0.001 18.316 18.457 multiply_cannon_multrec 24660 15.4 7.006 8.920 14.050 15.868 rebuild_ks_matrix 110 8.3 0.000 0.000 13.519 13.661 qs_ks_build_kohn_sham_matrix 110 9.3 0.030 0.062 13.518 13.660 qs_ot_get_derivative 99 11.5 0.001 0.001 12.560 12.691 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.855 11.989 mp_waitall_1 176588 16.5 7.469 9.791 7.469 9.791 multiply_cannon_metrocomm3 24660 15.4 0.073 0.076 5.119 7.958 init_scf_loop 11 6.9 0.000 0.000 7.482 7.483 multiply_cannon_sync_h2d 24660 15.4 6.391 7.330 6.391 7.330 dbcsr_mm_accdrv_process 52282 16.1 5.455 6.238 6.874 7.182 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.432 7.135 apply_single 110 13.6 0.000 0.001 6.432 7.135 qs_ot_get_p 110 10.4 0.001 0.001 6.450 6.581 sum_up_and_integrate 110 10.3 0.001 0.002 6.346 6.357 integrate_v_rspace 110 11.3 0.003 0.003 6.318 6.332 init_scf_run 11 5.9 0.000 0.001 6.073 6.074 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.073 6.074 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.854 5.863 calculate_rho_elec 110 8.6 0.039 0.047 5.853 5.863 ot_diis_step 99 11.5 0.010 0.010 5.708 5.708 prepare_preconditioner 11 7.9 0.000 0.000 5.469 5.485 make_preconditioner 11 8.9 0.000 0.000 5.469 5.485 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.715 5.443 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.054 5.208 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.569 4.589 make_m2s 4110 13.4 0.057 0.059 4.201 4.516 make_images 4110 14.4 0.410 0.468 4.092 4.402 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.102 4.103 pw_transfer 1331 11.6 0.067 0.074 3.530 3.664 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.422 3.560 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.539 3.557 cp_fm_diag_elpa_base 48 14.0 3.492 3.512 3.536 3.554 wfi_extrapolate 11 7.9 0.001 0.001 3.500 3.500 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.288 3.357 density_rs2pw 110 9.6 0.004 0.005 3.112 3.345 grid_integrate_task_list 110 12.3 3.159 3.314 3.159 3.314 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.238 3.239 fft_wrap_pw1pw2_140 451 13.1 0.461 0.475 2.925 3.063 calculate_dm_sparse 110 9.5 0.001 0.001 2.976 3.001 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.893 2.962 fft3d_ps 1111 14.6 1.108 1.337 2.518 2.671 make_images_data 4110 15.4 0.049 0.053 2.284 2.606 cp_fm_cholesky_invert 11 10.9 2.557 2.564 2.557 2.564 hybrid_alltoall_any 4261 16.3 0.106 0.457 1.978 2.495 calculate_first_density_matrix 1 7.0 0.000 0.001 2.477 2.480 grid_collocate_task_list 110 9.6 2.170 2.329 2.170 2.329 mp_sum_l 6594 12.7 1.734 2.293 1.734 2.293 potential_pw2rs 110 12.3 0.008 0.009 2.233 2.248 jit_kernel_multiply 11 16.4 1.055 2.122 1.055 2.122 mp_alltoall_d11v 2046 13.8 1.722 2.022 1.722 2.022 acc_transpose_blocks 24660 15.4 0.116 0.119 1.972 1.993 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.967 1.987 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.851 1.852 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.823 1.835 multiply_cannon_metrocomm4 22605 15.4 0.078 0.083 0.776 1.752 transfer_rs2pw 451 10.6 0.007 0.007 1.411 1.584 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.535 1.546 mp_irecv_dv 57340 16.2 0.647 1.526 0.647 1.526 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.385 1.487 mp_allgather_i34 2055 14.4 0.445 1.443 0.445 1.443 mp_waitany 10164 13.8 1.188 1.385 1.188 1.385 dbcsr_complete_redistribute 325 12.2 0.242 0.294 1.104 1.372 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=68.501000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=560.090909, yerr=7.267043 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/44ddd7182aebe9205a531925afe90506688fd6a0_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 665.714688E+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 0.040 0.060 59.704 59.705 qs_mol_dyn_low 1 2.0 0.003 0.003 59.379 59.389 qs_forces 11 3.9 0.003 0.003 59.303 59.304 qs_energies 11 4.9 0.001 0.002 56.126 56.129 scf_env_do_scf 11 5.9 0.000 0.001 48.319 48.319 scf_env_do_scf_inner_loop 99 6.5 0.003 0.006 39.844 39.844 velocity_verlet 10 3.0 0.002 0.002 32.445 32.447 dbcsr_multiply_generic 2055 12.4 0.108 0.113 28.727 29.020 qs_scf_new_mos 99 7.5 0.001 0.001 25.282 25.388 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.281 25.387 ot_scf_mini 99 9.5 0.002 0.002 24.032 24.169 multiply_cannon 2055 13.4 0.214 0.227 22.439 23.547 multiply_cannon_loop 2055 14.4 0.818 0.857 21.360 22.387 ot_mini 99 10.5 0.001 0.001 13.775 13.909 rebuild_ks_matrix 110 8.3 0.000 0.000 12.071 12.246 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.016 12.070 12.246 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.628 10.787 multiply_cannon_multrec 16440 15.4 3.597 4.434 9.820 10.599 mp_waitall_1 139946 16.5 7.015 10.133 7.015 10.133 qs_ot_get_derivative 99 11.5 0.001 0.001 9.332 9.468 init_scf_loop 11 6.9 0.000 0.000 8.439 8.440 multiply_cannon_metrocomm3 16440 15.4 0.046 0.048 4.454 7.478 prepare_preconditioner 11 7.9 0.000 0.000 6.701 6.721 make_preconditioner 11 8.9 0.000 0.000 6.701 6.721 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.016 6.377 dbcsr_mm_accdrv_process 34862 16.1 5.302 5.694 6.067 6.293 sum_up_and_integrate 110 10.3 0.001 0.002 6.243 6.257 integrate_v_rspace 110 11.3 0.003 0.003 6.217 6.232 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.718 5.727 calculate_rho_elec 110 8.6 0.059 0.059 5.717 5.726 qs_ot_get_p 110 10.4 0.001 0.001 5.431 5.591 init_scf_run 11 5.9 0.000 0.001 5.430 5.430 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.430 5.430 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.943 5.338 apply_single 110 13.6 0.000 0.000 4.942 5.338 make_m2s 4110 13.4 0.049 0.051 4.182 4.459 ot_diis_step 99 11.5 0.011 0.011 4.411 4.412 make_images 4110 14.4 0.400 0.526 4.068 4.345 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.180 3.810 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.783 3.787 multiply_cannon_sync_h2d 16440 15.4 3.202 3.735 3.202 3.735 pw_transfer 1331 11.6 0.066 0.072 3.551 3.561 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.443 3.456 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.407 3.408 grid_integrate_task_list 110 12.3 3.196 3.401 3.196 3.401 density_rs2pw 110 9.6 0.004 0.005 2.959 3.148 fft_wrap_pw1pw2_140 451 13.1 0.580 0.586 2.974 2.989 wfi_extrapolate 11 7.9 0.001 0.001 2.977 2.977 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.863 2.872 cp_fm_diag_elpa_base 48 14.0 2.797 2.829 2.862 2.870 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.777 2.778 make_images_data 4110 15.4 0.045 0.050 2.363 2.702 cp_fm_cholesky_invert 11 10.9 2.632 2.637 2.632 2.637 hybrid_alltoall_any 4261 16.3 0.109 0.387 2.077 2.588 calculate_dm_sparse 110 9.5 0.001 0.001 2.527 2.556 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.448 2.521 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.442 2.492 multiply_cannon_metrocomm4 14385 15.4 0.049 0.052 0.895 2.483 grid_collocate_task_list 110 9.6 2.219 2.434 2.219 2.434 fft3d_ps 1111 14.6 1.091 1.103 2.383 2.394 calculate_first_density_matrix 1 7.0 0.000 0.001 2.363 2.364 mp_irecv_dv 48980 15.7 0.819 2.350 0.819 2.350 mp_sum_l 6594 12.7 1.482 2.118 1.482 2.118 potential_pw2rs 110 12.3 0.011 0.011 2.108 2.113 mp_alltoall_d11v 2046 13.8 1.746 1.981 1.746 1.981 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.941 1.942 dbcsr_complete_redistribute 325 12.2 0.324 0.372 1.422 1.882 cp_fm_upper_to_full 70 14.2 1.410 1.796 1.410 1.796 acc_transpose_blocks 16440 15.4 0.077 0.079 1.583 1.672 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.637 1.653 cp_fm_cholesky_decompose 22 10.9 1.549 1.573 1.549 1.573 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.375 1.489 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.463 1.476 mp_allgather_i34 2055 14.4 0.391 1.464 0.391 1.464 transfer_rs2pw 451 10.6 0.005 0.006 1.253 1.427 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.973 1.423 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.286 1.296 mp_waitany 17072 13.8 1.125 1.270 1.125 1.270 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=59.705000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=630.272727, yerr=8.571527 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/44ddd7182aebe9205a531925afe90506688fd6a0_performance_tests/12/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 601.317074E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 4916280 0.0% 0.0% 100.0% average stack size 0.0 0.0 200.2 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 732.999680E+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 0.056 0.075 65.694 65.695 qs_mol_dyn_low 1 2.0 0.003 0.004 65.388 65.398 qs_forces 11 3.9 0.003 0.003 65.323 65.324 qs_energies 11 4.9 0.002 0.006 61.943 61.947 scf_env_do_scf 11 5.9 0.001 0.001 53.700 53.703 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 42.165 42.166 velocity_verlet 10 3.0 0.002 0.002 37.040 37.042 dbcsr_multiply_generic 2055 12.4 0.116 0.120 30.583 30.833 qs_scf_new_mos 99 7.5 0.001 0.001 27.671 27.778 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.671 27.778 ot_scf_mini 99 9.5 0.002 0.003 25.990 26.097 multiply_cannon 