=== 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: c078969bf945a475acf73b54de8647dac0b903ea ################# 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, # LIBGRPP 20231215 # # 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: 15.12.2023 # # \ 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=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2023.05.001 USE_HDF5 := 1.14.2 USE_LIBGRPP := 20231215 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.9.0 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(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_LIBGRPP),) USE_LIBGRPP := $(strip $(USE_LIBGRPP)) LIBGRPP_INC := $(INSTALL_PATH)/libgrpp-main-$(USE_LIBGRPP)/include LIBGRPP_LIB := $(INSTALL_PATH)/libgrpp-main-$(USE_LIBGRPP)/lib CFLAGS += -I$(LIBGRPP_INC) DFLAGS += -D__LIBGRPP LIBS += $(LIBGRPP_LIB)/liblibgrpp.a endif ifneq ($(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/c078969bf945a475acf73b54de8647dac0b903ea_performance_tests/01 job id: 50693173 --- 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/c078969bf945a475acf73b54de8647dac0b903ea_performance_tests/02 job id: 50693175 --- 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/c078969bf945a475acf73b54de8647dac0b903ea_performance_tests/03 job id: 50693177 --- 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/c078969bf945a475acf73b54de8647dac0b903ea_performance_tests/04 job id: 50693179 --- 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/c078969bf945a475acf73b54de8647dac0b903ea_performance_tests/05 job id: 50693181 --- 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/c078969bf945a475acf73b54de8647dac0b903ea_performance_tests/06 job id: 50693183 --- 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/c078969bf945a475acf73b54de8647dac0b903ea_performance_tests/07 job id: 50693185 --- 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/c078969bf945a475acf73b54de8647dac0b903ea_performance_tests/08 job id: 50693187 --- 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/c078969bf945a475acf73b54de8647dac0b903ea_performance_tests/09 job id: 50693189 --- 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/c078969bf945a475acf73b54de8647dac0b903ea_performance_tests/10 job id: 50693190 --- 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/c078969bf945a475acf73b54de8647dac0b903ea_performance_tests/11 job id: 50693193 --- 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/c078969bf945a475acf73b54de8647dac0b903ea_performance_tests/12 job id: 50693194 --- 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/c078969bf945a475acf73b54de8647dac0b903ea_performance_tests/13 job id: 50693196 --- 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/c078969bf945a475acf73b54de8647dac0b903ea_performance_tests/14 job id: 50693198 --- 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/c078969bf945a475acf73b54de8647dac0b903ea_performance_tests/15 job id: 50693200 --- 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/c078969bf945a475acf73b54de8647dac0b903ea_performance_tests/16 job id: 50693203 --- 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/c078969bf945a475acf73b54de8647dac0b903ea_performance_tests/17 job id: 50693206 --- 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/c078969bf945a475acf73b54de8647dac0b903ea_performance_tests/18 job id: 50693208 --- 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/c078969bf945a475acf73b54de8647dac0b903ea_performance_tests/19 job id: 50693210 --- 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/c078969bf945a475acf73b54de8647dac0b903ea_performance_tests/20 job id: 50693212 --- 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/c078969bf945a475acf73b54de8647dac0b903ea_performance_tests/21 job id: 50693213 --- 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/c078969bf945a475acf73b54de8647dac0b903ea_performance_tests/22 job id: 50693218 --- 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/c078969bf945a475acf73b54de8647dac0b903ea_performance_tests/23 job id: 50693221 --- 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/c078969bf945a475acf73b54de8647dac0b903ea_performance_tests/24 job id: 50693223 --- 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/c078969bf945a475acf73b54de8647dac0b903ea_performance_tests/25 job id: 50693225 --- 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/c078969bf945a475acf73b54de8647dac0b903ea_performance_tests/26 job id: 50693228 --- 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/c078969bf945a475acf73b54de8647dac0b903ea_performance_tests/27 job id: 50693230 --- 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/c078969bf945a475acf73b54de8647dac0b903ea_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 172669. MP_Allreduce 424 8. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.046 134.067 134.068 farming_run 1 2.0 133.410 133.412 134.025 134.028 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.492369E+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 0.005 0.024 114.915 114.915 qs_energies 1 2.0 0.000 0.000 114.697 114.700 mp2_main 1 3.0 0.000 0.000 112.619 112.622 mp2_gpw_main 1 4.0 0.027 0.034 111.603 111.607 mp2_ri_gpw_compute_in 1 5.0 0.171 0.172 92.903 93.022 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 54.565 54.683 mp2_eri_3c_integrate_gpw 272 7.0 0.151 0.162 40.968 45.720 get_2c_integrals 1 6.0 0.008 0.009 37.045 38.166 integrate_v_rspace 273 8.0 0.441 0.455 24.558 29.006 pw_transfer 6555 10.6 0.372 0.394 26.876 27.329 fft_wrap_pw1pw2 5465 11.4 0.044 0.047 25.459 25.750 grid_integrate_task_list 273 9.0 20.473 25.336 20.473 25.336 fft_wrap_pw1pw2_100 2178 12.4 0.030 0.034 22.989 23.296 compute_2c_integrals 1 7.0 0.002 0.002 19.414 19.415 compute_2c_integrals_loop_lm 1 8.0 0.001 0.003 18.717 19.146 mp2_eri_2c_integrate_gpw 1 9.0 2.360 2.417 18.716 19.145 rpa_ri_compute_en 1 5.0 0.019 0.023 18.596 18.721 cp_fm_cholesky_decompose 12 8.2 17.566 18.717 17.566 18.717 cholesky_decomp 1 7.0 0.000 0.000 16.469 17.622 fft3d_s 5443 13.4 16.136 16.345 16.158 16.365 ao_to_mo_and_store_B_mult_1 272 7.0 10.773 15.327 10.773 15.327 calculate_wavefunction 272 8.0 5.407 5.484 12.320 13.075 rpa_num_int 1 6.0 0.001 0.009 10.676 10.676 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.643 10.654 calc_mat_Q 8 8.0 0.000 0.000 9.503 9.583 contract_S_to_Q 8 9.0 0.000 0.000 8.927 9.008 calc_potential_gpw 544 9.5 0.004 0.005 8.250 8.769 parallel_gemm_fm 14 9.1 0.000 0.000 8.507 8.628 parallel_gemm_fm_cosma 14 10.1 8.507 8.628 8.507 8.628 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 8.149 8.446 potential_pw2rs 545 10.0 0.106 0.107 7.530 8.130 create_integ_mat 1 6.0 0.031 0.051 7.706 7.707 collocate_single_gaussian 272 10.0 0.038 0.041 7.402 7.633 array2fm 1 7.0 0.000 0.000 6.689 7.131 pw_scatter_s 2720 13.7 4.315 4.476 4.315 4.476 pw_gather_s 2722 13.2 3.441 3.783 3.441 3.783 array2fm_buffer_send 1 8.0 3.014 3.181 3.014 3.181 pw_poisson_solve 545 10.5 1.123 1.192 2.251 2.491 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.606374, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2809.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c078969bf945a475acf73b54de8647dac0b903ea_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 200775. 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 0.063 0.069 421.867 421.870 farming_run 1 2.0 420.826 420.832 421.781 421.791 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827141120 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788822 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.244484E+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 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 0.012 0.031 217.333 217.334 qs_energies 1 2.0 0.002 0.004 217.081 217.105 scf_env_do_scf 1 3.0 0.000 0.000 114.623 114.624 qs_ks_update_qs_env 5 5.0 0.000 0.000 113.724 113.733 rebuild_ks_matrix 4 6.0 0.000 0.000 113.723 113.732 qs_ks_build_kohn_sham_matrix 4 7.0 0.054 0.063 113.723 113.732 hfx_ks_matrix 4 8.0 0.001 0.001 113.375 113.380 integrate_four_center 4 9.0 0.152 0.467 113.375 113.379 integrate_four_center_main 4 10.0 0.109 0.559 101.872 104.722 integrate_four_center_bin 266 11.0 101.763 104.505 101.763 104.505 mp2_main 1 3.0 0.025 0.089 102.160 102.181 mp2_gpw_main 1 4.0 0.064 0.142 101.232 101.253 init_scf_loop 1 4.0 0.000 0.000 96.805 96.805 mp2_ri_gpw_compute_in 1 5.0 0.076 0.106 74.113 75.280 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.012 53.799 54.964 mp2_eri_3c_integrate_gpw 91 7.0 0.145 0.161 41.666 46.748 integrate_v_rspace 95 8.0 0.395 0.559 28.135 33.073 pw_transfer 2240 10.6 0.145 0.162 29.516 29.893 fft_wrap_pw1pw2 1868 11.4 0.018 0.020 28.526 28.917 grid_integrate_task_list 95 9.0 23.554 28.672 23.554 28.672 mp2_ri_gpw_compute_en 1 5.0 0.067 0.109 26.872 28.657 fft_wrap_pw1pw2_100 730 12.4 0.013 0.014 26.275 26.677 ao_to_mo_and_store_B_mult_1 91 7.0 10.464 25.887 10.464 25.887 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.839 1.908 24.971 24.979 get_2c_integrals 1 6.0 0.019 0.068 20.185 20.252 compute_2c_integrals 1 7.0 0.004 0.006 19.145 19.163 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 18.631 18.993 mp2_eri_2c_integrate_gpw 1 9.0 1.728 1.823 18.630 18.992 fft3d_s 1823 13.4 18.491 18.864 18.505 18.877 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 17.817 17.817 calculate_wavefunction 91 8.0 2.015 2.049 9.625 9.797 mp2_ri_gpw_compute_en_expansio 172 7.0 0.559 0.585 8.795 9.273 potential_pw2rs 186 10.0 0.033 0.035 8.406 8.892 local_gemm 172 8.0 8.236 8.699 8.236 8.699 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.164 8.571 mp2_ri_gpw_compute_en_comm 22 7.0 0.500 0.517 7.946 8.471 calc_potential_gpw 182 9.5 0.002 0.002 7.930 8.202 collocate_single_gaussian 91 10.0 0.021 0.062 7.762 7.985 mp_sync 37 10.5 3.489 6.957 3.489 6.957 integrate_four_center_load 4 10.0 0.000 0.000 6.770 6.775 hfx_load_balance 1 11.0 0.000 0.000 6.770 6.775 mp_sendrecv_dm3 2068 8.0 5.975 6.482 5.975 6.482 mp2_ri_gpw_compute_en_ener 172 7.0 6.346 6.442 6.346 6.442 pw_gather_s 912 13.2 4.477 4.941 4.477 4.941 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.250973, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1503.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c078969bf945a475acf73b54de8647dac0b903ea_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 453.386240E+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 592243. 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 0.072 0.114 54.928 54.931 qs_mol_dyn_low 1 2.0 0.003 0.004 54.454 54.462 qs_forces 11 3.9 0.072 0.094 54.389 54.390 qs_energies 11 4.9 0.039 0.089 52.786 52.813 scf_env_do_scf 11 5.9 0.001 0.004 46.154 46.159 scf_env_do_scf_inner_loop 108 6.5 0.002 0.008 43.975 43.975 dbcsr_multiply_generic 2286 12.5 0.094 0.098 33.674 34.293 qs_scf_new_mos 108 7.5 0.000 0.001 33.941 34.241 qs_scf_loop_do_ot 108 8.5 0.000 0.001 33.941 34.241 ot_scf_mini 108 9.5 0.002 0.002 32.218 32.425 multiply_cannon 2286 13.5 0.187 0.194 26.242 27.803 multiply_cannon_loop 2286 14.5 1.822 1.932 25.585 27.182 velocity_verlet 10 3.0 0.001 0.002 26.908 26.909 ot_mini 108 10.5 0.001 0.001 19.395 19.641 qs_ot_get_derivative 108 11.5 0.001 0.001 16.409 16.598 mp_waitall_1 245248 16.5 8.533 14.779 8.533 14.779 multiply_cannon_metrocomm3 54864 15.5 0.074 0.078 6.106 12.903 multiply_cannon_multrec 54864 15.5 3.609 5.690 7.576 10.906 qs_ot_get_p 119 10.4 0.001 0.001 8.151 8.443 rebuild_ks_matrix 119 8.3 0.000 0.000 7.864 8.051 qs_ks_build_kohn_sham_matrix 119 9.3 0.017 0.047 7.864 8.050 mp_sum_l 7287 12.8 5.519 7.235 5.519 7.235 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.917 7.073 multiply_cannon_sync_h2d 54864 15.5 5.155 6.656 5.155 6.656 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.962 6.450 dbcsr_mm_accdrv_process 76910 16.1 1.859 2.921 3.881 5.583 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.256 5.290 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.167 5.265 init_scf_run 11 5.9 0.000 0.001 5.107 5.107 scf_env_initial_rho_setup 11 6.9 0.020 0.061 5.107 5.107 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.459 4.459 sum_up_and_integrate 119 10.3 0.001 0.003 4.422 4.427 integrate_v_rspace 119 11.3 0.002 0.004 4.411 4.418 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.165 4.372 calculate_rho_elec 119 8.7 0.011 0.017 4.165 4.371 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.257 4.257 cp_fm_redistribute_end 50 14.0 2.172 4.233 2.178 4.235 cp_fm_diag_elpa_base 50 14.0 2.052 4.136 2.056 4.145 multiply_cannon_metrocomm1 54864 15.5 0.057 0.063 1.786 3.156 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.935 3.144 apply_single 119 13.6 0.000 0.000 2.935 3.144 calculate_dm_sparse 119 9.5 0.000 0.000 2.944 3.094 calculate_first_density_matrix 1 7.0 0.001 0.004 2.831 2.859 acc_transpose_blocks 54864 15.5 0.226 0.251 2.218 2.749 ot_diis_step 108 11.5 0.006 0.006 2.733 2.733 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.499 2.573 jit_kernel_multiply 13 15.8 1.953 2.566 1.953 2.566 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.270 2.272 density_rs2pw 119 9.7 0.004 0.004 2.130 2.217 wfi_extrapolate 11 7.9 0.001 0.001 2.180 2.180 init_scf_loop 11 6.9 0.005 0.030 2.163 2.168 grid_integrate_task_list 119 12.3 2.008 2.120 2.008 2.120 mp_sum_d 4137 12.0 1.453 2.024 1.453 2.024 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.942 1.994 potential_pw2rs 119 12.3 0.004 0.004 1.804 1.816 pw_transfer 1439 11.6 0.051 0.057 1.649 1.719 make_m2s 4572 13.5 0.053 0.055 1.619 1.665 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.574 1.647 make_images 4572 14.5 0.132 0.138 1.537 1.583 mp_alltoall_d11v 2130 13.8 1.353 1.577 1.353 1.577 transfer_rs2pw 487 10.6 0.005 0.006 1.449 1.571 mp_waitany 12084 13.8 1.351 1.535 1.351 1.535 acc_transpose_blocks_sync 164592 16.5 1.209 1.454 1.209 1.454 grid_collocate_task_list 119 9.7 1.355 1.432 1.355 1.432 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.392 1.423 fft3d_ps 1201 14.6 0.367 0.470 1.284 1.354 transfer_pw2rs 487 13.2 0.006 0.006 1.320 1.330 parallel_gemm_fm 81 9.0 0.000 0.000 1.305 1.310 parallel_gemm_fm_cosma 81 10.0 1.305 1.310 1.305 1.310 fft_wrap_pw1pw2_140 487 13.2 0.109 0.117 1.218 1.290 dbcsr_dot_sd 1205 11.9 0.049 0.060 0.769 1.155 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=54.931000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=432.272727, yerr=0.962091 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c078969bf945a475acf73b54de8647dac0b903ea_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 490.098688E+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 2442191. 