=== 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: bbd14a5fe6b002419d127c8ec63e3062d934e058 ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, # Cray-FFTW 3.3.8.10, COSMA 2.6.6, ELPA 2023.05.001, # HDF5 1.14.2, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.2.2, LIBVORI 220621, LIBXSMM 1.17, # PLUMED 2.9.0, SIRIUS 7.4.3, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Last update: 04.10.2023 # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2023.05.001 USE_HDF5 := 1.14.2 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.9.0 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a endif ifneq ($(USE_HDF5),) USE_HDF5 := $(strip $(USE_HDF5)) HDF5_INC := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/include HDF5_LIB := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/lib CFLAGS += -I$(HDF5_INC) DFLAGS += -D__HDF5 LIBS += $(HDF5_LIB)/libhdf5_fortran.a LIBS += $(HDF5_LIB)/libhdf5_hl.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bbd14a5fe6b002419d127c8ec63e3062d934e058_performance_tests/01 job id: 50205159 --- 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/bbd14a5fe6b002419d127c8ec63e3062d934e058_performance_tests/02 job id: 50205160 --- 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/bbd14a5fe6b002419d127c8ec63e3062d934e058_performance_tests/03 job id: 50205161 --- 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/bbd14a5fe6b002419d127c8ec63e3062d934e058_performance_tests/04 job id: 50205162 --- 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/bbd14a5fe6b002419d127c8ec63e3062d934e058_performance_tests/05 job id: 50205163 --- 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/bbd14a5fe6b002419d127c8ec63e3062d934e058_performance_tests/06 job id: 50205164 --- 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/bbd14a5fe6b002419d127c8ec63e3062d934e058_performance_tests/07 job id: 50205165 --- 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/bbd14a5fe6b002419d127c8ec63e3062d934e058_performance_tests/08 job id: 50205166 --- 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/bbd14a5fe6b002419d127c8ec63e3062d934e058_performance_tests/09 job id: 50205167 --- 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/bbd14a5fe6b002419d127c8ec63e3062d934e058_performance_tests/10 job id: 50205168 --- 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/bbd14a5fe6b002419d127c8ec63e3062d934e058_performance_tests/11 job id: 50205169 --- 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/bbd14a5fe6b002419d127c8ec63e3062d934e058_performance_tests/12 job id: 50205170 --- 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/bbd14a5fe6b002419d127c8ec63e3062d934e058_performance_tests/13 job id: 50205171 --- 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/bbd14a5fe6b002419d127c8ec63e3062d934e058_performance_tests/14 job id: 50205172 --- 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/bbd14a5fe6b002419d127c8ec63e3062d934e058_performance_tests/15 job id: 50205173 --- 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/bbd14a5fe6b002419d127c8ec63e3062d934e058_performance_tests/16 job id: 50205174 --- 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/bbd14a5fe6b002419d127c8ec63e3062d934e058_performance_tests/17 job id: 50205175 --- 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/bbd14a5fe6b002419d127c8ec63e3062d934e058_performance_tests/18 job id: 50205178 --- 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/bbd14a5fe6b002419d127c8ec63e3062d934e058_performance_tests/19 job id: 50205180 --- 