2055 13.4 0.236 0.254 23.579 24.674 multiply_cannon_loop 2055 14.4 1.410 1.462 22.236 22.881 ot_mini 99 10.5 0.001 0.001 14.997 15.107 multiply_cannon_multrec 24660 15.4 4.100 6.889 12.995 14.090 rebuild_ks_matrix 110 8.3 0.000 0.000 11.859 11.973 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.859 11.973 init_scf_loop 11 6.9 0.000 0.000 11.495 11.496 qs_ot_get_derivative 99 11.5 0.001 0.001 10.798 10.903 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.477 10.577 prepare_preconditioner 11 7.9 0.000 0.000 9.781 9.796 make_preconditioner 11 8.9 0.000 0.000 9.781 9.796 dbcsr_mm_accdrv_process 52304 16.0 7.761 9.000 8.738 9.642 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.032 9.469 sum_up_and_integrate 110 10.3 0.001 0.002 6.193 6.206 integrate_v_rspace 110 11.3 0.003 0.003 6.167 6.179 mp_waitall_1 121746 16.5 4.294 6.169 4.294 6.169 qs_ot_get_p 110 10.4 0.001 0.001 5.912 6.075 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.795 5.804 calculate_rho_elec 110 8.6 0.078 0.081 5.794 5.804 make_m2s 4110 13.4 0.059 0.060 5.403 5.656 make_images 4110 14.4 0.582 0.704 5.265 5.516 init_scf_run 11 5.9 0.000 0.001 5.466 5.467 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.466 5.466 cp_fm_upper_to_full 70 14.2 3.296 4.775 3.296 4.775 ot_diis_step 99 11.5 0.011 0.011 4.161 4.161 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.067 4.161 apply_single 110 13.6 0.000 0.000 4.067 4.160 qs_ot_p2m_diag 48 11.0 0.055 0.064 4.057 4.075 dbcsr_complete_redistribute 325 12.2 0.428 0.487 2.645 3.760 pw_transfer 1331 11.6 0.066 0.074 3.639 3.672 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.532 3.568 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.485 3.537 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.496 3.496 grid_integrate_task_list 110 12.3 3.275 3.415 3.275 3.415 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.349 3.403 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.110 3.217 multiply_cannon_metrocomm3 24660 15.4 0.038 0.039 1.553 3.169 density_rs2pw 110 9.6 0.004 0.004 2.918 3.105 fft_wrap_pw1pw2_140 451 13.1 0.610 0.630 3.038 3.079 calculate_dm_sparse 110 9.5 0.001 0.001 2.982 3.019 wfi_extrapolate 11 7.9 0.001 0.001 2.981 2.981 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.956 2.965 cp_fm_diag_elpa_base 48 14.0 2.812 2.873 2.954 2.964 make_images_data 4110 15.4 0.049 0.053 2.635 2.912 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.740 2.824 hybrid_alltoall_any 4261 16.3 0.123 0.464 2.297 2.808 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.788 2.790 mp_alltoall_i22 605 13.7 1.618 2.788 1.618 2.788 cp_fm_cholesky_invert 11 10.9 2.672 2.679 2.672 2.679 multiply_cannon_sync_h2d 24660 15.4 2.365 2.538 2.365 2.538 acc_transpose_blocks 24660 15.4 0.114 0.117 2.386 2.499 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.419 2.457 fft3d_ps 1111 14.6 1.084 1.118 2.414 2.425 grid_collocate_task_list 110 9.6 2.265 2.415 2.265 2.415 calculate_first_density_matrix 1 7.0 0.000 0.000 2.390 2.392 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.279 2.280 mp_alltoall_d11v 2046 13.8 1.761 2.067 1.761 2.067 potential_pw2rs 110 12.3 0.013 0.013 2.012 2.015 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.798 1.832 jit_kernel_multiply 10 15.7 0.630 1.823 0.630 1.823 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.610 1.714 cp_fm_cholesky_decompose 22 10.9 1.645 1.675 1.645 1.675 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.623 1.636 mp_sum_l 6594 12.7 0.922 1.602 0.922 1.602 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.546 1.560 multiply_cannon_metrocomm4 20550 15.4 0.063 0.066 0.865 1.513 acc_transpose_blocks_sync 73980 16.4 1.372 1.475 1.372 1.475 mp_allgather_i34 2055 14.4 0.431 1.457 0.431 1.457 mp_irecv_dv 62702 16.1 0.759 1.429 0.759 1.429 transfer_rs2pw 451 10.6 0.005 0.006 1.192 1.401 mp_waitany 13376 13.8 1.038 1.333 1.038 1.333 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=65.695000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=695.545455, yerr=8.239193 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/44ddd7182aebe9205a531925afe90506688fd6a0_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 864.374784E+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 0.017 0.152 55.599 55.619 qs_mol_dyn_low 1 2.0 0.003 0.003 55.244 55.271 qs_forces 11 3.9 0.003 0.003 55.177 55.178 qs_energies 11 4.9 0.001 0.002 51.531 51.535 scf_env_do_scf 11 5.9 0.000 0.001 43.340 43.340 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 35.646 35.647 velocity_verlet 10 3.0 0.002 0.002 31.039 31.042 dbcsr_multiply_generic 2055 12.4 0.105 0.109 23.554 23.777 qs_scf_new_mos 99 7.5 0.001 0.001 21.271 21.356 qs_scf_loop_do_ot 99 8.5 0.001 0.001 21.271 21.356 ot_scf_mini 99 9.5 0.002 0.002 20.016 20.084 multiply_cannon 2055 13.4 0.237 0.247 18.060 19.071 multiply_cannon_loop 2055 14.4 0.603 0.627 16.833 17.237 rebuild_ks_matrix 110 8.3 0.000 0.000 11.496 11.554 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 11.495 11.554 ot_mini 99 10.5 0.001 0.001 11.146 11.218 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.219 10.272 multiply_cannon_multrec 8220 15.4 3.427 5.205 7.943 9.257 init_scf_loop 11 6.9 0.000 0.000 7.646 7.648 mp_waitall_1 103326 16.6 5.892 7.604 5.892 7.604 qs_ot_get_derivative 99 11.5 0.001 0.001 7.362 7.430 sum_up_and_integrate 110 10.3 0.001 0.002 6.141 6.155 integrate_v_rspace 110 11.3 0.003 0.004 6.114 6.128 prepare_preconditioner 11 7.9 0.000 0.000 6.014 6.020 make_preconditioner 11 8.9 0.000 0.000 6.014 6.020 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.906 5.921 calculate_rho_elec 110 8.6 0.114 0.114 5.906 5.920 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.592 5.672 dbcsr_mm_accdrv_process 17442 15.9 3.188 4.480 4.377 5.577 init_scf_run 11 5.9 0.000 0.001 5.075 5.075 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.075 5.075 qs_ot_get_p 110 10.4 0.001 0.001 4.918 4.996 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 3.147 4.721 make_m2s 4110 13.4 0.038 0.039 4.147 4.333 make_images 4110 14.4 0.652 0.711 4.017 4.200 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.805 3.866 apply_single 110 13.6 0.000 0.000 3.805 3.866 pw_transfer 1331 11.6 0.066 0.071 3.827 3.837 ot_diis_step 99 11.5 0.012 0.012 3.762 3.762 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.719 3.733 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.542 3.546 grid_integrate_task_list 110 12.3 3.377 3.535 3.377 3.535 fft_wrap_pw1pw2_140 451 13.1 0.779 0.792 3.214 3.229 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.202 3.203 density_rs2pw 110 9.6 0.004 0.004 2.908 3.067 cp_fm_cholesky_invert 11 10.9 2.905 2.909 2.905 2.909 wfi_extrapolate 11 7.9 0.001 0.001 2.690 2.690 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.660 2.661 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.630 2.637 cp_fm_diag_elpa_base 48 14.0 2.575 2.596 2.628 2.636 hybrid_alltoall_any 4261 16.3 0.200 0.854 2.205 2.583 calculate_dm_sparse 110 9.5 0.001 0.001 2.510 2.557 grid_collocate_task_list 110 9.6 2.368 2.524 2.368 2.524 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.511 2.512 multiply_cannon_sync_h2d 8220 15.4 2.336 2.452 2.336 2.452 make_images_data 4110 15.4 0.040 0.047 2.180 2.442 fft3d_ps 1111 14.6 1.135 1.167 2.383 2.400 calculate_first_density_matrix 1 7.0 0.000 0.000 2.290 2.291 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.144 2.170 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.018 2.064 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.784 2.008 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.919 1.956 mp_alltoall_d11v 2046 13.8 1.728 1.956 1.728 1.956 potential_pw2rs 110 12.3 0.016 0.016 1.946 1.949 cp_fm_cholesky_decompose 22 10.9 1.651 1.671 1.651 1.671 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.605 1.611 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.488 1.606 dbcsr_complete_redistribute 325 12.2 0.560 0.596 1.443 1.536 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.459 1.473 multiply_cannon_metrocomm4 6165 15.4 0.019 0.020 0.483 1.431 mp_allgather_i34 2055 14.4 0.396 1.420 0.396 1.420 acc_transpose_blocks 8220 15.4 0.039 0.041 1.365 1.401 mp_irecv_dv 24056 15.7 0.456 1.385 0.456 1.385 multiply_cannon_metrocomm1 8220 15.4 0.022 0.023 0.742 1.381 qs_create_task_list 11 7.9 0.001 0.001 1.207 1.306 generate_qs_task_list 11 8.9 0.374 0.442 1.207 1.306 transfer_rs2pw 451 10.6 0.005 0.005 1.108 1.282 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.219 1.250 jit_kernel_multiply 6 15.5 0.870 1.239 0.870 1.239 mp_waitany 9240 13.8 1.034 1.212 1.034 1.212 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=55.619000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=811.636364, yerr=10.679400 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/44ddd7182aebe9205a531925afe90506688fd6a0_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 1.448481E+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 0.015 0.043 86.857 86.868 qs_mol_dyn_low 1 2.0 0.003 0.003 86.481 86.490 qs_forces 11 3.9 0.003 0.003 86.415 86.415 qs_energies 11 4.9 0.001 0.001 82.320 82.322 scf_env_do_scf 11 5.9 0.001 0.001 