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 0.018 0.053 43.980 43.982 qs_mol_dyn_low 1 2.0 0.003 0.004 43.666 43.672 qs_forces 11 3.9 0.002 0.003 43.325 43.326 qs_energies 11 4.9 0.005 0.023 41.496 41.501 scf_env_do_scf 11 5.9 0.002 0.010 35.901 35.903 scf_env_do_scf_inner_loop 108 6.5 0.013 0.027 32.889 32.890 dbcsr_multiply_generic 2286 12.5 0.102 0.105 23.054 23.388 qs_scf_new_mos 108 7.5 0.001 0.001 22.562 22.809 qs_scf_loop_do_ot 108 8.5 0.001 0.001 22.561 22.808 ot_scf_mini 108 9.5 0.003 0.004 21.602 21.776 velocity_verlet 10 3.0 0.002 0.008 20.971 20.978 multiply_cannon 2286 13.5 0.209 0.217 17.105 18.605 multiply_cannon_loop 2286 14.5 1.180 1.242 15.758 17.257 ot_mini 108 10.5 0.001 0.001 12.937 13.177 mp_waitall_1 200699 16.5 6.473 11.388 6.473 11.388 qs_ot_get_derivative 108 11.5 0.001 0.002 10.514 10.689 multiply_cannon_metrocomm3 27432 15.5 0.072 0.075 4.087 9.343 multiply_cannon_multrec 27432 15.5 1.808 4.111 6.132 8.883 rebuild_ks_matrix 119 8.3 0.000 0.000 7.814 7.959 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.019 7.813 7.959 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.871 7.006 dbcsr_mm_accdrv_process 47894 16.0 3.592 5.629 4.244 6.318 qs_ot_get_p 119 10.4 0.001 0.002 5.582 5.789 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.970 4.833 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.631 4.668 calculate_rho_elec 119 8.7 0.021 0.024 4.630 4.668 sum_up_and_integrate 119 10.3 0.001 0.002 4.613 4.625 integrate_v_rspace 119 11.3 0.002 0.003 4.599 4.610 mp_sum_l 7287 12.8 2.232 4.327 2.232 4.327 init_scf_run 11 5.9 0.000 0.001 4.247 4.248 scf_env_initial_rho_setup 11 6.9 0.002 0.003 4.247 4.248 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.056 4.092 apply_single 119 13.6 0.000 0.000 3.056 4.092 qs_ot_p2m_diag 50 11.0 0.015 0.063 3.627 3.686 make_m2s 4572 13.5 0.052 0.053 3.279 3.597 make_images 4572 14.5 0.206 0.244 3.189 3.510 cp_dbcsr_syevd 50 12.0 0.006 0.027 3.190 3.190 init_scf_loop 11 6.9 0.001 0.006 2.976 2.980 density_rs2pw 119 9.7 0.004 0.004 2.891 2.978 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.691 2.692 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.584 2.670 cp_fm_redistribute_end 50 14.0 1.356 2.642 1.367 2.647 cp_fm_diag_elpa_base 50 14.0 1.238 2.510 1.271 2.560 calculate_first_density_matrix 1 7.0 0.000 0.003 2.393 2.396 pw_transfer 1439 11.6 0.065 0.068 2.321 2.379 ot_diis_step 108 11.5 0.011 0.011 2.372 2.373 potential_pw2rs 119 12.3 0.006 0.006 2.300 2.338 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.230 2.289 calculate_dm_sparse 119 9.5 0.000 0.001 2.114 2.198 multiply_cannon_sync_h2d 27432 15.5 1.685 2.189 1.685 2.189 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.981 1.984 acc_transpose_blocks 27432 15.5 0.115 0.119 1.610 1.951 transfer_rs2pw 487 10.6 0.005 0.005 1.838 1.950 make_images_data 4572 15.5 0.047 0.053 1.432 1.947 prepare_preconditioner 11 7.9 0.000 0.000 1.915 1.943 make_preconditioner 11 8.9 0.000 0.001 1.915 1.943 fft3d_ps 1201 14.6 0.525 0.585 1.861 1.916 grid_integrate_task_list 119 12.3 1.825 1.913 1.825 1.913 make_full_inverse_cholesky 11 9.9 0.000 0.001 1.798 1.855 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.704 1.747 wfi_extrapolate 11 7.9 0.001 0.001 1.732 1.732 mp_alltoall_d11v 2130 13.8 1.523 1.694 1.523 1.694 fft_wrap_pw1pw2_140 487 13.2 0.127 0.133 1.603 1.664 hybrid_alltoall_any 4725 16.4 0.053 0.115 1.252 1.640 transfer_pw2rs 487 13.2 0.004 0.005 1.565 1.595 grid_collocate_task_list 119 9.7 1.296 1.375 1.296 1.375 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.321 1.372 jit_kernel_multiply 7 16.4 0.592 1.358 0.592 1.358 make_images_sizes 4572 15.5 0.005 0.005 0.964 1.325 mp_alltoall_i44 4572 16.5 0.960 1.320 0.960 1.320 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.286 1.297 mp_allgather_i34 2286 14.5 0.767 1.235 0.767 1.235 mp_sum_d 4137 12.0 0.637 1.123 0.637 1.123 mp_alltoall_z22v 1201 16.6 1.012 1.121 1.012 1.121 mp_sendrecv_dv 16779 12.7 0.996 1.023 0.996 1.023 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.009 1.023 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 0.991 0.992 acc_transpose_blocks_sync 82296 16.5 0.810 0.933 0.810 0.933 cp_fm_cholesky_invert 11 10.9 0.885 0.891 0.885 0.891 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=43.982000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=465.909091, yerr=2.193152 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c078969bf945a475acf73b54de8647dac0b903ea_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 523.247616E+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 0.031 0.175 32.641 32.642 qs_mol_dyn_low 1 2.0 0.004 0.015 32.183 32.202 qs_forces 11 3.9 0.006 0.016 32.119 32.121 qs_energies 11 4.9 0.003 0.018 30.519 30.524 scf_env_do_scf 11 5.9 0.002 0.015 25.431 25.431 scf_env_do_scf_inner_loop 108 6.5 0.012 0.020 22.853 22.854 dbcsr_multiply_generic 2286 12.5 0.096 0.100 16.792 16.849 qs_scf_new_mos 108 7.5 0.001 0.001 15.274 15.296 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.273 15.295 velocity_verlet 10 3.0 0.001 0.002 14.949 14.950 ot_scf_mini 108 9.5 0.002 0.002 14.518 14.541 multiply_cannon 2286 13.5 0.196 0.199 13.575 14.343 multiply_cannon_loop 2286 14.5 0.858 0.898 12.812 13.593 ot_mini 108 10.5 0.002 0.014 9.005 9.027 qs_ot_get_derivative 108 11.5 0.001 0.001 7.524 7.547 multiply_cannon_multrec 18288 15.5 1.872 2.848 6.933 7.236 rebuild_ks_matrix 119 8.3 0.000 0.000 6.064 6.091 qs_ks_build_kohn_sham_matrix 119 9.3 0.016 0.045 6.063 6.091 dbcsr_mm_accdrv_process 38222 16.0 4.904 5.657 4.966 5.726 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.343 5.368 init_scf_run 11 5.9 0.000 0.001 3.582 3.583 scf_env_initial_rho_setup 11 6.9 0.006 0.041 3.581 3.583 sum_up_and_integrate 119 10.3 0.002 0.006 3.577 3.582 integrate_v_rspace 119 11.3 0.003 0.014 3.564 3.570 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.785 3.439 qs_ot_get_p 119 10.4 0.001 0.001 3.294 3.315 mp_waitall_1 158411 16.6 2.488 3.308 2.488 3.308 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.202 3.205 calculate_rho_elec 119 8.7 0.031 0.031 3.201 3.205 init_scf_loop 11 6.9 0.001 0.005 2.548 2.548 calculate_first_density_matrix 1 7.0 0.001 0.010 2.364 2.372 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.041 2.345 apply_single 119 13.6 0.000 0.000 2.041 2.345 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.208 2.216 multiply_cannon_metrocomm3 18288 15.5 0.049 0.050 1.376 2.015 grid_integrate_task_list 119 12.3 1.819 1.927 1.819 1.927 cp_dbcsr_syevd 50 12.0 0.003 0.004 1.911 1.912 make_m2s 4572 13.5 0.044 0.045 1.748 1.889 calculate_dm_sparse 119 9.5 0.000 0.001 1.842 1.850 density_rs2pw 119 9.7 0.004 0.004 1.770 1.842 pw_transfer 1439 11.6 0.065 0.068 1.825 1.832 make_images 4572 14.5 0.193 0.206 1.663 1.802 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.732 1.740 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.695 1.700 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.696 1.700 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.672 1.684 cp_fm_diag_elpa_base 50 14.0 1.648 1.662 1.669 1.681 acc_transpose_blocks 18288 15.5 0.079 0.081 1.623 1.676 prepare_preconditioner 11 7.9 0.000 0.000 1.661 1.666 make_preconditioner 11 8.9 0.000 0.000 1.661 1.666 make_full_inverse_cholesky 11 9.9 0.000 0.001 1.518 1.607 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.513 1.520 ot_diis_step 108 11.5 0.011 0.011 1.464 1.464 mp_sum_l 7287 12.8 1.082 1.440 1.082 1.440 fft_wrap_pw1pw2_140 487 13.2 0.179 0.183 1.412 1.420 potential_pw2rs 119 12.3 0.007 0.008 1.299 1.300 fft3d_ps 1201 14.6 0.546 0.565 1.287 1.294 grid_collocate_task_list 119 9.7 1.234 1.292 1.234 1.292 qs_energies_init_hamiltonians 11 5.9 0.027 0.048 1.272 1.276 transfer_rs2pw 487 10.6 0.005 0.005 1.159 1.261 multiply_cannon_sync_h2d 18288 15.5 1.068 1.192 1.068 1.192 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.182 1.187 wfi_extrapolate 11 7.9 0.001 0.001 1.159 1.159 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.934 0.955 mp_alltoall_d11v 2130 13.8 0.709 0.907 0.709 0.907 make_images_data 4572 15.5 0.047 0.051 0.740 0.897 mp_waitany 9880 13.7 0.749 0.854 0.749 0.854 transfer_rs2pw_140 130 11.5 0.121 0.159 0.735 0.841 hybrid_alltoall_any 4725 16.4 0.058 0.114 0.639 0.827 qs_env_update_s_mstruct 11 6.9 0.000 0.001 0.763 0.827 acc_transpose_blocks_kernels 18288 16.5 0.219 0.227 0.789 0.797 transfer_pw2rs 487 13.2 0.004 0.004 0.792 0.796 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.792 0.793 acc_transpose_blocks_sync 54864 16.5 0.735 0.781 0.735 0.781 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.003 0.675 0.749 cp_fm_cholesky_invert 11 10.9 0.688 0.691 0.688 0.691 mp_alltoall_z22v 1201 16.6 0.594 0.662 0.594 0.662 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=32.642000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=498.363636, yerr=1.966664 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c078969bf945a475acf73b54de8647dac0b903ea_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 559.800320E+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 0.014 0.034 36.983 36.983 qs_mol_dyn_low 1 2.0 0.003 0.004 36.713 36.721 qs_forces 11 3.9 0.002 0.002 36.654 36.656 qs_energies 11 4.9 0.001 0.001 34.933 34.939 scf_env_do_scf 11 5.9 0.000 0.001 29.876 29.877 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 26.336 26.337 dbcsr_multiply_generic 2286 12.5 0.102 0.105 20.002 20.204 velocity_verlet 10 3.0 0.001 0.002 18.532 18.534 qs_scf_new_mos 108 7.5 0.001 0.001 18.133 18.197 qs_scf_loop_do_ot 108 8.5 0.001 0.001 18.133 18.196 ot_scf_mini 108 9.5 0.002 0.003 17.101 17.155 multiply_cannon 2286 13.5 0.220 0.227 16.236 16.719 multiply_cannon_loop 2286 14.5 1.500 1.602 15.271 15.693 ot_mini 108 10.5 0.001 0.001 10.533 10.599 multiply_cannon_multrec 27432 15.5 2.455 3.343 9.062 9.620 qs_ot_get_derivative 108 11.5 0.001 0.001 8.661 8.715 dbcsr_mm_accdrv_process 47916 15.9 5.833 7.449 6.506 7.854 rebuild_ks_matrix 119 8.3 0.000 0.000 6.451 6.506 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.451 6.506 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.719 5.768 qs_ot_get_p 119 10.4 0.001 0.001 3.754 3.822 init_scf_run 11 5.9 0.000 0.001 3.723 3.723 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.722 3.723 sum_up_and_integrate 119 10.3 0.001 0.001 3.545 3.552 integrate_v_rspace 119 11.3 0.003 0.003 3.533 3.540 init_scf_loop 11 6.9 0.000 0.000 3.519 3.520 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.041 3.514 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.436 3.475 calculate_rho_elec 119 8.7 0.040 0.046 3.435 3.474 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.176 2.641 apply_single 119 13.6 0.000 0.000 2.175 2.641 prepare_preconditioner 11 7.9 0.000 0.000 2.579 2.585 make_preconditioner 11 8.9 0.000 0.000 2.578 2.585 acc_transpose_blocks 27432 15.5 0.119 0.121 2.415 2.548 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.179 2.513 make_m2s 4572 13.5 0.054 0.055 2.297 2.422 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.344 2.371 mp_waitall_1 137007 16.6 1.875 2.369 1.875 2.369 make_images 4572 14.5 0.274 0.335 2.190 2.313 calculate_first_density_matrix 1 7.0 0.000 0.000 2.307 2.310 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.255 2.267 calculate_dm_sparse 119 9.5 0.000 0.000 2.216 2.266 density_rs2pw 119 9.7 0.004 0.004 1.897 2.066 pw_transfer 1439 11.6 0.065 0.069 2.029 2.063 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.937 1.974 grid_integrate_task_list 119 12.3 1.832 1.913 1.832 1.913 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.854 1.854 ot_diis_step 108 11.5 0.012 0.012 1.830 1.830 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.793 1.794 jit_kernel_multiply 9 15.9 0.605 1.680 0.605 1.680 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.621 1.634 mp_sum_l 7287 12.8 1.037 1.632 1.037 1.632 fft_wrap_pw1pw2_140 487 13.2 0.227 0.236 1.589 1.627 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.600 1.609 cp_fm_diag_elpa_base 50 14.0 1.565 1.581 1.598 1.607 acc_transpose_blocks_sync 82296 16.5 1.392 1.524 1.392 1.524 fft3d_ps 1201 14.6 0.592 0.653 1.425 1.450 multiply_cannon_metrocomm3 27432 15.5 0.040 0.041 0.818 1.388 wfi_extrapolate 11 7.9 0.001 0.001 1.362 1.362 potential_pw2rs 119 12.3 0.008 0.009 1.348 1.350 grid_collocate_task_list 119 9.7 1.249 1.345 1.249 1.345 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.311 1.319 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.291 1.309 cp_fm_upper_to_full 72 14.2 0.834 1.188 0.834 1.188 transfer_rs2pw 487 10.6 0.004 0.005 0.955 1.132 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.074 1.075 dbcsr_complete_redistribute 329 12.2 0.121 0.150 0.778 1.053 make_images_data 4572 15.5 0.048 0.052 0.851 0.976 hybrid_alltoall_any 4725 16.4 0.066 0.156 0.727 0.902 acc_transpose_blocks_kernels 27432 16.5 0.273 0.279 0.874 0.881 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.800 0.877 mp_alltoall_d11v 2130 13.8 0.737 0.853 0.737 0.853 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.574 0.843 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.821 0.825 multiply_cannon_metrocomm1 27432 15.5 0.035 0.036 0.199 0.791 transfer_pw2rs 487 13.2 0.004 0.004 0.769 0.772 cp_fm_cholesky_invert 11 10.9 0.768 0.771 0.768 0.771 mp_alltoall_z22v 1201 16.6 0.716 0.747 0.716 0.747 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=36.983000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=531.000000, yerr=3.567530 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c078969bf945a475acf73b54de8647dac0b903ea_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 616.411136E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63488. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.031 28.754 28.756 qs_mol_dyn_low 1 2.0 0.003 0.003 28.477 28.486 qs_forces 11 3.9 0.002 0.002 28.418 28.419 qs_energies 11 4.9 0.001 0.001 26.711 26.714 scf_env_do_scf 11 5.9 0.000 0.001 21.871 21.871 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 19.306 19.307 velocity_verlet 10 3.0 0.001 0.002 14.594 14.597 dbcsr_multiply_generic 2286 12.5 0.095 0.099 13.095 13.159 qs_scf_new_mos 108 7.5 0.001 0.001 11.669 11.690 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.668 11.689 multiply_cannon 2286 13.5 0.226 0.230 10.600 11.025 ot_scf_mini 108 9.5 0.002 0.002 10.963 10.986 multiply_cannon_loop 2286 14.5 0.637 0.653 9.715 9.856 multiply_cannon_multrec 9144 15.5 1.723 1.930 6.176 6.383 ot_mini 108 10.5 0.001 0.001 6.299 6.328 rebuild_ks_matrix 119 8.3 0.000 0.000 5.783 5.804 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.782 5.804 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.146 5.166 qs_ot_get_derivative 108 11.5 0.001 0.001 5.014 5.037 dbcsr_mm_accdrv_process 12550 15.8 3.691 4.359 4.344 4.428 sum_up_and_integrate 119 10.3 0.001 0.001 3.412 3.417 integrate_v_rspace 119 11.3 0.003 0.003 3.402 3.407 init_scf_run 11 5.9 0.000 0.001 3.401 3.401 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.401 3.401 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.352 3.359 calculate_rho_elec 119 8.7 0.060 0.061 3.352 3.358 qs_ot_get_p 119 10.4 0.001 0.001 2.895 2.934 init_scf_loop 11 6.9 0.000 0.000 2.545 2.546 calculate_first_density_matrix 1 7.0 0.000 0.000 2.270 2.271 pw_transfer 1439 11.6 0.065 0.068 2.105 2.114 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.012 2.022 grid_integrate_task_list 119 12.3 1.866 1.948 1.866 1.948 calculate_dm_sparse 119 9.5 0.000 0.001 1.852 1.867 make_m2s 4572 13.5 0.033 0.034 1.740 1.863 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.829 1.831 density_rs2pw 119 9.7 0.003 0.004 1.735 1.793 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.770 1.785 make_images 4572 14.5 0.270 0.302 1.651 1.773 prepare_preconditioner 11 7.9 0.000 0.000 1.702 1.706 make_preconditioner 11 8.9 0.000 0.000 1.702 1.706 fft_wrap_pw1pw2_140 487 13.2 0.324 0.331 1.677 1.690 mp_waitall_1 115863 16.7 1.227 1.671 1.227 1.671 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.595 1.620 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.587 1.588 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.576 1.577 acc_transpose_blocks 9144 15.5 0.042 0.044 1.506 1.531 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.415 1.427 fft3d_ps 1201 14.6 0.650 0.664 1.365 1.374 grid_collocate_task_list 119 9.7 1.295 1.363 1.295 1.363 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.329 1.342 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.299 1.308 cp_fm_diag_elpa_base 50 14.0 1.273 1.290 1.298 1.306 ot_diis_step 108 11.5 0.012 0.013 1.273 1.273 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.216 1.251 apply_single 119 13.6 0.000 0.000 1.216 1.251 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.242 1.248 qs_energies_init_hamiltonians 11 5.9 0.000 0.000 1.227 1.229 potential_pw2rs 119 12.3 0.010 0.011 1.225 1.227 jit_kernel_multiply 6 15.5 0.613 1.164 0.613 1.164 wfi_extrapolate 11 7.9 0.001 0.001 1.080 1.080 hybrid_alltoall_any 4725 16.4 0.065 0.179 0.758 0.929 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.875 0.921 make_images_data 4572 15.5 0.042 0.046 0.739 0.879 mp_alltoall_d11v 2130 13.8 0.787 0.853 0.787 0.853 cp_fm_cholesky_invert 11 10.9 0.850 0.853 0.850 0.853 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.808 0.815 transfer_rs2pw 487 10.6 0.004 0.004 0.748 0.811 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.757 0.759 acc_transpose_blocks_kernels 9144 16.5 0.118 0.121 0.736 0.739 acc_transpose_blocks_sync 27432 16.5 0.711 0.734 0.711 0.734 qs_env_update_s_mstruct 11 6.9 0.000 0.001 0.674 0.721 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 0.312 0.637 transfer_pw2rs 487 13.2 0.003 0.004 0.628 0.630 mp_alltoall_z22v 1201 16.6 0.589 0.622 0.589 0.622 jit_kernel_transpose 5 15.6 0.618 0.621 0.618 0.621 mp_allgather_i34 2286 14.5 0.198 0.596 0.198 0.596 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=28.756000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=581.636364, yerr=7.760889 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c078969bf945a475acf73b54de8647dac0b903ea_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 788.189184E+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 0.015 0.038 43.092 43.093 qs_mol_dyn_low 1 2.0 0.003 0.004 42.761 42.768 qs_forces 11 3.9 0.002 0.003 42.695 42.697 qs_energies 11 4.9 0.001 0.001 40.665 40.669 scf_env_do_scf 11 5.9 0.001 0.001 34.883 34.883 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 27.048 27.050 velocity_verlet 10 3.0 0.002 0.002 24.198 24.204 dbcsr_multiply_generic 2286 12.5 0.105 0.107 19.015 19.291 qs_scf_new_mos 108 7.5 0.001 0.001 17.335 17.436 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.334 17.435 ot_scf_mini 108 9.5 0.002 0.002 16.164 16.269 multiply_cannon 2286 13.5 0.302 0.309 15.023 16.081 multiply_cannon_loop 2286 14.5 0.873 0.881 13.745 14.817 ot_mini 108 10.5 0.001 0.001 9.833 9.954 multiply_cannon_multrec 9144 15.5 3.400 4.635 8.818 8.927 qs_ot_get_derivative 108 11.5 0.001 0.001 7.747 7.851 init_scf_loop 11 6.9 0.000 0.000 7.808 7.811 rebuild_ks_matrix 119 8.3 0.000 0.000 7.281 7.429 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.280 7.428 prepare_preconditioner 11 7.9 0.000 0.000 6.825 6.839 make_preconditioner 11 8.9 0.000 0.000 6.825 6.839 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.382 6.716 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.566 6.700 dbcsr_mm_accdrv_process 12550 15.8 4.733 6.545 5.279 6.590 cp_fm_upper_to_full 72 14.2 3.208 4.595 3.208 4.595 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.217 4.219 calculate_rho_elec 119 8.7 0.118 0.120 4.216 4.219 sum_up_and_integrate 119 10.3 0.001 0.001 3.832 3.839 integrate_v_rspace 119 11.3 0.003 0.004 3.821 3.828 init_scf_run 11 5.9 0.000 0.001 3.718 3.718 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.718 3.718 qs_ot_get_p 119 10.4 0.001 0.001 3.386 3.524 mp_waitall_1 94719 16.7 2.260 3.240 2.260 3.240 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.815 3.222 pw_transfer 1439 11.6 0.068 0.069 3.038 3.043 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.939 2.945 dbcsr_complete_redistribute 329 12.2 0.282 0.288 1.969 2.774 make_m2s 4572 13.5 0.037 0.037 2.425 2.577 fft_wrap_pw1pw2_140 487 13.2 0.652 0.660 2.510 2.517 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.233 2.508 apply_single 119 13.6 0.000 0.000 2.233 2.508 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.674 2.480 make_images 4572 14.5 0.353 0.386 2.304 2.456 calculate_dm_sparse 119 9.5 0.000 0.000 2.275 2.292 mp_alltoall_i22 627 13.8 1.418 2.269 1.418 2.269 density_rs2pw 119 9.7 0.004 0.004 2.243 2.264 calculate_first_density_matrix 1 7.0 0.000 0.000 2.256 2.258 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.437 2.237 multiply_cannon_metrocomm3 9144 15.5 0.021 0.022 1.251 2.180 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.090 2.144 grid_integrate_task_list 119 12.3 2.076 2.098 2.076 2.098 ot_diis_step 108 11.5 0.014 0.015 2.061 2.061 qs_ot_p2m_diag 50 11.0 0.043 0.044 1.997 1.998 acc_transpose_blocks 9144 15.5 0.045 0.046 1.830 1.863 fft3d_ps 1201 14.6 0.871 0.888 1.857 1.859 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.817 1.818 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.779 1.780 mp_sum_l 7287 12.8 1.051 1.764 1.051 1.764 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.635 1.635 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.567 1.614 grid_collocate_task_list 119 9.7 1.519 1.537 1.519 1.537 cp_fm_cholesky_invert 11 10.9 1.436 1.439 1.436 1.439 wfi_extrapolate 11 7.9 0.001 0.001 1.398 1.398 hybrid_alltoall_any 4725 16.4 0.091 0.150 1.141 1.383 potential_pw2rs 119 12.3 0.014 0.014 1.357 1.358 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.346 1.346 cp_fm_diag_elpa_base 50 14.0 1.200 1.252 1.344 1.344 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.310 1.328 make_images_data 4572 15.5 0.046 0.049 1.083 1.279 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.272 1.277 mp_alltoall_d11v 2130 13.8 1.156 1.182 1.156 1.182 acc_transpose_blocks_sync 27432 16.5 1.117 1.133 1.117 1.133 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.105 1.125 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.005 1.053 qs_create_task_list 11 7.9 0.001 0.001 0.928 0.941 generate_qs_task_list 11 8.9 0.368 0.387 0.927 0.940 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.924 0.938 mp_alltoall_z22v 1201 16.6 0.849 0.870 0.849 0.870 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=43.093000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=735.727273, yerr=16.652005 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c078969bf945a475acf73b54de8647dac0b903ea_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 502.620160E+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 1433069. 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 0.030 0.051 82.171 82.176 qs_mol_dyn_low 1 2.0 0.003 0.003 81.361 81.611 qs_forces 11 3.9 0.003 0.003 81.290 81.291 qs_energies 11 4.9 0.012 0.057 78.464 78.489 scf_env_do_scf 11 5.9 0.001 0.003 69.372 69.376 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 63.814 63.815 dbcsr_multiply_generic 2055 12.4 0.104 0.106 50.931 51.294 qs_scf_new_mos 99 7.5 0.000 0.001 46.921 47.063 qs_scf_loop_do_ot 99 8.5 0.000 0.001 46.920 47.062 ot_scf_mini 99 9.5 0.002 0.003 44.595 44.728 multiply_cannon 2055 13.4 0.182 0.188 42.737 43.545 multiply_cannon_loop 2055 14.4 1.793 1.839 41.769 42.645 velocity_verlet 10 3.0 0.001 0.001 42.601 42.601 ot_mini 99 10.5 0.001 0.001 26.472 26.583 qs_ot_get_derivative 99 11.5 0.001 0.001 19.686 19.815 multiply_cannon_multrec 49320 15.4 11.381 12.011 17.578 18.332 rebuild_ks_matrix 110 8.3 0.000 0.001 14.564 14.752 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.026 14.563 14.752 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.803 12.963 mp_waitall_1 220248 16.4 10.634 11.501 10.634 11.501 multiply_cannon_sync_h2d 49320 15.4 9.624 10.284 9.624 10.284 qs_ot_get_p 110 10.4 0.001 0.001 9.664 9.772 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.661 8.188 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.268 7.729 apply_single 110 13.6 0.000 0.000 7.268 7.728 multiply_cannon_metrocomm3 49320 15.4 0.085 0.088 6.449 7.684 sum_up_and_integrate 110 10.3 0.002 0.004 6.966 6.979 integrate_v_rspace 110 11.3 0.003 0.003 6.940 6.960 init_scf_run 11 5.9 0.000 0.001 6.923 6.923 scf_env_initial_rho_setup 11 6.9 0.001 0.002 6.923 6.923 ot_diis_step 99 11.5 0.005 0.006 6.607 6.607 qs_ot_p2m_diag 48 11.0 0.013 0.019 6.568 6.593 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.249 6.402 calculate_rho_elec 110 8.6 0.021 0.025 6.249 6.402 dbcsr_mm_accdrv_process 87628 16.1 3.077 3.155 6.065 6.332 cp_dbcsr_syevd 48 12.0 0.003 0.003 5.718 5.719 init_scf_loop 11 6.9 0.002 0.006 5.530 5.533 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.340 5.404 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.152 5.178 cp_fm_diag_elpa_base 48 14.0 5.140 5.164 5.151 5.177 mp_sum_l 6594 12.7 3.940 4.677 3.940 4.677 wfi_extrapolate 11 7.9 0.001 0.001 3.992 3.992 make_m2s 4110 13.4 0.060 0.065 3.824 3.936 make_images 4110 14.4 0.177 0.189 3.729 3.844 calculate_dm_sparse 110 9.5 0.001 0.001 3.693 3.813 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.745 3.749 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.399 3.451 grid_integrate_task_list 110 12.3 3.252 3.446 3.252 3.446 density_rs2pw 110 9.6 0.004 0.004 3.263 3.441 multiply_cannon_metrocomm1 49320 15.4 0.067 0.070 2.385 3.440 prepare_preconditioner 11 7.9 0.000 0.000 3.192 3.217 make_preconditioner 11 8.9 0.000 0.001 3.192 3.217 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.148 3.188 pw_transfer 1331 11.6 0.055 0.065 2.993 3.081 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.997 3.059 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 2.905 2.995 calculate_first_density_matrix 1 7.0 0.002 0.011 2.830 2.835 jit_kernel_multiply 13 15.9 2.702 2.716 2.702 2.716 acc_transpose_blocks 49320 15.4 0.213 0.218 2.628 2.685 fft_wrap_pw1pw2_140 451 13.1 0.312 0.334 2.491 2.592 mp_alltoall_d11v 2046 13.8 2.074 2.567 2.074 2.567 potential_pw2rs 110 12.3 0.005 0.006 2.534 2.552 grid_collocate_task_list 110 9.6 2.163 2.302 2.163 2.302 fft3d_ps 1111 14.6 0.792 0.882 2.195 2.255 mp_waitany 14300 13.8 1.836 2.117 1.836 2.117 transfer_rs2pw 451 10.6 0.005 0.006 1.884 2.046 mp_sum_d 3891 11.9 1.407 1.979 1.407 1.979 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.894 1.921 make_images_data 4110 15.4 0.044 0.048 1.752 1.883 cp_fm_cholesky_invert 11 10.9 1.776 1.779 1.776 1.779 hybrid_alltoall_any 4261 16.3 0.085 0.486 1.512 1.755 transfer_pw2rs 451 13.1 0.006 0.007 1.721 1.729 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.650 1.684 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=82.176000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=477.636364, yerr=2.532434 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c078969bf945a475acf73b54de8647dac0b903ea_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 593.686528E+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 3252905. 