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/bbd14a5fe6b002419d127c8ec63e3062d934e058_performance_tests/20 job id: 50205182 --- 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/bbd14a5fe6b002419d127c8ec63e3062d934e058_performance_tests/21 job id: 50205183 --- 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/bbd14a5fe6b002419d127c8ec63e3062d934e058_performance_tests/22 job id: 50205184 --- 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/bbd14a5fe6b002419d127c8ec63e3062d934e058_performance_tests/23 job id: 50205185 --- 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/bbd14a5fe6b002419d127c8ec63e3062d934e058_performance_tests/24 job id: 50205186 --- 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/bbd14a5fe6b002419d127c8ec63e3062d934e058_performance_tests/25 job id: 50205187 --- 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/bbd14a5fe6b002419d127c8ec63e3062d934e058_performance_tests/26 job id: 50205188 --- 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/bbd14a5fe6b002419d127c8ec63e3062d934e058_performance_tests/27 job id: 50205189 --- 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/bbd14a5fe6b002419d127c8ec63e3062d934e058_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 424 8. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.032 133.609 133.609 farming_run 1 2.0 133.085 133.123 133.579 133.580 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.488540E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 230 1134128. MP_Allreduce 571 1938539. MP_Sync 25 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 10 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.006 0.026 114.643 114.643 qs_energies 1 2.0 0.000 0.000 114.438 114.441 mp2_main 1 3.0 0.000 0.000 112.290 112.293 mp2_gpw_main 1 4.0 0.027 0.033 111.274 111.277 mp2_ri_gpw_compute_in 1 5.0 0.172 0.174 92.366 92.717 mp2_ri_gpw_compute_in_loop 1 6.0 0.003 0.003 54.705 55.055 mp2_eri_3c_integrate_gpw 272 7.0 0.151 0.162 41.113 46.160 get_2c_integrals 1 6.0 0.008 0.009 36.947 37.490 integrate_v_rspace 273 8.0 0.440 0.457 24.680 29.603 pw_transfer 6555 10.6 0.374 0.388 26.896 27.654 fft_wrap_pw1pw2 5465 11.4 0.045 0.047 25.539 26.226 grid_integrate_task_list 273 9.0 20.571 25.981 20.571 25.981 fft_wrap_pw1pw2_100 2178 12.4 0.028 0.030 23.072 23.739 compute_2c_integrals 1 7.0 0.002 0.002 19.390 19.392 compute_2c_integrals_loop_lm 1 8.0 0.001 0.003 18.615 19.108 mp2_eri_2c_integrate_gpw 1 9.0 2.380 2.424 18.614 19.107 rpa_ri_compute_en 1 5.0 0.024 0.025 18.802 19.064 cp_fm_cholesky_decompose 12 8.2 17.515 18.064 17.515 18.064 cholesky_decomp 1 7.0 0.000 0.000 16.393 16.942 fft3d_s 5443 13.4 16.126 16.502 16.146 16.523 ao_to_mo_and_store_B_mult_1 272 7.0 10.766 15.312 10.766 15.312 calculate_wavefunction 272 8.0 5.403 5.565 12.342 12.986 rpa_num_int 1 6.0 0.000 0.001 10.678 10.688 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.658 10.672 calc_mat_Q 8 8.0 0.000 0.000 9.496 9.586 contract_S_to_Q 8 9.0 0.000 0.000 8.917 9.005 calc_potential_gpw 544 9.5 0.004 0.004 8.237 8.691 parallel_gemm_fm 14 9.1 0.000 0.000 8.502 8.606 parallel_gemm_fm_cosma 14 10.1 8.502 8.606 8.502 8.606 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 8.089 8.400 potential_pw2rs 545 10.0 0.106 0.108 7.509 8.269 create_integ_mat 1 6.0 0.014 0.028 7.676 7.686 collocate_single_gaussian 272 10.0 0.039 0.042 7.344 7.647 array2fm 1 7.0 0.000 0.000 6.699 7.105 pw_scatter_s 2720 13.7 4.363 4.581 4.363 4.581 pw_gather_s 2722 13.2 3.481 3.799 3.481 3.799 array2fm_buffer_send 1 8.0 2.919 3.157 2.919 3.157 pw_poisson_solve 545 10.5 1.118 1.181 2.184 2.332 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.276719, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2794.