72.341 72.342 velocity_verlet 10 3.0 0.002 0.002 55.497 55.503 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 44.192 44.193 dbcsr_multiply_generic 2055 12.4 0.119 0.124 29.772 29.952 init_scf_loop 11 6.9 0.000 0.000 28.079 28.081 qs_scf_new_mos 99 7.5 0.001 0.001 27.083 27.173 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.082 27.172 prepare_preconditioner 11 7.9 0.000 0.000 26.099 26.114 make_preconditioner 11 8.9 0.000 0.000 26.099 26.114 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.391 25.570 ot_scf_mini 99 9.5 0.002 0.002 25.296 25.381 multiply_cannon 2055 13.4 0.335 0.364 22.566 23.232 multiply_cannon_loop 2055 14.4 0.834 0.862 20.747 21.162 cp_fm_upper_to_full 70 14.2 12.651 18.081 12.651 18.081 ot_mini 99 10.5 0.001 0.001 14.378 14.462 rebuild_ks_matrix 110 8.3 0.000 0.001 13.476 13.578 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 13.475 13.578 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.225 12.318 dbcsr_complete_redistribute 325 12.2 1.027 1.049 7.275 10.513 multiply_cannon_multrec 8220 15.4 4.079 4.233 9.779 9.918 qs_ot_get_derivative 99 11.5 0.001 0.001 9.600 9.686 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.254 9.497 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.693 8.907 mp_waitall_1 84994 16.7 7.563 8.556 7.563 8.556 mp_alltoall_i22 605 13.7 5.345 8.552 5.345 8.552 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.996 7.032 calculate_rho_elec 110 8.6 0.225 0.225 6.996 7.031 sum_up_and_integrate 110 10.3 0.002 0.002 6.676 6.691 integrate_v_rspace 110 11.3 0.004 0.004 6.648 6.664 make_m2s 4110 13.4 0.043 0.043 5.350 5.860 dbcsr_mm_accdrv_process 11614 15.7 3.883 4.155 5.552 5.795 qs_ot_get_p 110 10.4 0.001 0.001 5.614 5.715 init_scf_run 11 5.9 0.000 0.001 5.713 5.713 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.712 5.713 make_images 4110 14.4 0.886 0.937 5.162 5.670 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.896 5.423 apply_single 110 13.6 0.000 0.000 4.895 5.423 multiply_cannon_metrocomm3 8220 15.4 0.019 0.020 4.913 5.251 cp_fm_cholesky_invert 11 10.9 5.231 5.235 5.231 5.235 pw_transfer 1331 11.6 0.075 0.075 4.855 4.865 ot_diis_step 99 11.5 0.015 0.016 4.760 4.761 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 4.737 4.748 fft_wrap_pw1pw2_140 451 13.1 1.277 1.283 4.142 4.156 qs_ot_p2m_diag 48 11.0 0.151 0.156 3.991 4.000 grid_integrate_task_list 110 12.3 3.673 3.738 3.673 3.738 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.655 3.655 hybrid_alltoall_any 4261 16.3 0.263 0.564 2.817 3.536 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.051 3.536 density_rs2pw 110 9.6 0.004 0.004 3.461 3.477 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.473 3.473 make_images_data 4110 15.4 0.045 0.048 2.821 3.454 wfi_extrapolate 11 7.9 0.001 0.001 3.320 3.321 calculate_dm_sparse 110 9.5 0.001 0.001 3.190 3.217 multiply_cannon_sync_h2d 8220 15.4 3.109 3.143 3.109 3.143 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.899 2.902 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.884 2.884 cp_fm_diag_elpa_base 48 14.0 2.347 2.540 2.882 2.882 fft3d_ps 1111 14.6 1.297 1.301 2.792 2.802 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.688 2.732 grid_collocate_task_list 110 9.6 2.665 2.677 2.665 2.677 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.380 2.415 calculate_first_density_matrix 1 7.0 0.000 0.000 2.279 2.280 qs_env_update_s_mstruct 11 6.9 0.001 0.001 2.166 2.225 potential_pw2rs 110 12.3 0.022 0.022 2.208 2.215 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.090 2.176 mp_alltoall_d11v 2046 13.8 2.031 2.089 2.031 2.089 cp_fm_cholesky_decompose 22 10.9 1.963 1.980 1.963 1.980 qs_create_task_list 11 7.9 0.001 0.001 1.882 1.927 generate_qs_task_list 11 8.9 0.730 0.783 1.881 1.927 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.850 1.860 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.783 1.841 acc_transpose_blocks 8220 15.4 0.041 0.041 1.750 1.792 jit_kernel_multiply 10 15.1 1.471 1.753 1.471 1.753 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=86.868000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1286.545455, yerr=77.269733 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/44ddd7182aebe9205a531925afe90506688fd6a0_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 631.099392E+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 2361358. 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 0.022 0.040 205.860 205.862 qs_mol_dyn_low 1 2.0 0.003 0.005 205.408 205.422 qs_forces 11 3.9 0.005 0.005 205.141 205.142 qs_energies 11 4.9 0.004 0.021 199.565 199.581 scf_env_do_scf 11 5.9 0.001 0.002 183.220 183.224 scf_env_do_scf_inner_loop 117 6.6 0.004 0.018 162.997 162.998 dbcsr_multiply_generic 2507 12.6 0.181 0.185 125.770 126.747 qs_scf_new_mos 117 7.6 0.001 0.001 123.963 124.287 qs_scf_loop_do_ot 117 8.6 0.001 0.001 123.963 124.286 velocity_verlet 10 3.0 0.007 0.018 122.979 122.982 ot_scf_mini 117 9.6 0.003 0.004 117.258 117.574 multiply_cannon 2507 13.6 0.239 0.249 101.628 104.185 multiply_cannon_loop 2507 14.6 2.384 2.434 99.531 102.093 ot_mini 117 10.6 0.001 0.001 66.443 66.744 multiply_cannon_multrec 60168 15.6 31.839 33.957 41.577 43.712 qs_ot_get_derivative 117 11.6 0.001 0.001 41.510 41.793 rebuild_ks_matrix 128 8.3 0.001 0.001 33.356 33.799 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 33.356 33.799 mp_waitall_1 267128 16.5 28.832 31.560 28.832 31.560 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.955 30.368 qs_ot_get_p 128 10.4 0.001 0.001 28.468 28.851 multiply_cannon_sync_h2d 60168 15.6 26.393 28.674 26.393 28.674 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.333 25.599 apply_single 128 13.6 0.001 0.001 24.332 25.599 ot_diis_step 117 11.6 0.008 0.008 24.690 24.691 qs_ot_p2m_diag 83 11.4 0.079 0.091 21.966 22.047 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 20.193 20.432 init_scf_loop 11 6.9 0.001 0.004 20.151 20.152 cp_dbcsr_syevd 83 12.4 0.005 0.005 19.165 19.165 multiply_cannon_metrocomm3 60168 15.6 0.119 0.124 16.191 18.708 cp_fm_diag_elpa 83 13.4 0.000 0.000 16.260 16.297 cp_fm_diag_elpa_base 83 14.4 16.193 16.234 16.256 16.294 prepare_preconditioner 11 7.9 0.000 0.000 15.573 15.628 make_preconditioner 11 8.9 0.000 0.001 15.573 15.628 make_full_inverse_cholesky 11 9.9 0.000 0.000 14.818 15.002 make_m2s 5014 13.6 0.104 0.113 13.718 14.142 make_images 5014 14.6 0.398 0.416 13.535 13.967 sum_up_and_integrate 128 10.3 0.002 0.004 13.841 13.857 integrate_v_rspace 128 11.3 0.004 0.004 13.782 13.800 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.939 13.083 calculate_rho_elec 128 8.7 0.045 0.064 12.939 13.082 init_scf_run 11 5.9 0.000 0.001 12.180 12.180 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.179 12.180 mp_sum_l 7950 12.9 9.160 10.754 9.160 10.754 dbcsr_mm_accdrv_process 124484 16.2 4.687 4.812 9.303 9.812 wfi_extrapolate 11 7.9 0.001 0.001 8.986 8.987 cp_fm_cholesky_invert 11 10.9 8.860 8.868 8.860 8.868 calculate_dm_sparse 128 9.5 0.001 0.001 8.509 8.611 qs_ot_get_orbitals 117 10.6 0.001 0.001 8.227 8.317 multiply_cannon_metrocomm1 60168 15.6 0.095 0.100 6.315 8.257 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.957 8.109 pw_transfer 1547 11.6 0.074 0.087 7.482 7.658 make_images_data 5014 15.6 0.066 0.073 6.705 7.652 grid_integrate_task_list 128 12.3 7.070 7.487 7.070 7.487 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.279 7.453 density_rs2pw 128 9.7 0.006 0.007 6.603 7.188 hybrid_alltoall_any 5200 16.5 0.295 2.265 5.907 7.110 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.625 6.635 fft_wrap_pw1pw2_140 523 13.2 1.128 1.170 6.433 6.598 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.765 5.865 fft3d_ps 1291 14.7 2.191 2.843 5.285 5.624 grid_collocate_task_list 128 9.7 4.847 5.260 4.847 5.260 mp_alltoall_d11v 2415 14.1 4.232 5.258 4.232 5.258 cp_fm_cholesky_decompose 22 10.9 4.548 4.562 4.548 4.562 potential_pw2rs 128 12.3 0.009 0.010 4.350 4.363 mp_sum_d 4465 12.1 3.536 4.309 3.536 4.309 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=205.862000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=597.181818, yerr=6.739547 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/44ddd7182aebe9205a531925afe90506688fd6a0_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410023282688 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444707676160 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796579E+12 0.0% 0.0% 100.0% flops max/rank 2.166472E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500928 0.0% 0.0% 100.0% number of processed stacks 5925696 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.6 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 840.736768E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2385600 MPI messages size (bytes): total size 4.069300E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.705776E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70188 2295595008 32768 < size <= 131072 716032 54973693952 131072 < size <= 4194304 1363760 1386318135296 4194304 < size <= 16777216 153648 1453843137296 16777216 < size 67056 1171888537600 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4003 58199. MP_Allreduce 11086 960. MP_Sync 86 MP_Alltoall 1955 4796080. MP_SendRecv 11938 47072. MP_ISendRecv 11938 47072. MP_Wait 25718 MP_ISend 11660 212488. MP_IRecv 11660 212488. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.053 186.755 186.771 qs_mol_dyn_low 1 2.0 0.003 0.003 186.215 186.413 qs_forces 11 3.9 0.005 0.005 186.128 186.129 qs_energies 11 4.9 0.002 0.002 179.402 179.412 scf_env_do_scf 11 5.9 0.001 0.001 163.227 163.237 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 130.728 130.730 velocity_verlet 10 3.0 0.002 0.002 117.192 117.194 dbcsr_multiply_generic 2485 12.5 0.188 0.194 96.677 97.866 qs_scf_new_mos 116 7.6 0.001 0.001 92.784 93.360 qs_scf_loop_do_ot 116 8.6 0.001 0.001 92.783 93.359 ot_scf_mini 116 9.6 0.004 0.004 88.029 88.638 multiply_cannon 2485 13.5 0.504 0.571 76.810 81.583 multiply_cannon_loop 2485 14.5 1.557 1.630 73.409 76.036 ot_mini 116 10.6 0.001 0.001 49.869 50.478 mp_waitall_1 212858 16.6 23.810 39.381 23.810 39.381 multiply_cannon_multrec 29820 15.5 20.828 25.271 31.352 36.218 rebuild_ks_matrix 127 8.3 0.001 0.001 31.897 32.654 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.018 31.897 32.653 init_scf_loop 11 6.9 0.000 0.000 32.411 32.412 qs_ks_update_qs_env 127 7.6 0.001 0.001 28.655 29.339 multiply_cannon_metrocomm3 29820 15.5 0.097 0.102 15.667 29.052 qs_ot_get_derivative 116 11.6 0.001 0.002 28.070 28.677 prepare_preconditioner 11 7.9 0.000 0.000 28.116 28.192 make_preconditioner 11 8.9 0.000 0.000 28.116 28.192 make_full_inverse_cholesky 11 9.9 0.000 0.000 26.816 27.360 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 21.942 23.133 apply_single 127 13.6 0.001 0.001 21.942 23.132 ot_diis_step 116 11.6 0.014 0.015 21.619 21.621 qs_ot_get_p 127 10.4 0.001 0.001 20.352 21.028 multiply_cannon_sync_h2d 29820 15.5 17.860 20.230 17.860 20.230 cp_fm_cholesky_invert 11 10.9 16.422 16.434 16.422 16.434 make_m2s 4970 13.5 0.088 0.093 13.922 15.670 qs_ot_p2m_diag 82 11.4 0.186 0.214 15.480 15.514 make_images 4970 14.5 1.171 1.387 13.712 15.459 cp_dbcsr_syevd 82 12.4 0.005 0.006 14.304 14.304 sum_up_and_integrate 127 10.3 0.002 0.004 13.754 13.779 integrate_v_rspace 127 11.3 0.003 0.004 13.693 13.721 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.119 13.164 calculate_rho_elec 127 8.7 0.086 0.103 13.119 13.163 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 11.037 11.495 init_scf_run 11 5.9 0.000 0.001 11.392 11.394 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.392 11.393 cp_fm_diag_elpa 82 13.4 0.000 0.000 11.225 11.253 cp_fm_diag_elpa_base 82 14.4 10.977 11.062 11.219 11.243 multiply_cannon_metrocomm4 27335 15.5 0.105 0.119 3.786 10.864 dbcsr_mm_accdrv_process 61726 16.2 5.424 6.326 9.976 10.660 mp_irecv_dv 68888 16.3 3.582 10.460 3.582 10.460 make_images_data 4970 15.5 0.066 0.075 8.215 10.404 hybrid_alltoall_any 5155 16.4 0.349 1.497 6.987 9.757 pw_transfer 1535 11.6 0.084 0.098 8.398 8.463 wfi_extrapolate 11 7.9 0.001 0.001 8.317 8.317 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.176 8.238 grid_integrate_task_list 127 12.3 7.141 7.530 7.141 7.530 fft_wrap_pw1pw2_140 519 13.2 1.208 1.228 7.275 7.351 density_rs2pw 127 9.7 0.006 0.007 6.876 7.259 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.335 7.091 cp_fm_cholesky_decompose 22 10.9 6.814 6.902 6.814 6.902 calculate_dm_sparse 127 9.5 0.001 0.001 6.421 6.579 mp_sum_l 7884 12.9 4.165 6.254 4.165 6.254 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.054 6.064 fft3d_ps 1281 14.7 2.816 2.995 5.751 5.807 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.181 5.374 qs_ot_get_orbitals 116 10.6 0.001 0.001 5.312 5.373 grid_collocate_task_list 127 9.7 4.965 5.287 4.965 5.287 mp_allgather_i34 2485 14.5 1.924 5.055 1.924 5.055 mp_alltoall_d11v 2401 14.1 4.073 4.481 4.073 4.481 potential_pw2rs 127 12.3 0.015 0.017 4.323 4.335 dbcsr_complete_redistribute 393 12.7 0.772 0.875 3.053 3.867 mp_sum_d 4452 12.1 2.591 3.853 2.591 3.853 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=186.771000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=800.363636, yerr=0.771389 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/44ddd7182aebe9205a531925afe90506688fd6a0_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 978.784256E+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 4020 57952. MP_Allreduce 11127 998. 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 0.017 0.128 174.527 174.566 qs_mol_dyn_low 1 2.0 0.003 0.004 173.956 174.041 qs_forces 11 3.9 0.004 0.005 173.854 173.855 qs_energies 11 4.9 0.001 0.002 167.350 167.362 scf_env_do_scf 11 5.9 0.001 0.001 151.154 151.155 scf_env_do_scf_inner_loop 117 6.6 0.007 0.016 116.449 116.450 velocity_verlet 10 3.0 0.003 0.004 111.547 111.549 dbcsr_multiply_generic 2507 12.6 0.180 0.187 82.649 83.808 qs_scf_new_mos 117 7.6 0.001 0.001 79.551 79.907 qs_scf_loop_do_ot 117 8.6 0.001 0.001 79.551 79.906 ot_scf_mini 117 9.6 0.003 0.004 75.311 75.716 multiply_cannon 2507 13.6 0.509 0.524 62.907 68.518 multiply_cannon_loop 2507 14.6 1.125 1.183 59.922 62.768 ot_mini 117 10.6 0.001 0.001 42.495 42.878 init_scf_loop 11 6.9 0.000 0.000 34.608 34.609 mp_waitall_1 170520 16.6 24.609 34.592 24.609 34.592 prepare_preconditioner 11 7.9 0.000 0.000 30.570 30.611 make_preconditioner 11 8.9 0.000 0.000 30.570 30.611 rebuild_ks_matrix 128 8.3 0.001 0.001 30.116 30.588 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.018 30.116 30.588 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.189 29.653 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.150 27.576 multiply_cannon_multrec 20056 15.6 13.135 16.035 22.518 25.498 multiply_cannon_metrocomm3 20056 15.6 0.063 0.066 15.268 24.849 qs_ot_get_derivative 117 11.6 0.002 0.002 23.019 23.416 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.615 20.597 apply_single 128 13.6 0.001 0.001 19.615 20.597 ot_diis_step 117 11.6 0.018 0.019 19.370 19.370 qs_ot_get_p 128 10.4 0.001 0.001 18.180 18.636 make_m2s 5014 13.6 0.079 0.084 14.634 15.914 make_images 5014 14.6 1.156 1.248 14.400 15.668 multiply_cannon_sync_h2d 20056 15.6 13.632 15.174 13.632 15.174 cp_fm_cholesky_invert 11 10.9 14.120 14.128 14.120 14.128 qs_ot_p2m_diag 83 11.4 0.265 0.272 13.854 13.863 sum_up_and_integrate 128 10.3 0.002 0.003 13.820 13.846 integrate_v_rspace 128 11.3 0.003 0.004 13.761 13.784 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.430 13.468 calculate_rho_elec 128 8.7 0.130 0.145 13.429 13.468 cp_dbcsr_syevd 83 12.4 0.005 0.006 12.808 12.808 init_scf_run 11 5.9 0.000 0.001 11.213 11.213 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.213 11.213 make_images_data 5014 15.6 0.062 0.071 8.817 10.515 hybrid_alltoall_any 5200 16.5 0.448 2.039 7.676 10.263 cp_fm_diag_elpa 83 13.4 0.000 0.000 9.679 9.697 cp_fm_diag_elpa_base 83 14.4 9.274 9.421 9.676 9.694 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.069 9.352 multiply_cannon_metrocomm4 17549 15.6 0.067 0.077 3.465 9.241 mp_irecv_dv 50230 16.2 3.336 8.987 3.336 8.987 dbcsr_mm_accdrv_process 41502 16.2 5.617 5.955 8.840 8.972 pw_transfer 1547 11.6 0.085 0.103 8.571 8.683 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 8.347 8.465 grid_integrate_task_list 128 12.3 7.300 7.770 7.300 7.770 fft_wrap_pw1pw2_140 523 13.2 1.301 1.335 7.448 7.577 wfi_extrapolate 11 7.9 0.001 0.001 7.409 7.409 cp_fm_cholesky_decompose 22 10.9 7.225 7.309 7.225 7.309 cp_fm_upper_to_full 105 14.8 5.780 7.298 5.780 7.298 density_rs2pw 128 9.7 0.006 0.006 6.755 7.218 dbcsr_complete_redistribute 395 12.7 1.177 1.204 4.539 6.302 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.100 6.107 calculate_dm_sparse 128 9.5 0.001 0.001 5.836 5.934 fft3d_ps 1291 14.7 2.763 2.988 5.724 5.826 grid_collocate_task_list 128 9.7 5.154 5.565 5.154 5.565 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.205 5.340 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.568 5.276 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.408 5.166 mp_sum_l 7950 12.9 3.321 4.961 3.321 4.961 mp_alltoall_d11v 2415 14.1 4.301 4.782 4.301 4.782 mp_allgather_i34 2507 14.6 1.476 4.737 1.476 4.737 potential_pw2rs 128 12.3 0.021 0.023 4.240 4.249 transfer_fm_to_dbcsr 11 9.9 0.019 0.024 2.361 4.078 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.886 3.915 mp_alltoall_i22 716 14.1 1.980 3.872 1.980 3.872 qs_energies_init_hamiltonians 11 5.9 0.002 0.004 3.734 3.734 calculate_first_density_matrix 1 7.0 0.000 0.000 3.618 3.621 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.466 3.511 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=174.566000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=925.000000, yerr=20.823937 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/44ddd7182aebe9205a531925afe90506688fd6a0_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430458527744 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613083000832 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228655E+12 0.0% 0.0% 100.0% flops max/rank 4.387043E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806314816 0.0% 0.0% 100.0% number of processed stacks 6024576 0.0% 0.0% 100.0% average stack size 0.0 0.0 1129.8 marketing flops 145.647559E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 1.185497E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1152312 MPI messages size (bytes): total size 2.039354E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.769793E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 321600 36333158400 131072 < size <= 4194304 721560 792041881600 4194304 < size <= 16777216 70800 669921130560 16777216 < size 30960 541065216000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4030 57809. MP_Allreduce 11152 1080. MP_Sync 87 MP_Alltoall 1724 12509439. MP_SendRecv 5934 75008. MP_ISendRecv 5934 75008. MP_Wait 22612 MP_ISend 15064 244788. MP_IRecv 15064 244788. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.039 186.487 186.497 qs_mol_dyn_low 1 2.0 0.003 0.003 186.137 186.150 qs_forces 11 3.9 0.004 0.005 185.908 185.916 qs_energies 11 4.9 0.003 0.011 178.837 178.846 scf_env_do_scf 11 5.9 0.001 0.001 162.347 162.357 velocity_verlet 10 3.0 0.008 0.009 124.399 124.406 scf_env_do_scf_inner_loop 118 6.6 0.003 0.009 116.407 116.409 dbcsr_multiply_generic 2527 12.6 0.189 0.195 80.848 81.653 qs_scf_new_mos 118 7.6 0.001 0.001 80.179 80.515 qs_scf_loop_do_ot 118 8.6 0.001 0.001 80.178 80.514 ot_scf_mini 118 9.6 0.003 0.003 75.583 75.912 multiply_cannon 2527 13.6 0.553 0.581 56.156 59.071 multiply_cannon_loop 2527 14.6 1.836 1.911 52.508 54.551 init_scf_loop 11 6.9 0.000 0.000 45.819 45.819 ot_mini 118 10.6 0.001 0.001 43.233 43.551 prepare_preconditioner 11 7.9 0.000 0.000 41.723 41.749 make_preconditioner 11 8.9 0.000 0.000 41.723 41.749 make_full_inverse_cholesky 11 9.9 0.012 0.023 35.314 40.329 multiply_cannon_multrec 30324 15.6 13.753 18.855 26.686 31.354 rebuild_ks_matrix 129 8.3 0.001 0.001 29.281 29.624 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.019 29.280 29.623 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.446 26.746 mp_waitall_1 149060 16.7 17.093 26.305 17.093 26.305 qs_ot_get_derivative 118 11.6 0.002 0.002 23.391 23.720 make_m2s 5054 13.6 0.096 0.100 20.256 21.304 make_images 5054 14.6 1.975 2.272 19.954 21.007 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 19.180 19.764 apply_single 129 13.6 0.001 0.001 19.179 19.764 ot_diis_step 118 11.6 0.018 0.018 19.713 19.714 qs_ot_get_p 129 10.4 0.001 0.001 18.412 18.766 cp_fm_upper_to_full 105 14.8 11.416 16.827 11.416 16.827 cp_fm_cholesky_invert 11 10.9 15.779 15.787 15.779 15.787 multiply_cannon_metrocomm3 30324 15.6 0.051 0.054 6.407 15.314 qs_ot_p2m_diag 83 11.4 0.343 0.390 14.201 14.253 sum_up_and_integrate 129 10.3 0.002 0.003 13.973 13.998 integrate_v_rspace 129 11.3 0.003 0.006 13.913 13.940 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.674 13.704 calculate_rho_elec 129 8.7 0.174 0.190 13.674 13.703 dbcsr_mm_accdrv_process 62756 16.2 8.362 9.332 12.502 13.041 cp_dbcsr_syevd 83 12.4 0.005 0.006 12.817 12.817 dbcsr_complete_redistribute 395 12.7 1.514 1.630 8.965 12.718 make_images_data 5054 15.6 0.067 0.073 10.818 12.613 copy_fm_to_dbcsr 209 11.7 0.001 0.001 7.626 11.361 multiply_cannon_sync_h2d 30324 15.6 10.578 11.269 10.578 11.269 hybrid_alltoall_any 5240 16.5 0.533 2.231 9.700 11.165 init_scf_run 11 5.9 0.000 0.001 10.587 10.588 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.587 10.588 transfer_fm_to_dbcsr 11 9.9 0.002 0.007 6.387 10.058 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.766 10.006 cp_fm_diag_elpa 83 13.4 0.000 0.000 9.781 9.790 cp_fm_diag_elpa_base 83 14.4 8.838 9.153 9.775 9.784 mp_alltoall_i22 718 14.1 5.624 9.389 5.624 9.389 pw_transfer 1559 11.6 0.087 0.102 8.910 8.983 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 8.682 8.758 grid_integrate_task_list 129 12.3 7.624 7.911 7.624 7.911 fft_wrap_pw1pw2_140 527 13.2 1.457 1.495 7.763 7.852 wfi_extrapolate 11 7.9 0.001 0.001 7.528 7.528 cp_fm_cholesky_decompose 22 10.9 7.362 7.462 7.362 7.462 multiply_cannon_metrocomm4 25270 15.6 0.085 0.096 2.826 7.223 density_rs2pw 129 9.7 0.006 0.006 6.689 7.030 mp_irecv_dv 76685 16.2 2.669 6.953 2.669 6.953 calculate_dm_sparse 129 9.5 0.001 0.001 6.295 6.426 fft3d_ps 1301 14.7 2.863 2.933 5.840 5.883 grid_collocate_task_list 129 9.7 5.320 5.833 5.320 5.833 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.379 5.417 mp_alltoall_d11v 2423 14.1 4.750 5.054 4.750 5.054 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 4.577 4.691 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 4.412 4.412 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.305 4.400 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.262 4.323 potential_pw2rs 129 12.3 0.024 0.024 4.175 4.185 dbcsr_special_finalize 12635 15.6 0.067 0.070 3.593 3.807 mp_sum_l 8010 12.9 2.492 3.752 2.492 3.752 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=186.497000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1098.909091, yerr=30.515136 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/44ddd7182aebe9205a531925afe90506688fd6a0_performance_tests/19/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 5.865089E+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 1960712 0.0% 0.0% 100.0% average stack size 0.0 0.0 3445.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.553514E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 240672 MPI messages size (bytes): total size 1.331455E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.532238E+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 113904 59718500352 4194304 < size <= 16777216 104976 550376570880 16777216 < size 20208 721350232272 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8931 51. MP_Alltoall 9654 799394. MP_ISend 40068 2102573. MP_IRecv 40068 2101676. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4020 57947. MP_Allreduce 11127 1165. MP_Sync 87 MP_Alltoall 1712 18838222. MP_SendRecv 3840 122880. MP_ISendRecv 3840 122880. MP_Wait 16122 MP_ISend 10680 423556. MP_IRecv 10680 423556. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.034 166.915 166.917 qs_mol_dyn_low 1 2.0 0.003 0.003 166.462 166.477 qs_forces 11 3.9 0.004 0.004 166.360 166.365 qs_energies 11 4.9 0.002 0.002 158.977 158.984 scf_env_do_scf 11 5.9 0.001 0.001 141.713 141.723 velocity_verlet 10 3.0 0.002 0.002 109.883 109.887 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 106.833 106.834 dbcsr_multiply_generic 2507 12.6 0.181 0.190 71.557 71.868 qs_scf_new_mos 117 7.6 0.001 0.001 70.421 70.486 qs_scf_loop_do_ot 117 8.6 0.001 0.001 70.420 70.485 ot_scf_mini 117 9.6 0.003 0.004 66.023 66.070 multiply_cannon 2507 13.6 0.561 0.591 53.136 56.566 multiply_cannon_loop 2507 14.6 0.803 0.836 50.246 50.940 ot_mini 117 10.6 0.001 0.001 37.159 37.204 init_scf_loop 11 6.9 0.000 0.000 34.731 34.733 prepare_preconditioner 11 7.9 0.000 0.000 30.888 30.915 make_preconditioner 11 8.9 0.000 0.000 30.888 30.915 mp_waitall_1 125778 16.7 24.406 30.706 24.406 30.706 make_full_inverse_cholesky 11 9.9 0.015 0.027 28.820 29.106 rebuild_ks_matrix 128 8.3 0.001 0.001 28.838 28.908 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.018 28.837 28.908 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.304 26.366 multiply_cannon_multrec 10028 15.6 10.466 14.796 18.050 20.929 qs_ot_get_derivative 117 11.6 0.001 0.002 20.225 20.269 multiply_cannon_metrocomm3 10028 15.6 0.026 0.027 12.536 19.628 cp_fm_cholesky_invert 11 10.9 17.509 17.515 17.509 17.515 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 16.959 17.075 apply_single 128 13.6 0.001 0.001 16.958 17.075 ot_diis_step 117 11.6 0.020 0.021 16.867 16.867 qs_ot_get_p 128 10.4 0.001 0.001 16.178 16.235 make_m2s 5014 13.6 0.066 0.071 14.617 15.418 make_images 5014 14.6 2.183 2.614 14.310 15.107 sum_up_and_integrate 128 10.3 0.002 0.002 13.967 14.013 integrate_v_rspace 128 11.3 0.004 0.004 13.906 13.955 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.830 13.871 calculate_rho_elec 128 8.7 0.256 0.267 13.829 13.871 qs_ot_p2m_diag 83 11.4 0.495 0.501 12.500 12.518 cp_dbcsr_syevd 83 12.4 0.005 0.006 11.350 11.352 multiply_cannon_sync_h2d 10028 15.6 10.934 11.249 10.934 11.249 init_scf_run 11 5.9 0.000 0.001 10.353 10.353 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.352 10.353 hybrid_alltoall_any 5200 16.5 0.841 3.766 8.298 9.717 make_images_data 5014 15.6 0.055 0.066 8.409 9.629 pw_transfer 1547 11.6 0.086 0.095 9.134 9.171 fft_wrap_pw1pw2 1291 12.7 0.010 0.010 8.910 8.950 cp_fm_diag_elpa 83 13.4 0.000 0.000 8.523 8.536 cp_fm_diag_elpa_base 83 14.4 8.283 8.360 8.518 8.530 grid_integrate_task_list 128 12.3 7.764 8.046 7.764 8.046 fft_wrap_pw1pw2_140 523 13.2 1.809 1.852 7.928 7.970 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 7.928 7.964 dbcsr_mm_accdrv_process 20762 16.1 3.165 4.433 7.231 7.891 cp_fm_cholesky_decompose 22 10.9 7.708 7.849 7.708 7.849 wfi_extrapolate 11 7.9 0.001 0.001 7.374 7.374 multiply_cannon_metrocomm1 10028 15.6 0.030 0.030 4.264 6.964 density_rs2pw 128 9.7 0.005 0.006 6.565 6.824 calculate_dm_sparse 128 9.5 0.001 0.001 6.058 6.136 grid_collocate_task_list 128 9.7 5.575 5.809 5.575 5.809 fft3d_ps 1291 14.7 2.763 2.842 5.635 5.666 dbcsr_complete_redistribute 395 12.7 2.104 2.150 5.176 5.560 qs_energies_init_hamiltonians 11 5.9 0.006 0.011 5.311 5.316 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.140 5.178 mp_alltoall_d11v 