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 0.015 0.033 68.647 68.649 qs_mol_dyn_low 1 2.0 0.003 0.003 68.345 68.360 qs_forces 11 3.9 0.003 0.008 68.180 68.182 qs_energies 11 4.9 0.002 0.007 64.864 64.868 scf_env_do_scf 11 5.9 0.001 0.002 56.325 56.328 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 48.790 48.790 dbcsr_multiply_generic 2055 12.4 0.115 0.119 37.935 38.077 velocity_verlet 10 3.0 0.001 0.001 35.935 35.939 qs_scf_new_mos 99 7.5 0.001 0.001 32.997 33.114 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.996 33.113 multiply_cannon 2055 13.4 0.224 0.244 31.396 32.494 ot_scf_mini 99 9.5 0.006 0.031 31.327 31.463 multiply_cannon_loop 2055 14.4 1.155 1.189 30.124 30.871 ot_mini 99 10.5 0.001 0.001 18.293 18.435 multiply_cannon_multrec 24660 15.4 6.950 8.736 14.020 15.745 rebuild_ks_matrix 110 8.3 0.000 0.001 13.553 13.651 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.017 13.553 13.651 qs_ot_get_derivative 99 11.5 0.001 0.002 12.527 12.663 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.935 12.029 mp_waitall_1 176588 16.5 7.511 10.115 7.511 10.115 multiply_cannon_metrocomm3 24660 15.4 0.074 0.077 5.135 7.760 init_scf_loop 11 6.9 0.001 0.009 7.500 7.501 multiply_cannon_sync_h2d 24660 15.4 6.435 7.482 6.435 7.482 dbcsr_mm_accdrv_process 52282 16.1 5.642 6.403 6.901 7.217 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.466 7.140 apply_single 110 13.6 0.000 0.001 6.465 7.139 qs_ot_get_p 110 10.4 0.001 0.001 6.361 6.528 sum_up_and_integrate 110 10.3 0.002 0.004 6.348 6.361 integrate_v_rspace 110 11.3 0.003 0.004 6.321 6.333 init_scf_run 11 5.9 0.000 0.001 6.132 6.133 scf_env_initial_rho_setup 11 6.9 0.001 0.003 6.132 6.133 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.819 5.830 calculate_rho_elec 110 8.6 0.039 0.047 5.818 5.830 ot_diis_step 99 11.5 0.010 0.010 5.719 5.719 prepare_preconditioner 11 7.9 0.000 0.000 5.468 5.497 make_preconditioner 11 8.9 0.000 0.001 5.468 5.497 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.721 5.438 make_full_inverse_cholesky 11 9.9 0.000 0.001 5.061 5.223 make_m2s 4110 13.4 0.056 0.059 4.178 4.648 make_images 4110 14.4 0.409 0.467 4.068 4.533 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.502 4.522 cp_dbcsr_syevd 48 12.0 0.003 0.004 4.039 4.039 pw_transfer 1331 11.6 0.066 0.072 3.511 3.647 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.404 3.545 wfi_extrapolate 11 7.9 0.001 0.001 3.498 3.498 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.476 3.490 cp_fm_diag_elpa_base 48 14.0 3.429 3.445 3.472 3.486 grid_integrate_task_list 110 12.3 3.156 3.377 3.156 3.377 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.256 3.323 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.304 3.306 density_rs2pw 110 9.6 0.004 0.004 3.069 3.238 fft_wrap_pw1pw2_140 451 13.1 0.358 0.378 2.930 3.075 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.969 3.003 calculate_dm_sparse 110 9.5 0.001 0.001 2.971 3.000 hybrid_alltoall_any 4261 16.3 0.105 0.450 1.979 2.767 make_images_data 4110 15.4 0.049 0.052 2.262 2.751 fft3d_ps 1111 14.6 1.116 1.345 2.495 2.651 calculate_first_density_matrix 1 7.0 0.000 0.003 2.537 2.540 cp_fm_cholesky_invert 11 10.9 2.533 2.540 2.533 2.540 mp_sum_l 6594 12.7 1.732 2.355 1.732 2.355 grid_collocate_task_list 110 9.6 2.170 2.338 2.170 2.338 potential_pw2rs 110 12.3 0.008 0.009 2.228 2.236 acc_transpose_blocks 24660 15.4 0.116 0.119 2.001 2.060 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.955 1.969 mp_alltoall_d11v 2046 13.8 1.757 1.958 1.757 1.958 jit_kernel_multiply 10 16.4 0.889 1.901 0.889 1.901 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.874 1.876 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.781 1.789 transfer_rs2pw 451 10.6 0.006 0.007 1.407 1.652 mp_allgather_i34 2055 14.4 0.578 1.604 0.578 1.604 multiply_cannon_metrocomm4 22605 15.4 0.073 0.077 0.761 1.585 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.534 1.545 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.390 1.491 mp_irecv_dv 57340 16.2 0.637 1.467 0.637 1.467 mp_waitany 10164 13.8 1.166 1.402 1.166 1.402 dbcsr_complete_redistribute 325 12.2 0.235 0.298 1.120 1.379 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=68.649000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=561.363636, yerr=7.522968 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c078969bf945a475acf73b54de8647dac0b903ea_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 669.270016E+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 0.062 0.079 62.379 62.382 qs_mol_dyn_low 1 2.0 0.003 0.004 61.924 61.965 qs_forces 11 3.9 0.005 0.010 61.846 61.847 qs_energies 11 4.9 0.003 0.018 58.580 58.585 scf_env_do_scf 11 5.9 0.001 0.002 50.437 50.437 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 41.394 41.395 velocity_verlet 10 3.0 0.001 0.002 33.539 33.541 dbcsr_multiply_generic 2055 12.4 0.109 0.113 29.114 29.372 qs_scf_new_mos 99 7.5 0.001 0.001 26.343 26.439 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.342 26.438 ot_scf_mini 99 9.5 0.002 0.002 25.074 25.197 multiply_cannon 2055 13.4 0.212 0.220 22.548 23.800 multiply_cannon_loop 2055 14.4 0.814 0.838 21.300 22.448 ot_mini 99 10.5 0.001 0.001 14.005 14.124 rebuild_ks_matrix 110 8.3 0.000 0.001 12.444 12.592 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.019 12.443 12.592 multiply_cannon_multrec 16440 15.4 3.716 5.012 9.929 11.167 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.926 11.057 mp_waitall_1 139946 16.5 7.035 10.015 7.035 10.015 qs_ot_get_derivative 99 11.5 0.001 0.001 9.566 9.687 init_scf_loop 11 6.9 0.001 0.003 9.001 9.001 prepare_preconditioner 11 7.9 0.000 0.000 7.191 7.210 make_preconditioner 11 8.9 0.000 0.000 7.191 7.210 multiply_cannon_metrocomm3 16440 15.4 0.046 0.047 4.220 7.118 make_full_inverse_cholesky 11 9.9 0.000 0.001 6.496 6.863 sum_up_and_integrate 110 10.3 0.002 0.003 6.488 6.510 integrate_v_rspace 110 11.3 0.003 0.003 6.462 6.484 qs_ot_get_p 110 10.4 0.001 0.002 6.166 6.323 dbcsr_mm_accdrv_process 34862 16.1 5.396 5.741 6.058 6.157 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.983 5.993 calculate_rho_elec 110 8.6 0.059 0.059 5.982 5.993 init_scf_run 11 5.9 0.000 0.001 5.650 5.650 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.650 5.650 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.873 5.386 apply_single 110 13.6 0.000 0.000 4.873 5.386 make_m2s 4110 13.4 0.049 0.051 4.460 4.833 make_images 4110 14.4 0.401 0.525 4.344 4.717 ot_diis_step 99 11.5 0.010 0.011 4.408 4.408 qs_ot_p2m_diag 48 11.0 0.042 0.044 4.377 4.381 cp_dbcsr_syevd 48 12.0 0.003 0.004 3.989 3.990 multiply_cannon_sync_h2d 16440 15.4 3.231 3.926 3.231 3.926 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.298 3.870 pw_transfer 1331 11.6 0.066 0.071 3.824 3.846 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.717 3.741 grid_integrate_task_list 110 12.3 3.213 3.398 3.213 3.398 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.375 3.383 cp_fm_diag_elpa_base 48 14.0 3.297 3.335 3.372 3.380 density_rs2pw 110 9.6 0.004 0.004 3.189 3.369 fft_wrap_pw1pw2_140 451 13.1 0.460 0.471 3.167 3.191 wfi_extrapolate 11 7.9 0.001 0.001 3.095 3.095 make_images_data 4110 15.4 0.047 0.051 2.476 2.976 cp_fm_cholesky_invert 11 10.9 2.925 2.932 2.925 2.932 hybrid_alltoall_any 4261 16.3 0.108 0.382 2.226 2.885 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.809 2.810 fft3d_ps 1111 14.6 1.122 1.145 2.635 2.657 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.561 2.629 calculate_dm_sparse 110 9.5 0.001 0.001 2.550 2.582 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.447 2.497 calculate_first_density_matrix 1 7.0 0.000 0.003 2.459 2.460 multiply_cannon_metrocomm4 14385 15.4 0.047 0.051 0.872 2.406 grid_collocate_task_list 110 9.6 2.216 2.373 2.216 2.373 potential_pw2rs 110 12.3 0.011 0.011 2.301 2.322 mp_irecv_dv 48980 15.7 0.798 2.271 0.798 2.271 mp_sum_l 6594 12.7 1.487 2.120 1.487 2.120 mp_alltoall_d11v 2046 13.8 1.877 2.092 1.877 2.092 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.993 1.998 dbcsr_complete_redistribute 325 12.2 0.318 0.350 1.503 1.980 cp_fm_upper_to_full 70 14.2 1.468 1.861 1.468 1.861 cp_fm_cholesky_decompose 22 10.9 1.737 1.782 1.737 1.782 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.715 1.732 mp_allgather_i34 2055 14.4 0.557 1.679 0.557 1.679 acc_transpose_blocks 16440 15.4 0.078 0.081 1.585 1.654 transfer_rs2pw 451 10.6 0.005 0.005 1.389 1.564 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.039 1.511 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.476 1.489 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.361 1.481 mp_waitany 17072 13.8 1.169 1.385 1.169 1.385 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.318 1.327 mp_alltoall_z22v 1111 16.6 1.299 1.327 1.299 1.327 transfer_pw2rs 451 13.1 0.005 0.005 1.251 1.261 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=62.382000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=633.000000, yerr=8.409951 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c078969bf945a475acf73b54de8647dac0b903ea_performance_tests/12/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 601.317074E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 4916280 0.0% 0.0% 100.0% average stack size 0.0 0.0 200.2 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 740.147200E+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 0.010 0.031 65.839 65.840 qs_mol_dyn_low 1 2.0 0.003 0.004 65.617 65.629 qs_forces 11 3.9 0.003 0.003 65.477 65.478 qs_energies 11 4.9 0.001 0.002 62.090 62.093 scf_env_do_scf 11 5.9 0.000 0.001 53.527 53.529 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 41.944 41.944 velocity_verlet 10 3.0 0.012 0.027 37.246 37.250 dbcsr_multiply_generic 2055 12.4 0.117 0.123 30.514 30.744 qs_scf_new_mos 99 7.5 0.001 0.001 27.425 27.522 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.424 27.521 ot_scf_mini 99 9.5 0.003 0.003 25.738 25.824 multiply_cannon 2055 13.4 0.240 0.261 23.638 24.672 multiply_cannon_loop 2055 14.4 1.390 1.445 22.229 22.845 ot_mini 99 10.5 0.001 0.001 14.873 14.983 multiply_cannon_multrec 24660 15.4 4.118 6.809 13.062 14.215 rebuild_ks_matrix 110 8.3 0.000 0.000 11.885 11.981 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.885 11.980 init_scf_loop 11 6.9 0.000 0.001 11.544 11.545 qs_ot_get_derivative 99 11.5 0.001 0.002 10.703 10.796 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.486 10.573 prepare_preconditioner 11 7.9 0.000 0.000 9.828 9.844 make_preconditioner 11 8.9 0.000 0.000 9.828 9.844 dbcsr_mm_accdrv_process 52304 16.0 7.751 8.985 8.786 9.697 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.010 9.501 sum_up_and_integrate 110 10.3 0.001 0.002 6.195 6.208 integrate_v_rspace 110 11.3 0.003 0.003 6.168 6.181 qs_ot_get_p 110 10.4 0.001 0.001 5.781 5.902 mp_waitall_1 121746 16.5 4.170 5.847 4.170 5.847 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.821 5.832 calculate_rho_elec 110 8.6 0.078 0.081 5.820 5.832 make_m2s 4110 13.4 0.059 0.061 5.286 5.646 init_scf_run 11 5.9 0.000 0.001 5.574 5.574 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.573 5.574 make_images 4110 14.4 0.579 0.700 5.146 5.502 cp_fm_upper_to_full 70 14.2 3.387 4.922 3.387 4.922 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.055 4.147 apply_single 110 13.6 0.000 0.000 4.054 4.147 ot_diis_step 99 11.5 0.011 0.011 4.132 4.132 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.939 3.957 dbcsr_complete_redistribute 325 12.2 0.417 0.456 2.684 3.837 pw_transfer 1331 11.6 0.065 0.072 3.765 3.801 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.658 3.698 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.466 3.510 grid_integrate_task_list 110 12.3 3.294 3.447 3.294 3.447 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.378 3.379 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.317 3.367 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.157 3.298 multiply_cannon_metrocomm3 24660 15.4 0.037 0.038 1.505 3.277 fft_wrap_pw1pw2_140 451 13.1 0.531 0.543 3.163 3.206 density_rs2pw 110 9.6 0.004 0.004 2.958 3.087 calculate_dm_sparse 110 9.5 0.001 0.001 3.020 3.063 hybrid_alltoall_any 4261 16.3 0.123 0.462 2.247 2.962 wfi_extrapolate 11 7.9 0.001 0.001 2.950 2.950 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.809 2.938 make_images_data 4110 15.4 0.049 0.053 2.569 2.933 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.842 2.853 cp_fm_diag_elpa_base 48 14.0 2.693 2.753 2.839 2.851 mp_alltoall_i22 605 13.7 1.675 2.846 1.675 2.846 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.763 2.765 cp_fm_cholesky_invert 11 10.9 2.607 2.615 2.607 2.615 calculate_first_density_matrix 1 7.0 0.000 0.000 2.529 2.532 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 2.500 2.501 fft3d_ps 1111 14.6 1.154 1.196 2.483 2.499 acc_transpose_blocks 24660 15.4 0.112 0.115 2.389 2.494 multiply_cannon_sync_h2d 24660 15.4 2.377 2.452 2.377 2.452 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.408 2.445 grid_collocate_task_list 110 9.6 2.264 2.404 2.264 2.404 potential_pw2rs 110 12.3 0.012 0.013 2.023 2.028 mp_alltoall_d11v 2046 13.8 1.727 1.900 1.727 1.900 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.782 1.823 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.616 1.707 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.639 1.648 cp_fm_cholesky_decompose 22 10.9 1.593 1.629 1.593 1.629 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.575 1.593 jit_kernel_multiply 9 15.9 0.685 1.593 0.685 1.593 mp_sum_l 6594 12.7 0.897 1.560 0.897 1.560 multiply_cannon_metrocomm4 20550 15.4 0.059 0.062 0.838 1.518 acc_transpose_blocks_sync 73980 16.4 1.376 1.475 1.376 1.475 qs_env_update_s_mstruct 11 6.9 0.064 0.259 1.268 1.474 mp_allgather_i34 2055 14.4 0.475 1.452 0.475 1.452 mp_irecv_dv 62702 16.1 0.736 1.440 0.736 1.440 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=65.840000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=700.181818, yerr=10.364434 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c078969bf945a475acf73b54de8647dac0b903ea_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 873.172992E+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 0.011 0.028 55.161 55.162 qs_mol_dyn_low 1 2.0 0.003 0.004 54.892 54.902 qs_forces 11 3.9 0.003 0.003 54.825 54.826 qs_energies 11 4.9 0.001 0.002 51.183 51.187 scf_env_do_scf 11 5.9 0.000 0.001 42.855 42.856 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 35.225 35.226 velocity_verlet 10 3.0 0.002 0.002 31.059 31.062 dbcsr_multiply_generic 2055 12.4 0.132 0.134 23.261 23.403 qs_scf_new_mos 99 7.5 0.001 0.001 20.614 20.656 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.613 20.655 ot_scf_mini 99 9.5 0.002 0.002 19.358 19.377 multiply_cannon 2055 13.4 0.245 0.258 17.666 19.019 multiply_cannon_loop 2055 14.4 0.606 0.625 16.384 16.736 rebuild_ks_matrix 110 8.3 0.000 0.000 11.669 11.682 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.669 11.682 ot_mini 99 10.5 0.001 0.001 10.773 10.787 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.380 10.394 multiply_cannon_multrec 8220 15.4 3.199 4.335 7.594 8.424 init_scf_loop 11 6.9 0.000 0.001 7.584 7.586 mp_waitall_1 103326 16.6 5.839 7.344 5.839 7.344 qs_ot_get_derivative 99 11.5 0.001 0.002 7.083 7.103 sum_up_and_integrate 110 10.3 0.001 0.002 6.217 6.230 integrate_v_rspace 110 11.3 0.003 0.003 6.190 6.204 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.008 6.024 calculate_rho_elec 110 8.6 0.113 0.113 6.007 6.023 prepare_preconditioner 11 7.9 0.000 0.000 5.942 5.947 make_preconditioner 11 8.9 0.000 0.000 5.942 5.947 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.522 5.598 dbcsr_mm_accdrv_process 17442 15.9 3.110 4.145 4.255 5.183 init_scf_run 11 5.9 0.000 0.001 5.145 5.145 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.145 5.145 qs_ot_get_p 110 10.4 0.001 0.001 4.640 4.649 make_m2s 4110 13.4 0.037 0.038 4.228 4.516 make_images 