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bbd14a5fe6b002419d127c8ec63e3062d934e058_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 205321. MP_Allreduce 424 9. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.023 0.037 421.203 421.310 farming_run 1 2.0 420.393 420.501 421.165 421.263 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827128576 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788821 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.239974E+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 23252012. MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.033 216.777 216.832 qs_energies 1 2.0 0.000 0.000 216.529 216.607 scf_env_do_scf 1 3.0 0.000 0.000 114.408 114.437 qs_ks_update_qs_env 5 5.0 0.000 0.000 113.435 113.467 rebuild_ks_matrix 4 6.0 0.000 0.000 113.434 113.466 qs_ks_build_kohn_sham_matrix 4 7.0 0.054 0.060 113.434 113.466 hfx_ks_matrix 4 8.0 0.001 0.001 113.079 113.107 integrate_four_center 4 9.0 0.153 0.460 113.078 113.106 integrate_four_center_main 4 10.0 0.091 0.467 101.910 104.413 integrate_four_center_bin 268 11.0 101.818 104.371 101.818 104.371 mp2_main 1 3.0 0.002 0.013 101.817 101.866 mp2_gpw_main 1 4.0 0.047 0.071 100.883 100.935 init_scf_loop 1 4.0 0.000 0.000 96.060 96.084 mp2_ri_gpw_compute_in 1 5.0 0.064 0.064 73.955 74.941 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 53.760 54.745 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.160 41.472 46.427 integrate_v_rspace 95 8.0 0.397 0.567 27.958 32.768 pw_transfer 2240 10.6 0.144 0.167 29.549 29.975 fft_wrap_pw1pw2 1868 11.4 0.017 0.021 28.557 28.944 mp2_ri_gpw_compute_en 1 5.0 0.067 0.092 26.775 28.673 ao_to_mo_and_store_B_mult_1 91 7.0 10.615 28.597 10.615 28.597 grid_integrate_task_list 95 9.0 23.371 28.392 23.371 28.392 fft_wrap_pw1pw2_100 730 12.4 0.012 0.014 26.301 26.653 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.840 1.907 25.126 25.137 get_2c_integrals 1 6.0 0.000 0.000 20.097 20.130 compute_2c_integrals 1 7.0 0.002 0.003 19.074 19.078 fft3d_s 1823 13.4 18.490 18.944 18.503 18.957 compute_2c_integrals_loop_lm 1 8.0 0.001 0.002 18.683 18.950 mp2_eri_2c_integrate_gpw 1 9.0 1.730 1.837 18.682 18.950 scf_env_do_scf_inner_loop 4 4.0 0.000 0.001 18.347 18.351 calculate_wavefunction 91 8.0 2.011 2.043 9.617 9.813 mp2_ri_gpw_compute_en_expansio 172 7.0 0.556 0.587 8.812 9.213 potential_pw2rs 186 10.0 0.033 0.034 8.452 9.057 mp2_ri_gpw_compute_en_comm 22 7.0 0.492 0.511 8.095 8.710 local_gemm 172 8.0 8.256 8.667 8.256 8.667 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.187 8.556 calc_potential_gpw 182 9.5 0.002 0.002 7.906 8.129 collocate_single_gaussian 91 10.0 0.016 0.021 7.775 8.007 mp_sendrecv_dm3 2068 8.0 6.146 6.748 6.146 6.748 integrate_four_center_load 4 10.0 0.000 0.000 6.744 6.747 hfx_load_balance 1 11.0 0.000 0.000 6.744 6.747 mp2_ri_gpw_compute_en_ener 172 7.0 6.335 6.421 6.335 6.421 mp_sync 37 10.5 3.153 6.018 3.153 6.018 pw_gather_s 912 13.2 4.493 5.081 4.493 5.081 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=100.934847, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1498.