2415 14.1 4.528 4.955 4.528 4.955 mp_allgather_i34 2507 14.6 1.074 4.523 1.074 4.523 potential_pw2rs 128 12.3 0.027 0.028 4.148 4.153 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.042 4.081 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.590 3.897 copy_fm_to_dbcsr 209 11.7 0.001 0.002 3.344 3.670 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.557 3.620 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.558 3.591 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.435 3.552 multiply_cannon_metrocomm4 7521 15.6 0.027 0.028 1.706 3.552 mp_irecv_dv 28860 15.9 1.667 3.477 1.667 3.477 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.442 3.452 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=166.917000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1475.909091, yerr=19.261146 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/44ddd7182aebe9205a531925afe90506688fd6a0_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1389588037632 0.0% 0.0% 100.0% flops 32 x 32 x 32 1887638126592 0.0% 0.0% 100.0% flops 22 x 9 x 32 1929499951104 0.0% 0.0% 100.0% flops 9 x 22 x 32 1935094689792 0.0% 0.0% 100.0% flops 22 x 22 x 32 2675276685312 0.0% 0.0% 100.0% flops 32 x 32 x 9 4300336005120 0.0% 0.0% 100.0% flops 32 x 32 x 22 5255966228480 0.0% 0.0% 100.0% flops 9 x 32 x 32 5299016564736 0.0% 0.0% 100.0% flops 22 x 32 x 32 6476575801344 0.0% 0.0% 100.0% flops 9 x 32 x 9 11276331687936 0.0% 0.0% 100.0% flops 22 x 32 x 9 14799245598720 0.0% 0.0% 100.0% flops 9 x 32 x 22 14799245598720 0.0% 0.0% 100.0% flops 22 x 32 x 22 19338589388800 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 91.362404E+12 0.0% 0.0% 100.0% flops max/rank 11.427044E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6604657536 0.0% 0.0% 100.0% number of processed stacks 1915696 0.0% 0.0% 100.0% average stack size 0.0 0.0 3447.7 marketing flops 141.366701E+12 ------------------------------------------------------------------------------- # multiplications 2442 max memory usage/rank 3.180732E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 97680 MPI messages size (bytes): total size 1.109909E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.362705E+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 43928 34091302912 4194304 < size <= 16777216 43248 370189271040 16777216 < size 9888 705625925088 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4011 59005. MP_Allreduce 11105 1504. MP_Sync 86 MP_Alltoall 1676 36913843. MP_SendRecv 1750 218624. MP_ISendRecv 1750 218624. MP_Wait 9580 MP_ISend 6264 1080795. MP_IRecv 6264 1080795. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.023 0.064 286.707 286.722 qs_mol_dyn_low 1 2.0 0.003 0.003 286.083 286.125 qs_forces 11 3.9 0.005 0.005 285.979 285.984 qs_energies 11 4.9 0.002 0.002 277.116 277.123 scf_env_do_scf 11 5.9 0.001 0.001 254.876 254.881 velocity_verlet 10 3.0 0.014 0.015 205.820 205.827 scf_env_do_scf_inner_loop 114 6.6 0.004 0.008 130.852 130.853 init_scf_loop 11 6.9 0.000 0.000 123.754 123.756 prepare_preconditioner 11 7.9 0.000 0.000 118.976 118.996 make_preconditioner 11 8.9 0.000 0.000 118.975 118.996 make_full_inverse_cholesky 11 9.9 0.038 0.039 95.335 116.123 qs_scf_new_mos 114 7.6 0.001 0.001 88.732 88.806 qs_scf_loop_do_ot 114 8.6 0.001 0.001 88.731 88.805 ot_scf_mini 114 9.6 0.003 0.004 83.968 84.007 dbcsr_multiply_generic 2442 12.5 0.210 0.221 80.458 80.965 cp_fm_upper_to_full 104 14.8 53.141 75.659 53.141 75.659 multiply_cannon 2442 13.5 0.665 0.694 57.590 58.783 multiply_cannon_loop 2442 14.5 1.016 1.028 53.575 55.041 ot_mini 114 10.6 0.001 0.001 43.713 43.764 dbcsr_complete_redistribute 393 12.7 3.994 4.053 29.547 42.017 copy_fm_to_dbcsr 208 11.6 0.001 0.001 26.056 38.512 transfer_fm_to_dbcsr 11 9.9 0.030 0.030 23.598 35.826 mp_alltoall_i22 708 14.1 21.335 33.907 21.335 33.907 rebuild_ks_matrix 125 8.3 0.001 0.001 33.087 33.154 qs_ks_build_kohn_sham_matrix 125 9.3 0.017 0.017 33.087 33.154 cp_fm_cholesky_invert 11 10.9 33.131 33.137 33.131 33.137 mp_waitall_1 100996 16.8 27.451 31.684 27.451 31.684 qs_ks_update_qs_env 125 7.6 0.001 0.001 30.624 30.685 qs_ot_get_p 125 10.4 0.001 0.001 25.727 25.815 qs_ot_get_derivative 114 11.6 0.002 0.002 24.200 24.233 qs_ot_p2m_diag 82 11.4 0.869 0.875 21.667 21.698 cp_dbcsr_syevd 82 12.4 0.006 0.006 19.891 19.893 make_m2s 4884 13.5 0.074 0.076 18.319 19.842 multiply_cannon_metrocomm3 9768 15.5 0.024 0.026 18.270 19.748 ot_diis_step 114 11.6 0.021 0.024 19.483 19.483 make_images 4884 14.5 3.002 3.202 17.850 19.371 apply_preconditioner_dbcsr 125 12.6 0.000 0.000 18.931 19.094 apply_single 125 13.6 0.001 0.001 18.931 19.094 multiply_cannon_multrec 9768 15.5 10.136 11.953 17.794 17.897 cp_fm_diag_elpa 82 13.4 0.000 0.000 16.714 16.715 cp_fm_diag_elpa_base 82 14.4 12.385 13.880 16.709 16.710 qs_rho_update_rho_low 125 7.7 0.001 0.001 16.043 16.052 calculate_rho_elec 125 8.7 0.469 0.470 16.043 16.051 sum_up_and_integrate 125 10.3 0.002 0.002 15.623 15.707 integrate_v_rspace 125 11.3 0.004 0.005 15.562 15.646 multiply_cannon_sync_h2d 9768 15.5 13.983 14.013 13.983 14.013 hybrid_alltoall_any 5069 16.4 1.282 3.002 10.615 12.776 make_images_data 4884 15.5 0.063 0.067 10.433 12.640 init_scf_run 11 5.9 0.000 0.001 11.896 11.897 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.896 11.897 pw_transfer 1511 11.6 0.093 0.094 10.990 10.995 fft_wrap_pw1pw2 1261 12.7 0.011 0.011 10.758 10.763 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.662 9.689 fft_wrap_pw1pw2_140 511 13.2 2.983 3.027 9.601 9.604 dbcsr_mm_accdrv_process 20250 16.0 4.221 6.034 7.416 9.258 cp_fm_cholesky_decompose 22 10.9 8.839 8.865 8.839 8.865 wfi_extrapolate 11 7.9 0.001 0.001 8.810 8.810 grid_integrate_task_list 125 12.3 8.378 8.570 8.378 8.570 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 8.222 8.223 density_rs2pw 125 9.7 0.005 0.005 7.280 7.333 mp_alltoall_d11v 2385 14.1 6.742 6.872 6.742 6.872 calculate_dm_sparse 125 9.5 0.001 0.001 6.634 6.700 grid_collocate_task_list 125 9.7 6.270 6.341 6.270 6.341 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.238 6.290 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.132 6.188 fft3d_ps 1261 14.7 2.782 2.789 6.107 6.174 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=286.722000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2858.000000, yerr=156.460858 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/44ddd7182aebe9205a531925afe90506688fd6a0_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 1.260057E+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 0.013 0.032 85.207 85.208 qs_energies 1 2.0 0.000 0.000 84.708 84.716 ls_scf 1 3.0 0.000 0.000 83.814 83.822 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.726 72.941 multiply_cannon 111 7.7 0.017 0.020 55.848 57.116 multiply_cannon_loop 111 8.7 0.227 0.243 52.447 53.898 ls_scf_main 1 4.0 0.000 0.000 52.460 52.460 density_matrix_trs4 2 5.0 0.002 0.003 46.899 46.973 ls_scf_init_scf 1 4.0 0.000 0.000 28.312 28.313 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.107 27.146 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 24.959 24.983 mp_waitall_1 11031 10.9 22.288 24.934 22.288 24.934 multiply_cannon_multrec 2664 9.7 8.162 8.862 15.641 17.403 multiply_cannon_sync_h2d 2664 9.7 13.608 15.063 13.608 15.063 make_m2s 222 7.7 0.008 0.012 13.162 13.645 make_images 222 8.7 0.099 0.108 13.140 13.625 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.565 12.798 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.356 9.067 make_images_data 222 9.7 0.004 0.005 7.689 8.214 hybrid_alltoall_any 227 10.6 0.216 1.836 6.568 8.131 dbcsr_mm_accdrv_process 4760 10.4 0.590 0.703 7.097 8.116 dbcsr_mm_accdrv_process_sort 4760 11.4 6.303 7.227 6.303 7.227 calculate_norms 4752 9.8 5.523 6.198 5.523 6.198 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.019 5.152 mp_sum_l 887 5.1 3.183 4.706 3.183 4.706 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.388 3.651 multiply_cannon_metrocomm4 2442 9.7 0.012 0.014 2.061 3.548 mp_irecv_dv 6231 10.9 2.045 3.519 2.045 3.519 make_images_sizes 222 9.7 0.000 0.000 0.821 3.454 mp_alltoall_i44 222 10.7 0.821 3.454 0.821 3.454 arnoldi_extremal 4 6.8 0.000 0.000 3.209 3.239 arnoldi_normal_ev 4 7.8 0.001 0.003 3.209 3.239 build_subspace 16 8.4 0.009 0.012 3.113 3.114 ls_scf_post 1 4.0 0.000 0.000 3.042 3.050 ls_scf_store_result 1 5.0 0.000 0.000 2.860 2.900 dbcsr_special_finalize 555 9.7 0.005 0.006 2.297 2.835 dbcsr_merge_single_wm 555 10.7 0.451 0.583 2.289 2.825 make_images_pack 222 9.7 2.207 2.629 2.209 2.630 dbcsr_sort_data 658 11.4 2.095 2.585 2.095 2.585 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.324 2.565 dbcsr_matrix_vector_mult_local 304 10.0 2.067 2.462 2.069 2.464 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.324 2.408 buffer_matrices_ensure_size 222 8.7 1.752 2.085 1.752 2.085 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.838 1.839 rebuild_ks_matrix 3 7.3 0.000 0.000 1.829 1.830 qs_ks_build_kohn_sham_matrix 3 8.3 0.008 0.018 1.829 1.830 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.208000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1131.