4110 14.4 0.649 0.709 4.099 4.385 multiply_cannon_metrocomm3 8220 15.4 0.018 0.019 2.975 4.288 pw_transfer 1331 11.6 0.066 0.070 4.091 4.102 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.982 3.997 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.727 3.784 apply_single 110 13.6 0.000 0.000 3.727 3.783 ot_diis_step 99 11.5 0.012 0.012 3.666 3.666 grid_integrate_task_list 110 12.3 3.371 3.526 3.371 3.526 fft_wrap_pw1pw2_140 451 13.1 0.721 0.734 3.458 3.475 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.272 3.275 density_rs2pw 110 9.6 0.004 0.004 3.031 3.166 cp_dbcsr_syevd 48 12.0 0.003 0.003 2.932 2.933 cp_fm_cholesky_invert 11 10.9 2.815 2.819 2.815 2.819 hybrid_alltoall_any 4261 16.3 0.200 0.856 2.226 2.752 qs_energies_init_hamiltonians 11 5.9 0.002 0.008 2.730 2.730 wfi_extrapolate 11 7.9 0.001 0.001 2.707 2.707 make_images_data 4110 15.4 0.041 0.046 2.263 2.661 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.569 2.570 grid_collocate_task_list 110 9.6 2.369 2.562 2.369 2.562 fft3d_ps 1111 14.6 1.294 1.304 2.521 2.537 calculate_dm_sparse 110 9.5 0.001 0.001 2.484 2.521 multiply_cannon_sync_h2d 8220 15.4 2.378 2.470 2.378 2.470 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.424 2.431 cp_fm_diag_elpa_base 48 14.0 2.367 2.391 2.422 2.430 calculate_first_density_matrix 1 7.0 0.000 0.000 2.342 2.343 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.157 2.170 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.010 2.034 potential_pw2rs 110 12.3 0.015 0.015 2.019 2.022 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.781 1.992 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.911 1.922 mp_alltoall_d11v 2046 13.8 1.625 1.839 1.625 1.839 cp_fm_cholesky_decompose 22 10.9 1.664 1.683 1.664 1.683 qs_env_update_s_mstruct 11 6.9 0.033 0.037 1.549 1.667 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.617 1.623 dbcsr_complete_redistribute 325 12.2 0.589 0.617 1.486 1.577 mp_allgather_i34 2055 14.4 0.441 1.531 0.441 1.531 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.432 1.443 acc_transpose_blocks 8220 15.4 0.041 0.042 1.363 1.389 qs_create_task_list 11 7.9 0.001 0.001 1.215 1.314 generate_qs_task_list 11 8.9 0.375 0.443 1.215 1.313 transfer_rs2pw 451 10.6 0.005 0.005 1.117 1.284 multiply_cannon_metrocomm1 8220 15.4 0.022 0.023 0.783 1.211 mp_waitany 9240 13.8 1.060 1.211 1.060 1.211 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.159 1.186 jit_kernel_multiply 7 15.5 0.826 1.184 0.826 1.184 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=55.162000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=822.909091, yerr=12.999682 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c078969bf945a475acf73b54de8647dac0b903ea_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 1.395806E+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 0.045 0.098 89.261 89.295 qs_mol_dyn_low 1 2.0 0.003 0.004 88.844 88.865 qs_forces 11 3.9 0.003 0.003 88.743 88.744 qs_energies 11 4.9 0.002 0.002 84.545 84.547 scf_env_do_scf 11 5.9 0.000 0.001 74.465 74.465 velocity_verlet 10 3.0 0.002 0.002 56.911 56.921 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 45.739 45.740 dbcsr_multiply_generic 2055 12.4 0.121 0.125 29.923 30.078 init_scf_loop 11 6.9 0.000 0.000 28.652 28.654 qs_scf_new_mos 99 7.5 0.001 0.001 27.544 27.635 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.544 27.634 prepare_preconditioner 11 7.9 0.000 0.000 26.603 26.615 make_preconditioner 11 8.9 0.000 0.000 26.603 26.615 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.841 26.058 ot_scf_mini 99 9.5 0.002 0.002 25.719 25.796 multiply_cannon 2055 13.4 0.336 0.347 22.603 23.239 multiply_cannon_loop 2055 14.4 0.827 0.846 20.755 21.141 cp_fm_upper_to_full 70 14.2 12.826 18.293 12.826 18.293 ot_mini 99 10.5 0.001 0.001 14.493 14.569 rebuild_ks_matrix 110 8.3 0.001 0.001 14.130 14.224 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 14.130 14.223 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.818 12.902 dbcsr_complete_redistribute 325 12.2 1.002 1.042 7.353 10.494 multiply_cannon_multrec 8220 15.4 4.059 4.219 9.769 9.883 qs_ot_get_derivative 99 11.5 0.001 0.001 9.704 9.780 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.310 9.448 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.746 8.850 mp_waitall_1 84994 16.7 7.675 8.715 7.675 8.715 mp_alltoall_i22 605 13.7 5.373 8.549 5.373 8.549 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.602 7.642 calculate_rho_elec 110 8.6 0.222 0.223 7.601 7.642 sum_up_and_integrate 110 10.3 0.002 0.002 7.040 7.060 integrate_v_rspace 110 11.3 0.004 0.004 7.012 7.032 make_m2s 4110 13.4 0.042 0.043 5.452 5.979 qs_ot_get_p 110 10.4 0.001 0.001 5.874 5.973 pw_transfer 1331 11.6 0.075 0.075 5.875 5.886 make_images 4110 14.4 0.886 0.941 5.265 5.790 dbcsr_mm_accdrv_process 11614 15.7 3.838 4.172 5.560 5.785 init_scf_run 11 5.9 0.000 0.001 5.771 5.771 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.771 5.771 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 5.758 5.768 cp_fm_cholesky_invert 11 10.9 5.442 5.447 5.442 5.447 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.884 5.401 apply_single 110 13.6 0.000 0.000 4.884 5.401 multiply_cannon_metrocomm3 8220 15.4 0.019 0.020 4.944 5.232 fft_wrap_pw1pw2_140 451 13.1 1.368 1.375 5.046 5.057 ot_diis_step 99 11.5 0.015 0.016 4.760 4.761 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.205 4.214 density_rs2pw 110 9.6 0.004 0.004 4.019 4.048 grid_integrate_task_list 110 12.3 3.714 3.791 3.714 3.791 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.698 3.700 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.681 3.682 hybrid_alltoall_any 4261 16.3 0.264 0.564 2.829 3.577 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.086 3.555 make_images_data 4110 15.4 0.045 0.048 2.860 3.506 fft3d_ps 1111 14.6 1.886 1.890 3.462 3.468 wfi_extrapolate 11 7.9 0.001 0.001 3.377 3.377 calculate_dm_sparse 110 9.5 0.001 0.001 3.195 3.224 multiply_cannon_sync_h2d 8220 15.4 3.124 3.142 3.124 3.142 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.078 3.079 cp_fm_diag_elpa_base 48 14.0 2.541 2.737 3.077 3.077 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.890 2.892 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.717 2.759 grid_collocate_task_list 110 9.6 2.696 2.730 2.696 2.730 potential_pw2rs 110 12.3 0.021 0.021 2.495 2.504 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.377 2.408 calculate_first_density_matrix 1 7.0 0.000 0.000 2.274 2.275 qs_env_update_s_mstruct 11 6.9 0.009 0.018 2.205 2.262 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.108 2.213 mp_alltoall_d11v 2046 13.8 2.117 2.175 2.117 2.175 cp_fm_cholesky_decompose 22 10.9 2.047 2.063 2.047 2.063 qs_create_task_list 11 7.9 0.001 0.001 1.882 1.929 generate_qs_task_list 11 8.9 0.732 0.787 1.882 1.929 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.910 1.920 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.783 1.827 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=89.295000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1249.000000, yerr=57.105962 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c078969bf945a475acf73b54de8647dac0b903ea_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 1.086553E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 11851392 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.8 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 634.306560E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10258080 MPI messages size (bytes): total size 4.456715E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.459031E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3542056 94711185408 32768 < size <= 131072 1282176 73356279808 131072 < size <= 4194304 5107038 3151762421624 4194304 < size <= 16777216 259842 1136842803272 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4007 57688. MP_Allreduce 11097 796. MP_Sync 86 MP_Alltoall 2210 3042734. MP_SendRecv 24130 18752. MP_ISendRecv 24130 18752. MP_Wait 42150 MP_ISend 15900 108037. MP_IRecv 15900 108037. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.069 0.097 205.981 205.996 qs_mol_dyn_low 1 2.0 0.003 0.004 205.448 205.477 qs_forces 11 3.9 0.005 0.005 205.313 205.315 qs_energies 11 4.9 0.001 0.002 199.689 199.715 scf_env_do_scf 11 5.9 0.001 0.001 183.167 183.171 scf_env_do_scf_inner_loop 116 6.6 0.003 0.009 162.989 162.992 dbcsr_multiply_generic 2485 12.5 0.172 0.179 124.918 125.798 qs_scf_new_mos 116 7.6 0.001 0.001 123.885 124.199 qs_scf_loop_do_ot 116 8.6 0.001 0.001 123.884 124.198 velocity_verlet 10 3.0 0.006 0.021 121.904 121.905 ot_scf_mini 116 9.6 0.003 0.003 117.193 117.528 multiply_cannon 2485 13.5 0.235 0.244 100.953 102.810 multiply_cannon_loop 2485 14.5 2.389 2.443 98.808 100.634 ot_mini 116 10.6 0.001 0.001 65.801 66.114 multiply_cannon_multrec 59640 15.5 31.567 33.472 41.302 43.452 qs_ot_get_derivative 116 11.6 0.001 0.001 41.051 41.341 rebuild_ks_matrix 127 8.3 0.001 0.001 33.337 33.667 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.029 33.337 33.666 mp_waitall_1 264810 16.5 28.602 32.066 28.602 32.066 qs_ks_update_qs_env 127 7.6 0.001 0.001 29.894 30.231 qs_ot_get_p 127 10.4 0.001 0.001 29.235 29.535 multiply_cannon_sync_h2d 59640 15.5 26.284 28.337 26.284 28.337 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 24.273 25.145 apply_single 127 13.6 0.001 0.001 24.273 25.144 ot_diis_step 116 11.6 0.007 0.008 24.498 24.499 qs_ot_p2m_diag 82 11.4 0.078 0.090 22.646 22.730 init_scf_loop 11 6.9 0.000 0.000 20.104 20.106 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 19.716 19.992 cp_dbcsr_syevd 82 12.4 0.005 0.005 19.900 19.901 multiply_cannon_metrocomm3 59640 15.5 0.123 0.129 16.146 18.411 cp_fm_diag_elpa 82 13.4 0.000 0.000 16.970 17.006 cp_fm_diag_elpa_base 82 14.4 16.903 16.941 16.966 17.003 prepare_preconditioner 11 7.9 0.000 0.000 15.534 15.592 make_preconditioner 11 8.9 0.000 0.000 15.533 15.592 make_full_inverse_cholesky 11 9.9 0.000 0.000 14.787 15.003 make_m2s 4970 13.5 0.103 0.113 13.706 14.024 sum_up_and_integrate 127 10.3 0.002 0.005 13.871 13.893 make_images 4970 14.5 0.396 0.414 13.528 13.858 integrate_v_rspace 127 11.3 0.003 0.004 13.813 13.839 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.068 13.222 calculate_rho_elec 127 8.7 0.045 0.063 13.067 13.222 init_scf_run 11 5.9 0.000 0.001 12.333 12.334 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.333 12.334 mp_sum_l 7884 12.9 9.012 10.449 9.012 10.449 dbcsr_mm_accdrv_process 123452 16.2 4.711 4.835 9.309 9.804 wfi_extrapolate 11 7.9 0.001 0.001 9.077 9.077 cp_fm_cholesky_invert 11 10.9 8.800 8.807 8.800 8.807 calculate_dm_sparse 127 9.5 0.001 0.001 8.502 8.611 qs_ot_get_orbitals 116 10.6 0.001 0.001 8.099 8.183 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.997 8.142 multiply_cannon_metrocomm1 59640 15.5 0.096 0.100 6.004 7.708 make_images_data 4970 15.5 0.068 0.072 6.752 7.703 pw_transfer 1535 11.6 0.073 0.090 7.522 7.693 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 7.322 7.496 grid_integrate_task_list 127 12.3 7.030 7.462 7.030 7.462 hybrid_alltoall_any 5155 16.4 0.295 2.259 5.922 7.255 density_rs2pw 127 9.7 0.006 0.006 6.705 7.204 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.668 6.679 fft_wrap_pw1pw2_140 519 13.2 0.856 0.897 6.437 6.615 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.739 5.838 fft3d_ps 1281 14.7 2.177 2.818 5.334 5.650 mp_alltoall_d11v 2401 14.1 4.292 5.472 4.292 5.472 grid_collocate_task_list 127 9.7 4.797 5.203 4.797 5.203 cp_fm_cholesky_decompose 22 10.9 4.582 4.597 4.582 4.597 potential_pw2rs 127 12.3 0.009 0.010 4.449 4.470 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=205.996000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=598.000000, yerr=6.120012 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c078969bf945a475acf73b54de8647dac0b903ea_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 843.890688E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100943E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834845264 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 58214. MP_Allreduce 11083 960. MP_Sync 87 MP_Alltoall 1969 4571856. MP_SendRecv 12032 47072. MP_ISendRecv 12032 47072. MP_Wait 25916 MP_ISend 11748 212467. MP_IRecv 11748 212467. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.043 0.056 189.192 189.192 qs_mol_dyn_low 1 2.0 0.003 0.003 188.724 188.737 qs_forces 11 3.9 0.005 0.006 188.621 188.621 qs_energies 11 4.9 0.002 0.002 181.870 181.880 scf_env_do_scf 11 5.9 0.001 0.001 165.446 165.456 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 132.810 132.812 velocity_verlet 10 3.0 0.001 0.002 118.726 118.729 dbcsr_multiply_generic 2507 12.6 0.190 0.198 97.756 99.088 qs_scf_new_mos 117 7.6 0.001 0.001 94.293 94.781 qs_scf_loop_do_ot 117 8.6 0.001 0.001 94.292 94.780 ot_scf_mini 117 9.6 0.004 0.004 89.481 90.074 multiply_cannon 2507 13.6 0.565 0.639 77.599 81.515 multiply_cannon_loop 2507 14.6 1.561 1.630 74.114 77.157 ot_mini 117 10.6 0.001 0.001 50.280 50.867 mp_waitall_1 214728 16.6 24.000 39.870 24.000 39.870 multiply_cannon_multrec 30084 15.6 20.983 26.288 31.625 37.428 rebuild_ks_matrix 128 8.3 0.001 0.001 32.279 33.107 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.019 32.278 33.107 init_scf_loop 11 6.9 0.000 0.001 32.549 32.550 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.030 29.785 multiply_cannon_metrocomm3 30084 15.6 0.101 0.108 15.632 29.354 qs_ot_get_derivative 117 11.6 0.001 0.002 28.257 28.834 prepare_preconditioner 11 7.9 0.000 0.000 28.242 28.307 make_preconditioner 11 8.9 0.000 0.000 28.242 28.307 make_full_inverse_cholesky 11 9.9 0.000 0.000 26.968 27.494 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.096 23.253 apply_single 128 13.6 0.001 0.001 22.096 23.253 qs_ot_get_p 128 10.4 0.001 0.001 21.159 21.930 multiply_cannon_sync_h2d 30084 15.6 18.180 21.918 18.180 21.918 ot_diis_step 117 11.6 0.014 0.015 21.845 21.847 cp_fm_cholesky_invert 11 10.9 16.472 16.484 16.472 16.484 qs_ot_p2m_diag 83 11.4 0.189 0.217 16.173 16.208 make_m2s 5014 13.6 0.087 0.091 14.173 15.897 make_images 5014 14.6 1.179 1.389 13.960 15.678 cp_dbcsr_syevd 83 12.4 0.006 0.006 14.972 14.973 sum_up_and_integrate 128 10.3 0.002 0.004 13.942 13.969 integrate_v_rspace 128 11.3 0.003 0.004 13.882 13.911 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.314 13.347 calculate_rho_elec 128 8.7 0.088 0.105 13.313 13.346 cp_fm_diag_elpa 83 13.4 0.000 0.001 11.812 11.845 cp_fm_diag_elpa_base 83 14.4 11.560 11.656 11.806 11.835 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 11.243 11.682 init_scf_run 11 5.9 0.000 0.001 11.634 11.636 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.634 11.636 multiply_cannon_metrocomm4 27577 15.6 0.099 0.112 3.833 11.087 mp_irecv_dv 69486 16.3 3.633 10.682 3.633 10.682 dbcsr_mm_accdrv_process 62242 16.2 5.563 6.376 10.088 10.629 make_images_data 5014 15.6 0.067 0.078 8.356 10.549 hybrid_alltoall_any 5200 16.5 0.349 1.491 7.143 10.031 pw_transfer 1547 11.6 0.084 0.099 8.517 8.587 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.294 8.372 wfi_extrapolate 11 7.9 0.001 0.001 8.338 8.338 grid_integrate_task_list 128 12.3 7.166 7.530 7.166 7.530 fft_wrap_pw1pw2_140 523 13.2 0.934 0.955 7.358 7.448 density_rs2pw 128 9.7 0.006 0.006 6.995 7.402 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.266 7.083 cp_fm_cholesky_decompose 22 10.9 6.916 7.007 6.916 7.007 calculate_dm_sparse 128 9.5 0.001 0.001 6.495 6.624 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.268 6.275 mp_sum_l 7950 12.9 4.181 6.157 4.181 6.157 fft3d_ps 1291 14.7 2.825 3.015 5.849 5.898 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.345 5.533 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.364 5.434 grid_collocate_task_list 128 9.7 5.005 5.376 5.005 5.376 mp_allgather_i34 2507 14.6 1.935 4.961 1.935 4.961 mp_alltoall_d11v 2415 14.1 4.154 4.725 4.154 4.725 potential_pw2rs 128 12.3 0.015 0.017 4.413 4.429 dbcsr_complete_redistribute 395 12.7 0.785 0.888 3.122 3.956 mp_sum_d 4472 12.1 2.645 3.941 2.645 3.941 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=189.192000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=803.727273, yerr=1.212879 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c078969bf945a475acf73b54de8647dac0b903ea_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420241154048 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 11528896499712 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 93.514757E+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 6755939872 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 961.724416E+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 931531083728 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 58208. MP_Allreduce 11082 999. MP_Sync 87 MP_Alltoall 1712 9388896. MP_SendRecv 7936 75008. MP_ISendRecv 7936 75008. MP_Wait 21820 MP_ISend 11748 275205. MP_IRecv 11748 275205. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.053 0.097 177.076 177.078 qs_mol_dyn_low 1 2.0 0.003 0.004 176.495 176.509 qs_forces 11 3.9 0.010 0.019 175.809 175.811 qs_energies 11 4.9 0.002 0.009 169.272 169.286 scf_env_do_scf 11 5.9 0.001 0.002 151.452 151.454 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 116.706 116.707 velocity_verlet 10 3.0 0.001 0.002 112.067 112.087 dbcsr_multiply_generic 2507 12.6 0.183 0.187 84.078 85.118 qs_scf_new_mos 117 7.6 0.001 0.001 79.822 80.187 qs_scf_loop_do_ot 117 8.6 0.001 0.001 79.821 80.187 ot_scf_mini 117 9.6 0.003 0.004 75.570 75.961 multiply_cannon 2507 13.6 0.510 0.535 64.648 69.662 multiply_cannon_loop 2507 14.6 1.129 1.196 61.753 64.327 ot_mini 117 10.6 0.001 0.001 42.580 42.968 mp_waitall_1 170520 16.6 26.052 36.038 26.052 36.038 init_scf_loop 11 6.9 0.001 0.004 34.650 34.651 prepare_preconditioner 11 7.9 0.000 0.000 30.643 30.687 make_preconditioner 11 8.9 0.000 0.004 30.643 30.687 rebuild_ks_matrix 128 8.3 0.001 0.001 30.136 30.592 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.022 30.136 30.592 make_full_inverse_cholesky 11 9.9 0.000 0.001 28.191 29.664 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.163 27.581 multiply_cannon_metrocomm3 20056 15.6 0.063 0.067 16.028 26.067 multiply_cannon_multrec 20056 15.6 12.868 16.473 22.398 26.022 qs_ot_get_derivative 117 11.6 0.002 0.002 22.853 23.232 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.767 20.873 apply_single 128 13.6 0.001 0.001 19.767 20.873 ot_diis_step 117 11.6 0.018 0.018 19.619 19.619 qs_ot_get_p 128 10.4 0.001 0.002 18.304 18.816 make_m2s 5014 13.6 0.078 0.085 14.593 15.760 make_images 5014 14.6 1.163 1.260 14.361 15.524 multiply_cannon_sync_h2d 20056 15.6 13.633 15.296 13.633 15.296 qs_ot_p2m_diag 83 11.4 0.266 0.273 14.187 14.192 cp_fm_cholesky_invert 11 10.9 14.044 14.052 14.044 14.052 sum_up_and_integrate 128 10.3 0.002 0.003 13.858 13.884 integrate_v_rspace 128 11.3 0.003 0.004 13.799 13.827 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.434 13.465 calculate_rho_elec 128 8.7 0.130 0.144 13.434 13.464 cp_dbcsr_syevd 83 12.4 0.006 0.007 13.139 13.140 init_scf_run 11 5.9 0.000 0.001 12.761 12.761 scf_env_initial_rho_setup 11 6.9 0.002 0.002 12.761 12.761 make_images_data 5014 15.6 0.065 0.074 8.743 10.280 cp_fm_diag_elpa 83 13.4 0.000 0.000 10.005 10.021 cp_fm_diag_elpa_base 83 14.4 9.585 9.731 10.001 10.017 hybrid_alltoall_any 5200 16.5 0.448 2.050 7.529 9.743 multiply_cannon_metrocomm4 17549 15.6 0.065 0.074 3.536 9.679 mp_irecv_dv 50230 16.2 3.407 9.415 3.407 9.415 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 8.914 9.197 dbcsr_mm_accdrv_process 41502 16.2 5.851 6.139 8.987 9.107 pw_transfer 1547 11.6 0.084 0.100 8.722 8.825 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.499 8.606 grid_integrate_task_list 128 12.3 7.371 7.756 7.371 7.756 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 7.720 7.727 fft_wrap_pw1pw2_140 523 13.2 1.066 1.107 7.574 7.683 cp_fm_upper_to_full 105 14.8 5.983 7.582 5.983 7.582 wfi_extrapolate 11 7.9 0.001 0.001 7.391 7.391 density_rs2pw 128 9.7 0.006 0.006 6.834 7.170 cp_fm_cholesky_decompose 22 10.9 7.096 7.134 7.096 7.134 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 6.723 6.857 dbcsr_complete_redistribute 395 12.7 1.251 1.306 4.741 6.593 calculate_dm_sparse 128 9.5 0.001 0.001 5.858 5.971 fft3d_ps 1291 14.7 2.828 3.066 5.805 5.892 grid_collocate_task_list 128 9.7 5.161 5.519 5.161 5.519 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.592 5.450 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.631 5.235 calculate_first_density_matrix 1 7.0 0.001 0.004 5.182 5.184 mp_alltoall_d11v 2415 14.1 4.189 4.569 4.189 4.569 mp_allgather_i34 2507 14.6 1.385 4.552 1.385 4.552 mp_sum_l 7950 12.9 3.042 4.504 3.042 4.504 potential_pw2rs 128 12.3 0.020 0.021 4.299 4.307 transfer_fm_to_dbcsr 11 9.9 0.019 0.023 2.432 4.276 acc_transpose_blocks 20056 15.6 0.102 0.107 2.910 4.098 mp_alltoall_i22 716 14.1 2.011 4.053 2.011 4.053 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.965 4.001 multiply_cannon_metrocomm1 20056 15.6 0.039 0.042 1.656 3.916 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.768 3.774 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=177.078000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=918.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c078969bf945a475acf73b54de8647dac0b903ea_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420243808256 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528908111872 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514772E+12 0.0% 0.0% 100.0% flops max/rank 4.353792E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755942624 0.0% 0.0% 100.0% number of processed stacks 5977344 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.144984E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1143192 MPI messages size (bytes): total size 2.023815E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.770320E+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 319024 36042702848 131072 < size <= 4194304 715736 785529176064 4194304 < size <= 16777216 70320 665379572320 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4020 57949. MP_Allreduce 11127 1081. MP_Sync 87 MP_Alltoall 1712 12503107. MP_SendRecv 5888 75008. MP_ISendRecv 5888 75008. MP_Wait 22442 MP_ISend 14952 244818. MP_IRecv 14952 244818. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.026 0.040 187.000 187.002 qs_mol_dyn_low 1 2.0 0.003 0.004 186.517 186.531 qs_forces 11 3.9 0.004 0.004 186.409 186.419 qs_energies 11 4.9 0.002 0.012 179.362 179.371 scf_env_do_scf 11 5.9 0.002 0.013 162.191 162.202 velocity_verlet 10 3.0 0.001 0.002 123.643 123.645 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 116.005 116.006 dbcsr_multiply_generic 2507 12.6 0.191 0.197 80.219 80.911 qs_scf_new_mos 117 7.6 0.001 0.001 79.837 80.196 qs_scf_loop_do_ot 117 8.6 0.001 0.001 79.836 80.195 ot_scf_mini 117 9.6 0.003 0.004 75.262 75.637 multiply_cannon 2507 13.6 0.552 0.584 55.823 58.286 multiply_cannon_loop 2507 14.6 1.815 1.925 52.187 54.055 init_scf_loop 11 6.9 0.001 0.009 46.057 46.058 ot_mini 117 10.6 0.001 0.001 42.715 43.104 prepare_preconditioner 11 7.9 0.000 0.000 41.934 41.967 make_preconditioner 11 8.9 0.000 0.001 41.934 41.967 make_full_inverse_cholesky 11 9.9 0.012 0.023 35.498 40.548 multiply_cannon_multrec 30084 15.6 13.650 18.623 26.698 31.446 rebuild_ks_matrix 128 8.3 0.001 0.001 29.220 29.591 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.025 29.219 29.590 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.387 26.726 mp_waitall_1 147882 16.7 16.715 26.441 16.715 26.441 qs_ot_get_derivative 117 11.6 0.001 0.002 23.106 23.492 make_m2s 5014 13.6 0.092 0.096 20.000 20.754 make_images 5014 14.6 1.963 2.257 19.695 20.451 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 19.046 19.612 apply_single 128 13.6 0.001 0.001 19.046 19.612 ot_diis_step 117 11.6 0.017 0.018 19.474 19.476 qs_ot_get_p 128 10.4 0.001 0.002 18.776 19.176 cp_fm_upper_to_full 105 14.8 11.527 16.993 11.527 16.993 cp_fm_cholesky_invert 11 10.9 15.891 15.900 15.891 15.900 multiply_cannon_metrocomm3 30084 15.6 0.049 0.051 6.313 14.932 qs_ot_p2m_diag 83 11.4 0.343 0.389 14.659 14.712 sum_up_and_integrate 128 10.3 0.002 0.003 13.922 13.949 integrate_v_rspace 128 11.3 0.003 0.004 13.863 13.891 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.710 13.746 calculate_rho_elec 128 8.7 0.172 0.188 13.709 13.745 cp_dbcsr_syevd 83 12.4 0.005 0.006 13.260 13.262 dbcsr_mm_accdrv_process 62264 16.2 8.511 9.230 12.613 13.109 dbcsr_complete_redistribute 395 12.7 1.482 1.592 9.100 12.977 make_images_data 5014 15.6 0.066 0.072 10.619 12.321 multiply_cannon_sync_h2d 30084 15.6 10.514 11.712 10.514 11.712 copy_fm_to_dbcsr 209 11.7 0.001 0.002 7.761 11.652 init_scf_run 11 5.9 0.000 0.001 11.227 11.228 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.226 11.228 hybrid_alltoall_any 5200 16.5 0.529 2.205 9.655 11.190 transfer_fm_to_dbcsr 11 9.9 0.001 0.014 6.414 10.221 cp_fm_diag_elpa 83 13.4 0.000 0.000 10.085 10.099 cp_fm_diag_elpa_base 83 14.4 9.112 9.430 10.078 10.091 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.694 9.969 mp_alltoall_i22 716 14.1 5.661 9.516 5.661 9.516 pw_transfer 1547 11.6 0.085 0.101 9.116 9.183 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.891 8.964 fft_wrap_pw1pw2_140 523 13.2 1.249 1.290 7.939 8.027 grid_integrate_task_list 128 12.3 7.507 7.880 7.507 7.880 wfi_extrapolate 11 7.9 0.001 0.001 7.592 7.592 cp_fm_cholesky_decompose 22 10.9 7.357 7.457 7.357 7.457 multiply_cannon_metrocomm4 25070 15.6 0.079 0.089 2.797 7.451 mp_irecv_dv 76098 16.2 2.645 7.171 2.645 7.171 density_rs2pw 128 9.7 0.006 0.006 6.770 7.116 calculate_dm_sparse 128 9.5 0.001 0.001 6.313 6.406 fft3d_ps 1291 14.7 2.991 3.069 5.933 5.990 grid_collocate_task_list 128 9.7 5.279 5.688 5.279 5.688 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.503 5.572 mp_alltoall_d11v 2415 14.1 4.918 5.443 4.918 5.443 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.455 4.566 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.437 4.553 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 4.429 4.430 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.236 4.313 potential_pw2rs 128 12.3 0.023 0.024 4.218 4.228 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=187.002000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1085.181818, yerr=13.927202 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c078969bf945a475acf73b54de8647dac0b903ea_performance_tests/19/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 5.820059E+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 1944496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3448.5 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.548603E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 238560 MPI messages size (bytes): total size 1.321104E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.537828E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 112800 59139686400 4194304 < size <= 16777216 104112 545846722560 16777216 < size 20064 716108638608 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8852 52. MP_Alltoall 9584 804353. MP_ISend 39716 2104723. MP_IRecv 39716 2103824. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4003 58187. MP_Allreduce 11085 1167. MP_Sync 86 MP_Alltoall 1700 18828160. MP_SendRecv 3810 122880. MP_ISendRecv 3810 122880. MP_Wait 16000 MP_ISend 10600 423612. MP_IRecv 10600 423612. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.030 169.160 169.160 qs_mol_dyn_low 1 2.0 0.003 0.003 168.731 168.744 qs_forces 11 3.9 0.004 0.004 168.640 168.644 qs_energies 11 4.9 0.002 0.002 161.190 161.195 scf_env_do_scf 11 5.9 0.001 0.001 143.404 143.418 velocity_verlet 10 3.0 0.002 0.003 110.537 110.541 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 107.596 107.597 dbcsr_multiply_generic 2485 12.5 0.187 0.192 71.376 71.820 qs_scf_new_mos 116 7.6 0.001 0.001 70.706 70.824 qs_scf_loop_do_ot 116 8.6 0.001 0.001 70.705 70.823 ot_scf_mini 116 9.6 0.003 0.004 66.328 66.426 multiply_cannon 2485 13.5 0.571 0.596 52.896 56.110 multiply_cannon_loop 2485 14.5 0.801 0.845 49.965 50.829 ot_mini 116 10.6 0.001 0.001 37.019 37.097 init_scf_loop 11 6.9 0.000 0.000 35.657 35.659 prepare_preconditioner 11 7.9 0.000 0.000 31.776 31.803 make_preconditioner 11 8.9 0.000 0.000 31.776 31.803 mp_waitall_1 124680 16.7 24.336 30.466 24.336 30.466 make_full_inverse_cholesky 11 9.9 0.015 0.027 29.557 29.853 rebuild_ks_matrix 127 8.3 0.001 0.001 29.094 29.194 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.017 29.093 29.194 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.513 26.602 multiply_cannon_multrec 9940 15.5 10.338 14.250 17.981 20.993 qs_ot_get_derivative 116 11.6 0.002 0.002 20.213 20.302 multiply_cannon_metrocomm3 9940 15.5 0.024 0.026 12.293 19.470 cp_fm_cholesky_invert 11 10.9 18.064 18.070 18.064 18.070 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 16.819 16.978 apply_single 127 13.6 0.001 0.001 16.818 16.978 qs_ot_get_p 127 10.4 0.001 0.001 16.737 16.833 ot_diis_step 116 11.6 0.020 0.020 16.737 16.737 make_m2s 4970 13.5 0.064 0.068 14.679 15.482 make_images 4970 14.5 2.157 2.587 14.374 15.172 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.185 14.213 calculate_rho_elec 127 8.7 0.253 0.264 14.184 14.212 sum_up_and_integrate 127 10.3 0.002 0.002 14.112 14.159 integrate_v_rspace 127 11.3 0.004 0.004 14.052 14.100 qs_ot_p2m_diag 82 11.4 0.489 0.495 13.020 13.036 cp_dbcsr_syevd 82 12.4 0.005 0.006 11.865 11.866 multiply_cannon_sync_h2d 9940 15.5 10.787 11.024 10.787 11.024 init_scf_run 11 5.9 0.000 0.001 10.831 10.831 scf_env_initial_rho_setup 11 6.9 0.015 0.018 10.831 10.831 pw_transfer 1535 11.6 0.084 0.092 9.888 9.919 hybrid_alltoall_any 5155 16.4 0.835 3.754 8.302 9.728 fft_wrap_pw1pw2 1281 12.7 0.010 0.010 9.666 9.705 make_images_data 4970 15.5 0.055 0.064 8.438 9.693 cp_fm_diag_elpa 82 13.4 0.000 0.000 9.033 9.044 cp_fm_diag_elpa_base 82 14.4 8.792 8.875 9.027 9.039 fft_wrap_pw1pw2_140 519 13.2 1.669 1.703 8.562 8.603 cp_fm_cholesky_decompose 22 10.9 7.916 8.014 7.916 8.014 dbcsr_mm_accdrv_process 20590 16.1 3.205 4.317 7.295 8.012 grid_integrate_task_list 127 12.3 7.750 7.967 7.750 7.967 qs_ot_get_derivative_diag 76 12.4 0.002 0.003 7.898 7.962 wfi_extrapolate 11 7.9 0.001 0.001 7.355 7.355 density_rs2pw 127 9.7 0.005 0.005 6.932 7.249 multiply_cannon_metrocomm1 9940 15.5 0.030 0.031 4.376 7.101 calculate_dm_sparse 127 9.5 0.001 0.001 6.115 6.184 fft3d_ps 1281 14.7 3.121 3.223 6.132 6.181 dbcsr_complete_redistribute 393 12.7 2.176 2.317 5.490 5.763 grid_collocate_task_list 127 9.7 5.534 5.761 5.534 5.761 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.410 5.422 qs_energies_init_hamiltonians 11 5.9 0.006 0.008 5.375 5.376 mp_alltoall_d11v 2401 14.1 4.651 4.836 4.651 4.836 mp_allgather_i34 2485 14.5 1.104 4.519 1.104 4.519 potential_pw2rs 127 12.3 0.026 0.026 4.329 4.333 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.131 4.181 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.662 3.983 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.594 3.919 multiply_cannon_metrocomm4 7455 15.5 0.026 0.028 1.730 3.728 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.602 3.662 mp_irecv_dv 28618 15.9 1.691 3.655 1.691 3.655 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.517 3.555 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.421 3.517 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.480 3.491 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=169.160000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1458.909091, yerr=23.418930 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c078969bf945a475acf73b54de8647dac0b903ea_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456039424 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613072052224 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176077312 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176077312 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.233020E+12 0.0% 0.0% 100.0% flops max/rank 11.786061E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383904 0.0% 0.0% 100.0% number of processed stacks 1980288 0.0% 0.0% 100.0% average stack size 0.0 0.0 3437.1 marketing flops 145.650931E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 3.203559E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101160 MPI messages size (bytes): total size 1.144970E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.318403E+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 45648 35433480192 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592466352 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 58558. MP_Allreduce 11184 1500. MP_Sync 88 MP_Alltoall 1724 36993632. MP_SendRecv 1806 218624. MP_ISendRecv 1806 218624. MP_Wait 9876 MP_ISend 6456 1080169. MP_IRecv 6456 1080169. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.047 0.055 295.260 295.269 qs_mol_dyn_low 1 2.0 0.003 0.004 294.624 294.639 qs_forces 11 3.9 0.005 0.005 294.519 294.521 qs_energies 11 4.9 0.002 0.002 285.574 285.583 scf_env_do_scf 11 5.9 0.001 0.001 262.795 262.801 velocity_verlet 10 3.0 0.020 0.021 213.308 213.315 scf_env_do_scf_inner_loop 118 6.6 0.004 0.009 136.745 136.748 init_scf_loop 11 6.9 0.000 0.000 125.775 125.779 prepare_preconditioner 11 7.9 0.000 0.000 120.906 120.923 make_preconditioner 11 8.9 0.000 0.000 120.906 120.923 make_full_inverse_cholesky 11 9.9 0.038 0.039 96.564 118.076 qs_scf_new_mos 118 7.6 0.001 0.001 91.285 91.346 qs_scf_loop_do_ot 118 8.6 0.001 0.001 91.284 91.345 ot_scf_mini 118 9.6 0.004 0.004 86.339 86.343 dbcsr_multiply_generic 2529 12.6 0.270 0.279 82.873 83.339 cp_fm_upper_to_full 106 14.8 54.276 77.651 54.276 77.651 multiply_cannon 2529 13.6 0.684 0.734 59.433 61.008 multiply_cannon_loop 2529 14.6 1.064 1.082 55.348 56.594 ot_mini 118 10.6 0.001 0.001 44.609 44.612 dbcsr_complete_redistribute 397 12.7 4.002 4.049 30.283 43.694 copy_fm_to_dbcsr 210 11.7 0.001 0.002 26.775 40.145 transfer_fm_to_dbcsr 11 9.9 0.030 0.031 24.299 37.449 mp_alltoall_i22 720 14.1 22.064 35.541 22.064 35.541 rebuild_ks_matrix 129 8.3 0.001 0.001 35.168 35.173 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.018 35.167 35.172 cp_fm_cholesky_invert 11 10.9 33.263 33.269 33.263 33.269 mp_waitall_1 104580 16.8 28.128 32.799 28.128 32.799 qs_ks_update_qs_env 129 7.6 0.001 0.001 32.639 32.654 qs_ot_get_p 129 10.4 0.001 0.001 26.424 26.449 qs_ot_get_derivative 118 11.6 0.002 0.002 25.078 25.085 qs_ot_p2m_diag 84 11.4 0.891 0.904 22.284 22.312 cp_dbcsr_syevd 84 12.4 0.006 0.006 20.461 20.462 make_m2s 5058 13.6 0.074 0.076 18.853 20.359 multiply_cannon_metrocomm3 10116 15.6 0.025 0.025 18.884 20.088 make_images 5058 14.6 3.099 3.301 18.369 19.875 ot_diis_step 118 11.6 0.022 0.022 19.506 19.506 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 18.789 18.876 apply_single 129 13.6 0.001 0.001 18.789 18.876 multiply_cannon_multrec 10116 15.6 10.471 12.379 18.358 18.426 qs_rho_update_rho_low 129 7.7 0.001 0.001 17.543 17.562 calculate_rho_elec 129 8.7 0.482 0.483 17.543 17.561 cp_fm_diag_elpa 84 13.4 0.000 0.000 17.214 17.215 cp_fm_diag_elpa_base 84 14.4 12.673 14.349 17.209 17.210 sum_up_and_integrate 129 10.3 0.002 0.002 16.668 16.761 integrate_v_rspace 129 11.3 0.004 0.005 16.604 16.698 multiply_cannon_sync_h2d 10116 15.6 14.400 14.440 14.400 14.440 pw_transfer 1559 11.6 0.096 0.097 13.374 13.387 hybrid_alltoall_any 5245 16.5 1.328 3.084 10.782 13.150 fft_wrap_pw1pw2 1301 12.7 0.011 0.012 13.133 13.145 make_images_data 5058 15.6 0.064 0.070 10.542 12.773 init_scf_run 11 5.9 0.000 0.001 12.340 12.340 scf_env_initial_rho_setup 11 6.9 0.004 0.005 12.340 12.340 fft_wrap_pw1pw2_140 527 13.2 3.237 3.271 11.682 11.692 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 9.889 9.897 dbcsr_mm_accdrv_process 20934 16.1 4.288 6.185 7.635 9.634 wfi_extrapolate 11 7.9 0.001 0.001 9.025 9.025 cp_fm_cholesky_decompose 22 10.9 9.000 9.024 9.000 9.024 grid_integrate_task_list 129 12.3 8.621 8.809 8.621 8.809 density_rs2pw 129 9.7 0.005 0.005 8.470 8.526 qs_energies_init_hamiltonians 11 5.9 0.015 0.028 8.240 8.241 fft3d_ps 1301 14.7 4.010 4.032 7.562 7.601 mp_alltoall_d11v 2429 14.1 6.965 7.086 6.965 7.086 calculate_dm_sparse 129 9.5 0.001 0.001 6.792 6.868 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.543 6.592 grid_collocate_task_list 129 9.7 6.467 6.539 6.467 6.539 copy_dbcsr_to_fm 187 11.8 0.004 0.004 6.151 6.263 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=295.269000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2918.818182, yerr=149.195307 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c078969bf945a475acf73b54de8647dac0b903ea_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 1.260929E+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 0.031 0.126 88.195 88.196 qs_energies 1 2.0 0.000 0.001 87.449 87.466 ls_scf 1 3.0 0.000 0.002 86.538 86.552 dbcsr_multiply_generic 111 6.7 0.015 0.016 75.278 75.440 multiply_cannon 111 7.7 0.018 0.020 58.500 59.597 multiply_cannon_loop 111 8.7 0.230 0.247 55.065 56.276 ls_scf_main 1 4.0 0.000 0.006 52.551 52.563 density_matrix_trs4 2 5.0 0.002 0.004 46.936 47.022 ls_scf_init_scf 1 4.0 0.000 0.004 30.835 30.836 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.678 29.729 mp_waitall_1 11031 10.9 23.135 28.296 23.135 28.296 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 27.567 27.584 multiply_cannon_multrec 2664 9.7 8.152 9.029 15.630 17.536 multiply_cannon_sync_h2d 2664 9.7 13.234 15.637 13.234 15.637 make_m2s 222 7.7 0.009 0.012 13.148 13.584 make_images 222 8.7 0.100 0.110 13.125 13.563 multiply_cannon_metrocomm1 2664 9.7 0.011 0.012 10.106 13.326 multiply_cannon_metrocomm3 2664 9.7 0.010 0.012 5.679 9.903 make_images_data 222 9.7 0.005 0.006 7.687 8.240 dbcsr_mm_accdrv_process 4760 10.4 0.599 0.724 7.095 8.115 hybrid_alltoall_any 227 10.6 0.218 1.850 6.636 7.991 dbcsr_mm_accdrv_process_sort 4760 11.4 6.297 7.221 6.297 7.221 calculate_norms 4752 9.8 5.499 6.237 5.499 6.237 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.093 5.221 mp_sum_l 887 5.1 3.144 4.514 3.144 4.514 make_images_sizes 222 9.7 0.000 0.000 0.692 3.877 mp_alltoall_i44 222 10.7 0.692 3.876 0.692 3.876 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.041 3.574 mp_irecv_dv 6231 10.9 2.024 3.545 2.024 3.545 arnoldi_extremal 4 6.8 0.000 0.002 3.366 3.388 arnoldi_normal_ev 4 7.8 0.001 0.006 3.365 3.387 compute_matrix_preconditioner 1 6.0 0.000 0.002 3.364 3.369 build_subspace 16 8.4 0.009 0.014 3.255 3.258 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.311 3.255 ls_scf_post 1 4.0 0.001 0.019 3.152 3.165 ls_scf_store_result 1 5.0 0.000 0.000 2.898 2.950 dbcsr_special_finalize 555 9.7 0.006 0.007 2.406 2.831 dbcsr_merge_single_wm 555 10.7 0.466 0.610 2.397 2.821 acc_transpose_blocks 2664 9.7 0.018 0.020 2.537 2.760 make_images_pack 222 9.7 2.213 2.636 2.215 2.637 acc_transpose_blocks_kernels 2664 10.7 0.035 0.040 2.395 2.611 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.338 2.590 jit_kernel_transpose 1 13.0 2.360 2.575 2.360 2.575 dbcsr_sort_data 658 11.4 2.188 2.516 2.188 2.516 dbcsr_matrix_vector_mult_local 304 10.0 2.069 2.479 2.071 2.481 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.329 2.413 buffer_matrices_ensure_size 222 8.7 1.762 2.119 1.762 2.119 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.780 1.782 rebuild_ks_matrix 3 7.3 0.000 0.000 1.770 1.772 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.006 1.770 1.772 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=88.196000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1132.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c078969bf945a475acf73b54de8647dac0b903ea_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 2.165567E+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 0.032 0.071 89.666 89.666 qs_energies 1 2.0 0.000 0.000 89.140 89.144 ls_scf 1 3.0 0.000 0.000 87.798 87.802 dbcsr_multiply_generic 111 6.7 0.015 0.016 74.057 74.321 multiply_cannon 111 7.7 0.028 0.044 52.712 56.682 ls_scf_main 1 4.0 0.000 0.000 54.297 54.303 multiply_cannon_loop 111 8.7 0.136 0.147 49.875 52.990 density_matrix_trs4 2 5.0 0.002 0.003 48.677 48.888 ls_scf_init_scf 1 4.0 0.000 0.000 29.836 29.838 mp_waitall_1 9105 10.9 20.512 28.878 20.512 28.878 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.626 28.736 multiply_cannon_multrec 1332 9.7 13.371 17.283 22.745 27.765 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.255 26.266 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.363 19.643 make_m2s 222 7.7 0.007 0.008 15.024 15.598 make_images 222 8.7 1.376 1.702 14.994 15.569 dbcsr_mm_accdrv_process 4041 10.4 0.333 0.512 8.974 10.569 dbcsr_mm_accdrv_process_sort 4041 11.4 8.493 10.039 8.493 10.039 make_images_data 222 9.7 0.004 0.005 8.679 9.543 hybrid_alltoall_any 227 10.6 0.543 2.566 8.013 9.388 mp_sum_l 887 5.1 4.862 7.676 4.862 7.676 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.229 7.647 mp_irecv_dv 3311 11.0 3.208 7.591 3.208 7.591 calculate_norms 2376 9.8 6.059 6.769 6.059 6.769 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.686 6.177 multiply_cannon_sync_h2d 1332 9.7 4.858 5.969 4.858 5.969 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.876 5.102 arnoldi_extremal 4 6.8 0.000 0.000 4.598 4.611 arnoldi_normal_ev 4 7.8 0.001 0.004 4.598 4.611 build_subspace 16 8.4 0.014 0.021 4.345 4.349 ls_scf_post 1 4.0 0.000 0.000 3.665 3.668 ls_scf_store_result 1 5.0 0.000 0.000 3.395 3.488 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.131 3.360 dbcsr_matrix_vector_mult_local 304 10.0 2.749 3.231 2.751 3.233 mp_allgather_i34 111 8.7 0.807 3.031 0.807 3.031 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.120 2.754 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.598 2.688 dbcsr_data_new 4174 10.1 2.118 2.396 2.118 2.396 make_images_pack 222 9.7 1.822 2.125 1.825 2.128 dbcsr_sort_data 436 11.2 1.830 2.093 1.830 2.093 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.853 1.855 rebuild_ks_matrix 3 7.3 0.000 0.000 1.840 1.842 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.840 1.842 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=89.666000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1775.