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/44ddd7182aebe9205a531925afe90506688fd6a0_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 2.191794E+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 0.050 0.087 90.153 90.154 qs_energies 1 2.0 0.004 0.039 89.574 89.618 ls_scf 1 3.0 0.002 0.017 88.216 88.232 dbcsr_multiply_generic 111 6.7 0.016 0.017 74.055 74.436 multiply_cannon 111 7.7 0.026 0.038 52.661 56.932 ls_scf_main 1 4.0 0.001 0.015 54.509 54.511 multiply_cannon_loop 111 8.7 0.135 0.146 49.960 53.207 density_matrix_trs4 2 5.0 0.002 0.004 48.801 49.014 ls_scf_init_scf 1 4.0 0.000 0.002 29.999 30.000 mp_waitall_1 9105 10.9 20.741 29.361 20.741 29.361 ls_scf_init_matrix_S 1 5.0 0.000 0.002 28.664 28.792 multiply_cannon_multrec 1332 9.7 13.292 17.301 22.654 27.800 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.308 26.319 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.583 20.048 make_m2s 222 7.7 0.006 0.008 15.001 15.489 make_images 222 8.7 1.367 1.702 14.970 15.459 dbcsr_mm_accdrv_process 4041 10.4 0.323 0.509 8.959 10.539 dbcsr_mm_accdrv_process_sort 4041 11.4 8.481 10.019 8.481 10.019 make_images_data 222 9.7 0.004 0.005 8.698 9.492 hybrid_alltoall_any 227 10.6 0.541 2.530 8.140 9.353 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.205 7.913 mp_irecv_dv 3311 11.0 3.185 7.882 3.185 7.882 mp_sum_l 887 5.1 4.938 7.878 4.938 7.878 calculate_norms 2376 9.8 6.053 6.798 6.053 6.798 multiply_cannon_sync_h2d 1332 9.7 4.866 6.623 4.866 6.623 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.728 6.223 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.915 5.139 arnoldi_extremal 4 6.8 0.000 0.000 4.771 4.789 arnoldi_normal_ev 4 7.8 0.001 0.005 4.771 4.789 build_subspace 16 8.4 0.014 0.021 4.522 4.524 ls_scf_post 1 4.0 0.002 0.016 3.707 3.722 ls_scf_store_result 1 5.0 0.000 0.000 3.402 3.529 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.190 3.449 dbcsr_matrix_vector_mult_local 304 10.0 2.752 3.237 2.754 3.238 mp_allgather_i34 111 8.7 0.694 2.861 0.694 2.861 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.615 2.708 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.105 2.507 dbcsr_data_new 4174 10.1 2.108 2.406 2.108 2.406 make_images_pack 222 9.7 1.816 2.128 1.819 2.130 dbcsr_sort_data 436 11.2 1.822 2.017 1.822 2.017 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.948 1.951 rebuild_ks_matrix 3 7.3 0.000 0.000 1.935 1.938 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.935 1.938 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=90.154000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1779.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/44ddd7182aebe9205a531925afe90506688fd6a0_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 2.888040E+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 0.024 0.043 91.445 91.446 qs_energies 1 2.0 0.000 0.000 90.882 90.886 ls_scf 1 3.0 0.000 0.000 89.471 89.475 dbcsr_multiply_generic 111 6.7 0.016 0.016 74.124 74.414 ls_scf_main 1 4.0 0.000 0.000 55.944 55.950 multiply_cannon 111 7.7 0.032 0.063 51.463 55.653 multiply_cannon_loop 111 8.7 0.116 0.129 48.729 51.911 density_matrix_trs4 2 5.0 0.002 0.003 50.033 50.192 mp_waitall_1 7281 11.0 22.965 31.779 22.965 31.779 ls_scf_init_scf 1 4.0 0.000 0.000 29.902 29.905 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.724 28.789 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.363 26.376 multiply_cannon_multrec 888 9.7 12.647 15.365 21.297 24.605 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.753 21.422 make_m2s 222 7.7 0.006 0.007 16.309 17.017 make_images 222 8.7 1.581 1.860 16.271 16.979 make_images_data 222 9.7 0.004 0.005 9.506 10.441 hybrid_alltoall_any 227 10.6 0.643 2.957 8.992 10.160 dbcsr_mm_accdrv_process 3754 10.4 0.291 0.490 8.183 9.444 dbcsr_mm_accdrv_process_sort 3754 11.4 7.749 8.954 7.749 8.954 mp_sum_l 887 5.1 4.649 7.721 4.649 7.721 multiply_cannon_sync_h2d 888 9.7 6.074 7.292 6.074 7.292 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.448 7.030 mp_irecv_dv 2335 11.1 2.433 6.968 2.433 6.968 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.349 6.632 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.564 6.033 arnoldi_extremal 4 6.8 0.000 0.000 5.161 5.179 arnoldi_normal_ev 4 7.8 0.001 0.005 5.161 5.179 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.766 4.946 build_subspace 16 8.4 0.014 0.020 4.850 4.856 calculate_norms 1584 9.8 4.344 4.755 4.344 4.755 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.468 3.802 mp_allgather_i34 111 8.7 0.850 3.708 0.850 3.708 ls_scf_post 1 4.0 0.000 0.000 3.625 3.630 dbcsr_matrix_vector_mult_local 304 10.0 3.038 3.612 3.040 3.614 ls_scf_store_result 1 5.0 0.000 0.000 3.364 3.448 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.842 2.962 dbcsr_data_new 4116 9.9 2.103 2.460 2.103 2.460 dbcsr_sort_data 325 11.1 1.902 2.130 1.902 2.130 make_images_sizes 222 9.7 0.000 0.000 0.944 2.039 mp_alltoall_i44 222 10.7 0.944 2.038 0.944 2.038 dbcsr_finalize 304 7.8 0.026 0.032 1.608 1.887 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.884 1.886 make_images_pack 222 9.7 1.627 1.876 1.629 1.879 rebuild_ks_matrix 3 7.3 0.000 0.000 1.865 1.867 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.865 1.867 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=91.446000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2248.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/44ddd7182aebe9205a531925afe90506688fd6a0_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 3.373875E+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 0.049 0.102 97.220 97.221 qs_energies 1 2.0 0.006 0.047 96.490 96.544 ls_scf 1 3.0 0.002 0.018 94.833 94.850 dbcsr_multiply_generic 111 6.7 0.017 0.017 78.170 78.423 ls_scf_main 1 4.0 0.000 0.002 58.549 58.552 multiply_cannon 111 7.7 0.045 0.107 51.660 56.223 density_matrix_trs4 2 5.0 0.002 0.004 52.467 52.574 multiply_cannon_loop 111 8.7 0.152 0.167 46.619 49.234 ls_scf_init_scf 1 4.0 0.000 0.002 33.010 33.011 ls_scf_init_matrix_S 1 5.0 0.000 0.001 31.398 31.449 mp_waitall_1 6369 11.0 22.552 29.378 22.552 29.378 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.924 28.936 multiply_cannon_multrec 1332 9.7 14.138 17.413 22.116 25.033 make_m2s 222 7.7 0.007 0.007 21.054 22.477 make_images 222 8.7 3.134 3.590 21.004 22.429 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.244 16.806 make_images_data 222 9.7 0.004 0.004 11.748 13.372 hybrid_alltoall_any 227 10.6 0.800 3.790 11.098 12.903 dbcsr_mm_accdrv_process 3641 10.4 0.270 0.482 7.623 9.137 dbcsr_mm_accdrv_process_sort 3641 11.4 7.184 8.644 7.184 8.644 mp_sum_l 887 5.1 3.896 7.369 3.896 7.369 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.076 6.161 mp_irecv_dv 3229 10.9 2.051 6.083 2.051 6.083 multiply_cannon_sync_h2d 1332 9.7 5.473 6.002 5.473 6.002 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.965 5.808 arnoldi_extremal 4 6.8 0.000 0.000 5.248 5.266 arnoldi_normal_ev 4 7.8 0.001 0.005 5.248 5.266 build_subspace 16 8.4 0.014 0.021 4.911 4.919 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.393 4.727 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.489 4.639 mp_allgather_i34 111 8.7 2.156 4.553 2.156 4.553 calculate_norms 2376 9.8 4.174 4.534 4.174 4.534 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.609 3.920 dbcsr_matrix_vector_mult_local 304 10.0 3.210 3.710 3.212 3.712 ls_scf_post 1 4.0 0.002 0.019 3.272 3.287 dbcsr_sort_data 658 11.4 3.077 3.278 3.077 3.278 dbcsr_special_finalize 555 9.7 0.006 0.007 2.837 3.108 dbcsr_merge_single_wm 555 10.7 0.541 0.648 2.829 3.100 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.002 3.059 ls_scf_store_result 1 5.0 0.000 0.000 2.969 3.033 dbcsr_data_release 10477 10.7 1.580 2.419 1.580 2.419 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.212 2.214 rebuild_ks_matrix 3 7.3 0.000 0.000 2.189 2.190 qs_ks_build_kohn_sham_matrix 3 8.3 0.027 0.031 2.189 2.190 dbcsr_finalize 304 7.8 0.049 0.061 1.796 1.966 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.221000, 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/44ddd7182aebe9205a531925afe90506688fd6a0_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 4.728996E+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 0.039 0.118 91.768 91.769 qs_energies 1 2.0 0.000 0.000 90.918 90.921 ls_scf 1 3.0 0.000 0.000 88.951 88.953 dbcsr_multiply_generic 111 6.7 0.017 0.019 70.553 70.729 ls_scf_main 1 4.0 0.000 0.000 56.187 56.187 multiply_cannon 111 7.7 0.067 0.166 52.496 55.729 multiply_cannon_loop 111 8.7 0.088 0.094 49.938 51.061 density_matrix_trs4 2 5.0 0.002 0.003 49.251 49.344 ls_scf_init_scf 1 4.0 0.000 0.000 29.334 29.336 mp_waitall_1 5436 11.0 24.096 28.305 24.096 28.305 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.103 28.133 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.949 25.965 multiply_cannon_multrec 444 9.7 13.689 16.779 20.858 22.391 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.723 15.258 make_m2s 222 7.7 0.005 0.005 13.313 14.230 make_images 222 8.7 2.039 2.478 13.245 14.161 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.110 13.607 make_images_data 222 9.7 0.003 0.004 8.080 9.591 hybrid_alltoall_any 227 10.6 0.801 3.822 8.025 9.332 dbcsr_mm_accdrv_process 