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c078969bf945a475acf73b54de8647dac0b903ea_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 2.931429E+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 0.062 0.109 93.915 93.916 qs_energies 1 2.0 0.002 0.016 93.173 93.197 ls_scf 1 3.0 0.001 0.008 91.575 91.593 dbcsr_multiply_generic 111 6.7 0.016 0.016 75.892 76.204 ls_scf_main 1 4.0 0.001 0.029 57.007 57.013 multiply_cannon 111 7.7 0.038 0.073 52.596 56.554 multiply_cannon_loop 111 8.7 0.118 0.131 49.816 53.861 density_matrix_trs4 2 5.0 0.003 0.011 51.095 51.256 mp_waitall_1 7281 11.0 23.928 33.786 23.928 33.786 ls_scf_init_scf 1 4.0 0.001 0.004 30.877 30.888 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.369 29.459 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.895 26.911 multiply_cannon_multrec 888 9.7 12.723 15.468 21.404 24.730 multiply_cannon_metrocomm3 888 9.7 0.004 0.005 11.272 23.149 make_m2s 222 7.7 0.006 0.007 16.596 17.370 make_images 222 8.7 1.590 1.853 16.558 17.330 make_images_data 222 9.7 0.004 0.005 9.609 10.855 hybrid_alltoall_any 227 10.6 0.642 2.944 9.153 10.596 dbcsr_mm_accdrv_process 3754 10.4 0.307 0.516 8.207 9.446 dbcsr_mm_accdrv_process_sort 3754 11.4 7.754 8.930 7.754 8.930 mp_sum_l 887 5.1 4.994 8.540 4.994 8.540 multiply_cannon_sync_h2d 888 9.7 6.061 7.367 6.061 7.367 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.522 7.019 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.872 6.974 mp_irecv_dv 2335 11.1 2.506 6.956 2.506 6.956 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.696 6.888 arnoldi_extremal 4 6.8 0.000 0.000 5.080 5.096 arnoldi_normal_ev 4 7.8 0.003 0.009 5.080 5.096 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.803 5.009 build_subspace 16 8.4 0.014 0.020 4.782 4.790 calculate_norms 1584 9.8 4.357 4.776 4.357 4.776 mp_allgather_i34 111 8.7 0.891 3.867 0.891 3.867 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.435 3.773 ls_scf_post 1 4.0 0.002 0.015 3.690 3.704 dbcsr_matrix_vector_mult_local 304 10.0 3.026 3.607 3.028 3.610 ls_scf_store_result 1 5.0 0.000 0.000 3.438 3.512 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.839 2.933 dbcsr_data_new 4116 9.9 2.093 2.452 2.093 2.452 make_images_sizes 222 9.7 0.000 0.000 1.118 2.424 mp_alltoall_i44 222 10.7 1.117 2.423 1.117 2.423 dbcsr_sort_data 325 11.1 1.889 2.161 1.889 2.161 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.125 2.138 rebuild_ks_matrix 3 7.3 0.000 0.000 2.107 2.121 qs_ks_build_kohn_sham_matrix 3 8.3 0.009 0.068 2.107 2.121 dbcsr_finalize 304 7.8 0.026 0.032 1.617 1.900 make_images_pack 222 9.7 1.649 1.889 1.652 1.892 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=93.916000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2244.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c078969bf945a475acf73b54de8647dac0b903ea_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 3.336315E+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 0.041 0.092 96.782 96.783 qs_energies 1 2.0 0.000 0.000 96.171 96.176 ls_scf 1 3.0 0.000 0.000 94.463 94.470 dbcsr_multiply_generic 111 6.7 0.017 0.017 78.231 78.493 ls_scf_main 1 4.0 0.000 0.000 58.688 58.689 multiply_cannon 111 7.7 0.042 0.084 51.716 55.783 density_matrix_trs4 2 5.0 0.002 0.003 52.584 52.700 multiply_cannon_loop 111 8.7 0.153 0.168 46.671 49.514 ls_scf_init_scf 1 4.0 0.000 0.000 32.540 32.543 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.330 31.396 mp_waitall_1 6369 11.0 22.474 29.323 22.474 29.323 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.823 28.834 multiply_cannon_multrec 1332 9.7 14.209 16.938 22.221 25.685 make_m2s 222 7.7 0.007 0.009 21.123 22.514 make_images 222 8.7 3.150 3.617 21.072 22.466 multiply_cannon_metrocomm3 1332 9.7 0.004 0.004 9.078 16.319 make_images_data 222 9.7 0.004 0.005 11.722 13.398 hybrid_alltoall_any 227 10.6 0.797 3.735 11.153 12.929 dbcsr_mm_accdrv_process 3641 10.4 0.303 0.482 7.636 9.166 dbcsr_mm_accdrv_process_sort 3641 11.4 7.194 8.673 7.194 8.673 mp_sum_l 887 5.1 3.799 6.802 3.799 6.802 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.106 6.225 mp_irecv_dv 3229 10.9 2.081 6.145 2.081 6.145 multiply_cannon_sync_h2d 1332 9.7 5.473 5.997 5.473 5.997 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.915 5.342 arnoldi_extremal 4 6.8 0.000 0.000 5.185 5.201 arnoldi_normal_ev 4 7.8 0.001 0.005 5.185 5.201 build_subspace 16 8.4 0.014 0.021 4.852 4.859 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.510 4.668 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.510 4.653 calculate_norms 2376 9.8 4.179 4.529 4.179 4.529 mp_allgather_i34 111 8.7 2.149 4.481 2.149 4.481 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.579 3.877 dbcsr_matrix_vector_mult_local 304 10.0 3.197 3.700 3.200 3.702 dbcsr_sort_data 658 11.4 3.083 3.439 3.083 3.439 ls_scf_post 1 4.0 0.000 0.000 3.235 3.239 dbcsr_special_finalize 555 9.7 0.006 0.007 2.835 3.206 dbcsr_merge_single_wm 555 10.7 0.539 0.662 2.826 3.198 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.025 3.069 ls_scf_store_result 1 5.0 0.000 0.000 2.977 3.038 dbcsr_data_release 10477 10.7 1.604 2.422 1.604 2.422 dbcsr_finalize 304 7.8 0.049 0.061 1.807 1.982 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=96.783000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2710.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c078969bf945a475acf73b54de8647dac0b903ea_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 4.757172E+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 0.062 92.930 92.930 qs_energies 1 2.0 0.000 0.000 92.125 92.130 ls_scf 1 3.0 0.000 0.000 90.072 90.076 dbcsr_multiply_generic 111 6.7 0.017 0.019 71.462 71.661 ls_scf_main 1 4.0 0.000 0.000 57.048 57.049 multiply_cannon 111 7.7 0.070 0.130 52.910 55.762 multiply_cannon_loop 111 8.7 0.088 0.095 50.310 52.067 density_matrix_trs4 2 5.0 0.002 0.003 50.093 50.155 ls_scf_init_scf 1 4.0 0.000 0.000 29.575 29.576 mp_waitall_1 5436 11.0 24.853 28.579 24.853 28.579 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.316 28.348 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.024 26.030 multiply_cannon_multrec 444 9.7 13.723 16.270 20.881 22.223 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.003 15.664 make_m2s 222 7.7 0.005 0.005 13.748 14.730 make_images 222 8.7 2.048 2.489 13.680 14.660 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.193 14.495 hybrid_alltoall_any 227 10.6 0.802 3.831 8.338 10.129 make_images_data 222 9.7 0.003 0.004 8.476 9.834 multiply_cannon_sync_h2d 444 9.7 6.740 8.168 6.740 8.168 dbcsr_mm_accdrv_process 3003 10.4 0.368 0.431 6.849 7.992 dbcsr_mm_accdrv_process_sort 3003 11.4 6.470 7.561 6.470 7.561 arnoldi_extremal 4 6.8 0.000 0.000 5.844 5.852 arnoldi_normal_ev 4 7.8 0.002 0.005 5.844 5.852 build_subspace 16 8.4 0.015 0.020 5.450 5.463 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.416 4.638 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.178 4.373 mp_sum_l 887 5.1 2.806 4.360 2.806 4.360 dbcsr_matrix_vector_mult_local 304 10.0 3.710 4.177 3.712 4.180 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.552 3.779 mp_allgather_i34 111 8.7 1.175 3.774 1.175 3.774 mp_irecv_dv 1241 11.2 1.540 3.750 1.540 3.750 calculate_norms 792 9.8 3.603 3.706 3.603 3.706 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.529 3.585 ls_scf_post 1 4.0 0.000 0.000 3.448 3.453 make_images_sizes 222 9.7 0.000 0.000 0.863 3.360 mp_alltoall_i44 222 10.7 0.863 3.360 0.863 3.360 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.993 3.291 ls_scf_store_result 1 5.0 0.000 0.000 3.238 3.274 dbcsr_finalize 304 7.8 0.062 0.077 2.198 2.268 dbcsr_data_new 4608 9.7 1.792 2.262 1.792 2.262 dbcsr_merge_all 275 8.9 0.480 0.522 2.055 2.128 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.039 2.040 qs_energies_init_hamiltonians 1 3.0 0.021 0.029 2.037 2.037 rebuild_ks_matrix 3 7.3 0.000 0.000 2.008 2.008 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.008 2.008 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=92.930000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3772.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c078969bf945a475acf73b54de8647dac0b903ea_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 8.886608E+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 0.085 0.111 108.687 108.687 qs_energies 1 2.0 0.000 0.000 107.231 107.234 ls_scf 1 3.0 0.000 0.000 104.199 104.202 dbcsr_multiply_generic 111 6.7 0.024 0.027 77.657 77.794 ls_scf_main 1 4.0 0.000 0.000 64.049 64.049 density_matrix_trs4 2 5.0 0.002 0.003 55.000 55.063 multiply_cannon 111 7.7 0.173 0.225 51.352 53.537 multiply_cannon_loop 111 8.7 0.099 0.100 48.320 48.811 ls_scf_init_scf 1 4.0 0.000 0.000 36.328 36.329 ls_scf_init_matrix_S 1 5.0 0.000 0.000 34.758 34.769 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 32.034 32.045 mp_waitall_1 4527 11.1 22.433 26.047 22.433 26.047 make_m2s 222 7.7 0.005 0.005 22.721 23.837 make_images 222 8.7 3.584 3.899 22.612 23.727 multiply_cannon_multrec 444 9.7 17.917 18.567 22.591 23.299 hybrid_alltoall_any 227 10.6 1.655 3.623 12.772 15.810 make_images_data 222 9.7 0.004 0.004 13.068 15.247 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.782 13.568 multiply_cannon_sync_h2d 444 9.7 8.778 8.844 8.778 8.844 arnoldi_extremal 4 6.8 0.000 0.000 7.456 7.468 arnoldi_normal_ev 4 7.8 0.003 0.009 7.455 7.468 build_subspace 16 8.4 0.026 0.036 6.896 6.911 dbcsr_matrix_vector_mult 304 9.0 0.017 0.033 5.514 5.668 dbcsr_matrix_vector_mult_local 304 10.0 5.073 5.389 5.075 5.392 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.180 5.275 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.882 5.139 dbcsr_mm_accdrv_process 1814 10.4 0.267 0.358 4.500 4.622 dbcsr_mm_accdrv_process_sort 1814 11.4 4.164 4.292 4.164 4.292 ls_scf_post 1 4.0 0.000 0.000 3.822 3.825 make_images_sizes 222 9.7 0.000 0.000 1.494 3.773 mp_alltoall_i44 222 10.7 1.494 3.773 1.494 3.773 mp_allgather_i34 111 8.7 1.126 3.668 1.126 3.668 ls_scf_store_result 1 5.0 0.000 0.000 3.525 3.561 calculate_norms 792 9.8 3.241 3.321 3.241 3.321 compute_matrix_preconditioner 1 6.0 0.002 0.002 3.242 3.248 dbcsr_finalize 304 7.8 0.082 0.090 3.073 3.188 qs_energies_init_hamiltonians 1 3.0 0.004 0.024 3.001 3.001 dbcsr_merge_all 275 8.9 0.883 0.922 2.857 2.966 dbcsr_complete_redistribute 5 7.6 1.425 1.467 2.740 2.846 acc_transpose_blocks 444 9.7 0.003 0.003 2.450 2.770 acc_transpose_blocks_kernels 444 10.7 0.006 0.006 2.366 2.687 jit_kernel_transpose 1 13.0 2.360 2.681 2.360 2.681 matrix_ls_to_qs 2 6.0 0.000 0.000 2.410 2.527 dbcsr_sort_data 325 11.1 2.439 2.512 2.439 2.512 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.460 2.462 rebuild_ks_matrix 3 7.3 0.000 0.000 2.394 2.396 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.394 2.396 dbcsr_new_transposed 4 7.5 0.241 0.250 2.338 2.357 dbcsr_data_new 6591 9.6 1.810 2.258 1.810 2.258 dbcsr_frobenius_norm 74 6.6 2.056 2.138 2.191 2.219 dbcsr_add_d 103 6.2 0.000 0.000 2.124 2.207 dbcsr_add_anytype 103 7.2 0.858 0.890 2.124 2.207 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=108.687000, 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/c078969bf945a475acf73b54de8647dac0b903ea_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 594.288640E+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 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 0.032 0.192 243.105 243.120 qs_mol_dyn_low 1 2.0 0.004 0.019 241.368 241.418 qs_forces 5 3.8 0.006 0.048 241.206 241.238 qs_energies 5 4.8 0.004 0.029 238.114 238.141 scf_env_do_scf 5 5.8 0.000 0.002 222.763 222.766 scf_env_do_scf_inner_loop 105 6.6 0.004 0.028 194.426 194.445 qs_scf_new_mos 105 7.6 0.000 0.001 151.353 151.502 qs_scf_loop_do_ot 105 8.6 0.001 0.001 151.353 151.502 ot_scf_mini 105 9.6 0.003 0.004 141.480 141.613 dbcsr_multiply_generic 1445 12.2 0.126 0.143 134.167 134.629 multiply_cannon 1445 13.2 0.276 0.287 114.568 116.401 multiply_cannon_loop 1445 14.2 2.839 2.984 112.559 113.882 velocity_verlet 4 3.0 0.005 0.022 109.127 109.128 ot_mini 105 10.6 0.001 0.002 60.925 61.036 qs_ot_get_p 112 10.4 0.001 0.002 50.550 50.849 multiply_cannon_multrec 69360 15.2 29.663 35.025 39.559 44.633 mp_waitall_1 488190 16.1 36.148 44.307 36.148 44.307 qs_ot_p2m_diag 40 11.0 0.020 0.030 39.328 39.437 qs_ot_get_derivative 55 11.6 0.001 0.001 39.014 39.158 cp_dbcsr_syevd 40 12.0 0.002 0.004 36.005 36.006 multiply_cannon_metrocomm3 69360 15.2 0.205 0.216 25.773 33.462 multiply_cannon_sync_h2d 69360 15.2 29.093 33.120 29.093 33.120 cp_fm_syevd 40 13.0 0.000 0.001 30.560 30.696 rebuild_ks_matrix 110 8.4 0.000 0.000 29.977 30.159 qs_ks_build_kohn_sham_matrix 110 9.4 0.018 0.042 29.977 30.158 init_scf_loop 7 6.6 0.001 0.005 28.294 28.299 qs_ks_update_qs_env 112 7.6 0.001 0.001 27.520 27.681 cp_fm_redistribute_end 40 14.0 12.822 25.582 12.829 25.585 cp_fm_syevd_base 40 14.0 12.744 25.505 12.744 25.505 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 23.221 23.459 apply_single 62 13.6 0.000 0.000 23.221 23.459 prepare_preconditioner 7 7.6 0.000 0.000 23.198 23.233 make_preconditioner 7 8.6 0.000 0.003 23.197 23.233 ot_new_cg_direction 55 11.6 0.001 0.003 21.183 21.184 qs_rho_update_rho_low 110 7.6 0.001 0.002 19.306 19.645 calculate_rho_elec 110 8.6 0.030 0.032 19.305 19.644 make_full_inverse_cholesky 7 9.6 0.000 0.001 15.747 15.802 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.738 14.891 density_rs2pw 110 9.6 0.005 0.006 13.519 14.061 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 13.717 13.807 mp_sum_l 4764 12.2 12.206 12.892 12.206 12.892 init_scf_run 5 5.8 0.000 0.001 12.410 12.411 scf_env_initial_rho_setup 5 6.8 0.002 0.004 12.409 12.411 pw_transfer 1645 12.4 0.080 0.103 11.756 12.008 fft_wrap_pw1pw2 1425 13.5 0.012 0.015 11.617 11.873 calculate_dm_sparse 110 9.5 0.000 0.001 11.480 11.605 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 10.861 10.948 cp_fm_cholesky_invert 7 10.6 10.867 10.878 10.867 10.878 transfer_rs2pw 445 10.6 0.007 0.008 10.124 10.782 dbcsr_mm_accdrv_process 154766 15.8 6.244 6.442 9.763 10.586 qs_vxc_create 110 10.4 0.003 0.027 10.360 10.421 fft_wrap_pw1pw2_240 915 15.0 0.830 0.917 10.044 10.273 check_diag 80 13.5 8.685 8.944 9.934 10.085 sum_up_and_integrate 60 10.3 0.001 0.003 9.551 9.562 integrate_v_rspace 60 11.3 0.002 0.026 9.533 9.544 fft3d_pb 915 16.0 2.400 2.747 8.228 8.557 multiply_cannon_metrocomm1 69360 15.2 0.099 0.106 4.743 8.366 acc_transpose_blocks 69360 15.2 0.364 0.382 7.440 8.081 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.003 7.657 7.675 make_m2s 2890 13.2 0.079 0.087 6.981 7.587 xc_rho_set_and_dset_create 110 12.4 0.076 0.096 7.273 7.540 make_images 2890 14.2 0.240 0.259 6.874 7.481 calculate_first_density_matrix 1 7.0 0.001 0.013 7.276 7.292 make_full_single_inverse 7 9.6 0.001 0.001 7.171 7.206 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 7.082 7.144 xc_vxc_pw_create 60 11.3 0.039 0.051 6.924 6.983 xc_pw_derive 510 13.4 0.005 0.007 6.114 6.192 potential_pw2rs 60 12.3 0.002 0.003 5.722 5.783 mp_alltoall_z22v 2340 17.7 5.258 5.554 5.258 5.554 mp_waitany 7680 13.5 4.593 5.322 4.593 5.322 acc_transpose_blocks_kernels 69360 16.2 0.857 0.898 4.659 5.228 transfer_rs2pw_30 110 11.6 1.287 1.363 4.356 4.913 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=243.120000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=563.200000, yerr=3.487119 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: c078969bf945a475acf73b54de8647dac0b903ea Summary: empty Status: OK