3003 10.4 0.330 0.452 6.864 7.978 multiply_cannon_sync_h2d 444 9.7 6.745 7.738 6.745 7.738 dbcsr_mm_accdrv_process_sort 3003 11.4 6.488 7.580 6.488 7.580 arnoldi_extremal 4 6.8 0.000 0.000 5.793 5.802 arnoldi_normal_ev 4 7.8 0.002 0.005 5.793 5.802 build_subspace 16 8.4 0.015 0.020 5.398 5.407 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.447 4.680 mp_sum_l 887 5.1 2.799 4.540 2.799 4.540 dbcsr_matrix_vector_mult 304 9.0 0.011 0.022 4.180 4.366 dbcsr_matrix_vector_mult_local 304 10.0 3.725 4.177 3.727 4.179 calculate_norms 792 9.8 3.613 3.720 3.613 3.720 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.952 3.684 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.529 3.560 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.491 3.552 mp_irecv_dv 1241 11.2 1.516 3.537 1.516 3.537 mp_allgather_i34 111 8.7 1.136 3.485 1.136 3.485 ls_scf_post 1 4.0 0.000 0.000 3.429 3.432 ls_scf_store_result 1 5.0 0.000 0.000 3.190 3.225 make_images_sizes 222 9.7 0.000 0.000 0.844 3.211 mp_alltoall_i44 222 10.7 0.843 3.211 0.843 3.211 dbcsr_finalize 304 7.8 0.062 0.077 2.197 2.279 dbcsr_data_new 4608 9.7 1.803 2.192 1.803 2.192 dbcsr_merge_all 275 8.9 0.482 0.538 2.056 2.119 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.007 2.008 rebuild_ks_matrix 3 7.3 0.000 0.000 1.975 1.975 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.975 1.975 qs_energies_init_hamiltonians 1 3.0 0.004 0.014 1.951 1.952 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=91.769000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3723.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/44ddd7182aebe9205a531925afe90506688fd6a0_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 8.836866E+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 0.077 0.097 109.200 109.200 qs_energies 1 2.0 0.000 0.000 107.541 107.546 ls_scf 1 3.0 0.000 0.000 104.556 104.560 dbcsr_multiply_generic 111 6.7 0.024 0.028 78.171 78.306 ls_scf_main 1 4.0 0.000 0.000 63.787 63.788 density_matrix_trs4 2 5.0 0.002 0.003 54.843 54.895 multiply_cannon 111 7.7 0.149 0.264 48.726 51.116 multiply_cannon_loop 111 8.7 0.098 0.099 45.739 49.228 ls_scf_init_scf 1 4.0 0.000 0.000 36.988 36.990 ls_scf_init_matrix_S 1 5.0 0.000 0.000 35.486 35.507 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 32.786 32.793 mp_waitall_1 4527 11.1 21.778 25.468 21.778 25.468 make_m2s 222 7.7 0.005 0.005 22.768 23.857 make_images 222 8.7 3.578 3.868 22.661 23.748 multiply_cannon_multrec 444 9.7 17.855 18.562 22.546 23.284 hybrid_alltoall_any 227 10.6 1.659 3.638 12.900 15.943 make_images_data 222 9.7 0.003 0.004 13.163 15.319 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.049 10.484 multiply_cannon_sync_h2d 444 9.7 8.787 8.839 8.787 8.839 arnoldi_extremal 4 6.8 0.000 0.000 7.499 7.507 arnoldi_normal_ev 4 7.8 0.003 0.009 7.498 7.506 build_subspace 16 8.4 0.026 0.036 6.946 6.958 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.565 5.712 dbcsr_matrix_vector_mult_local 304 10.0 5.105 5.413 5.107 5.416 mp_sum_l 887 5.1 4.310 5.374 4.310 5.374 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.078 5.171 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.850 5.106 dbcsr_mm_accdrv_process 1814 10.4 0.289 0.356 4.506 4.637 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.674 4.357 dbcsr_mm_accdrv_process_sort 1814 11.4 4.170 4.306 4.170 4.306 compute_matrix_preconditioner 1 6.0 0.002 0.002 4.148 4.154 ls_scf_post 1 4.0 0.000 0.000 3.780 3.786 acc_transpose_blocks 444 9.7 0.003 0.003 0.659 3.635 make_images_sizes 222 9.7 0.000 0.000 1.465 3.607 mp_alltoall_i44 222 10.7 1.464 3.606 1.464 3.606 acc_transpose_blocks_kernels 444 10.7 0.006 0.006 0.577 3.552 jit_kernel_transpose 1 13.0 0.571 3.546 0.571 3.546 ls_scf_store_result 1 5.0 0.000 0.000 3.502 3.538 mp_allgather_i34 111 8.7 1.095 3.478 1.095 3.478 calculate_norms 792 9.8 3.234 3.271 3.234 3.271 dbcsr_finalize 304 7.8 0.082 0.089 3.086 3.167 qs_energies_init_hamiltonians 1 3.0 0.025 0.029 2.952 2.953 dbcsr_merge_all 275 8.9 0.890 0.924 2.871 2.946 dbcsr_complete_redistribute 5 7.6 1.426 1.466 2.743 2.850 matrix_ls_to_qs 2 6.0 0.000 0.000 2.401 2.515 dbcsr_sort_data 325 11.1 2.443 2.488 2.443 2.488 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.385 2.387 dbcsr_data_new 6591 9.6 1.844 2.344 1.844 2.344 rebuild_ks_matrix 3 7.3 0.000 0.000 2.318 2.320 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 2.318 2.320 dbcsr_new_transposed 4 7.5 0.242 0.263 2.281 2.288 dbcsr_frobenius_norm 74 6.6 2.054 2.135 2.212 2.245 dbcsr_add_d 103 6.2 0.000 0.000 2.131 2.208 dbcsr_add_anytype 103 7.2 0.859 0.891 2.130 2.207 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=109.200000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6953.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/44ddd7182aebe9205a531925afe90506688fd6a0_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 588.886016E+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 4809148. 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 0.034 0.210 228.458 228.464 qs_mol_dyn_low 1 2.0 0.006 0.059 226.794 227.051 qs_forces 5 3.8 0.017 0.045 226.547 226.550 qs_energies 5 4.8 0.004 0.076 223.376 223.472 scf_env_do_scf 5 5.8 0.000 0.002 208.237 208.240 scf_env_do_scf_inner_loop 105 6.6 0.004 0.028 181.147 181.166 qs_scf_new_mos 105 7.6 0.000 0.001 141.234 141.407 qs_scf_loop_do_ot 105 8.6 0.001 0.001 141.234 141.406 dbcsr_multiply_generic 1445 12.2 0.127 0.137 133.405 133.771 ot_scf_mini 105 9.6 0.005 0.146 131.435 131.612 multiply_cannon 1445 13.2 0.275 0.286 114.271 116.087 multiply_cannon_loop 1445 14.2 2.848 2.996 112.603 113.607 velocity_verlet 4 3.0 0.002 0.035 107.326 107.326 ot_mini 105 10.6 0.001 0.032 60.085 60.234 multiply_cannon_multrec 69360 15.2 29.740 34.754 39.880 44.980 qs_ot_get_p 112 10.4 0.001 0.001 41.133 41.356 mp_waitall_1 488190 16.1 34.483 41.014 34.483 41.014 qs_ot_get_derivative 55 11.6 0.002 0.043 38.369 38.534 multiply_cannon_sync_h2d 69360 15.2 29.131 33.188 29.131 33.188 multiply_cannon_metrocomm3 69360 15.2 0.202 0.214 25.538 32.826 qs_ot_p2m_diag 40 11.0 0.020 0.032 30.166 30.357 rebuild_ks_matrix 110 8.4 0.000 0.000 29.086 29.258 qs_ks_build_kohn_sham_matrix 110 9.4 0.011 0.013 29.086 29.258 init_scf_loop 7 6.6 0.000 0.011 27.059 27.060 cp_dbcsr_syevd 40 12.0 0.002 0.002 26.949 26.950 qs_ks_update_qs_env 112 7.6 0.001 0.001 26.717 26.879 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 22.965 23.217 apply_single 62 13.6 0.000 0.000 22.965 23.217 prepare_preconditioner 7 7.6 0.000 0.000 22.112 22.143 make_preconditioner 7 8.6 0.000 0.000 22.112 22.143 cp_fm_syevd 40 13.0 0.000 0.000 21.839 21.963 ot_new_cg_direction 55 11.6 0.001 0.004 21.024 21.026 qs_rho_update_rho_low 110 7.6 0.001 0.034 16.781 17.163 calculate_rho_elec 110 8.6 0.030 0.040 16.779 17.129 cp_fm_redistribute_end 40 14.0 8.546 17.048 8.551 17.049 cp_fm_syevd_base 40 14.0 8.491 16.996 8.491 16.996 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.991 15.155 make_full_inverse_cholesky 7 9.6 0.000 0.000 14.816 14.887 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 14.051 14.151 mp_sum_l 4764 12.2 12.486 13.158 12.486 13.158 init_scf_run 5 5.8 0.000 0.001 12.309 12.310 scf_env_initial_rho_setup 5 6.8 0.002 0.020 12.309 12.310 pw_transfer 1645 12.4 0.080 0.095 11.972 12.155 fft_wrap_pw1pw2 1425 13.5 0.012 0.015 11.833 12.019 calculate_dm_sparse 110 9.5 0.000 0.001 11.499 11.691 density_rs2pw 110 9.6 0.005 0.007 10.996 11.576 dbcsr_mm_accdrv_process 154766 15.8 6.222 6.439 10.007 10.826 qs_vxc_create 110 10.4 0.003 0.006 10.618 10.653 fft_wrap_pw1pw2_240 915 15.0 1.149 1.222 10.401 10.560 cp_fm_cholesky_invert 7 10.6 10.408 10.416 10.408 10.416 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 10.129 10.203 check_diag 80 13.5 8.704 8.935 9.541 9.664 fft3d_pb 915 16.0 2.386 2.591 8.600 8.773 sum_up_and_integrate 60 10.3 0.001 0.002 8.459 8.470 integrate_v_rspace 60 11.3 0.002 0.002 8.442 8.453 transfer_rs2pw 445 10.6 0.007 0.008 7.673 8.251 acc_transpose_blocks 69360 15.2 0.350 0.365 7.524 8.077 xc_rho_set_and_dset_create 110 12.4 0.077 0.099 7.446 7.700 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.003 7.443 7.460 calculate_first_density_matrix 1 7.0 0.001 0.023 7.308 7.323 xc_vxc_pw_create 60 11.3 0.039 0.050 7.121 7.155 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 7.015 7.067 make_full_single_inverse 7 9.6 0.001 0.001 7.014 7.047 make_m2s 2890 13.2 0.079 0.088 6.393 6.953 multiply_cannon_metrocomm1 69360 15.2 0.096 0.104 4.594 6.927 make_images 2890 14.2 0.239 0.261 6.286 6.848 xc_pw_derive 510 13.4 0.005 0.006 6.390 6.447 mp_alltoall_z22v 2340 17.7 5.621 5.921 5.621 5.921 acc_transpose_blocks_kernels 69360 16.2 0.852 0.889 4.771 5.222 mp_waitany 7680 13.5 4.378 4.994 4.378 4.994 multiply_cannon_metrocomm4 67915 15.2 0.185 0.201 2.019 4.720 potential_pw2rs 60 12.3 0.003 0.003 4.657 4.688 wfi_extrapolate 5 7.8 0.000 0.001 4.674 4.674 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=228.464000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=560.600000, yerr=1.959592 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 44ddd7182aebe9205a531925afe90506688fd6a0 Summary: empty Status: OK