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bbd14a5fe6b002419d127c8ec63e3062d934e058_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 453.857280E+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 MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.009 0.026 63.967 63.968 qs_mol_dyn_low 1 2.0 0.003 0.003 63.746 63.755 qs_forces 11 3.9 0.003 0.003 63.524 63.525 qs_energies 11 4.9 0.002 0.002 61.783 61.801 scf_env_do_scf 11 5.9 0.001 0.001 55.134 55.134 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 52.182 52.183 qs_scf_new_mos 108 7.5 0.000 0.001 39.089 39.428 qs_scf_loop_do_ot 108 8.5 0.000 0.001 39.088 39.427 dbcsr_multiply_generic 2286 12.5 0.094 0.098 37.094 37.612 ot_scf_mini 108 9.5 0.002 0.002 37.224 37.445 velocity_verlet 10 3.0 0.001 0.002 31.557 31.559 multiply_cannon 2286 13.5 0.189 0.204 27.398 29.233 multiply_cannon_loop 2286 14.5 1.820 1.933 26.224 28.040 ot_mini 108 10.5 0.001 0.001 22.452 22.721 qs_ot_get_derivative 108 11.5 0.001 0.001 19.232 19.395 mp_waitall_1 245248 16.5 10.329 16.491 10.329 16.491 multiply_cannon_metrocomm3 54864 15.5 0.072 0.076 6.485 13.567 multiply_cannon_multrec 54864 15.5 3.569 5.568 7.602 10.953 qs_ot_get_p 119 10.4 0.001 0.002 9.791 10.090 rebuild_ks_matrix 119 8.3 0.000 0.000 9.641 9.814 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.013 9.641 9.813 qs_ks_update_qs_env 119 7.6 0.001 0.001 8.502 8.663 mp_sum_l 7287 12.8 6.423 8.297 6.423 8.297 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 6.459 6.907 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 6.605 6.713 qs_ot_p2m_diag 50 11.0 0.004 0.006 6.244 6.279 multiply_cannon_sync_h2d 54864 15.5 5.019 6.207 5.019 6.207 qs_rho_update_rho_low 119 7.7 0.000 0.001 5.985 6.148 calculate_rho_elec 119 8.7 0.011 0.016 5.985 6.147 sum_up_and_integrate 119 10.3 0.001 0.002 5.868 5.875 integrate_v_rspace 119 11.3 0.002 0.002 5.857 5.865 dbcsr_mm_accdrv_process 76910 16.1 1.843 2.893 3.948 5.648 cp_dbcsr_syevd 50 12.0 0.002 0.003 5.315 5.316 init_scf_run 11 5.9 0.000 0.001 5.209 5.209 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.209 5.209 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.849 4.851 cp_fm_redistribute_end 50 14.0 2.459 4.791 2.480 4.798 cp_fm_diag_elpa_base 50 14.0 2.301 4.631 2.311 4.652 density_rs2pw 119 9.7 0.004 0.004 3.691 3.788 multiply_cannon_metrocomm1 54864 15.5 0.055 0.060 2.176 3.618 pw_transfer 1439 11.6 0.051 0.055 3.169 3.299 calculate_dm_sparse 119 9.5 0.000 0.000 3.109 3.265 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.023 3.247 apply_single 119 13.6 0.000 0.000 3.022 3.247 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 3.094 3.228 potential_pw2rs 119 12.3 0.004 0.004 3.086 3.159 make_m2s 4572 13.5 0.054 0.056 2.959 3.073 make_images 4572 14.5 0.133 0.139 2.876 2.991 fft3d_ps 1201 14.6 0.377 0.490 2.814 2.950 init_scf_loop 11 6.9 0.000 0.000 2.914 2.915 ot_diis_step 108 11.5 0.006 0.006 2.877 2.878 acc_transpose_blocks 54864 15.5 0.223 0.247 2.211 2.741 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.663 2.736 jit_kernel_multiply 13 15.8 2.041 2.673 2.041 2.673 wfi_extrapolate 11 7.9 0.001 0.001 2.660 2.661 fft_wrap_pw1pw2_140 487 13.2 0.103 0.109 2.362 2.496 calculate_first_density_matrix 1 7.0 0.000 0.001 2.429 2.437 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.431 2.433 mp_alltoall_d11v 2130 13.8 2.224 2.393 2.224 2.393 transfer_rs2pw 487 10.6 0.005 0.006 2.135 2.235 mp_sum_d 4137 12.0 1.535 2.195 1.535 2.195 transfer_pw2rs 487 13.2 0.005 0.006 2.114 2.189 grid_integrate_task_list 119 12.3 2.001 2.120 2.001 2.120 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.969 2.021 mp_alltoall_z22v 1201 16.6 1.752 1.852 1.752 1.852 make_images_sizes 4572 15.5 0.004 0.005 1.451 1.805 mp_alltoall_i44 4572 16.5 1.446 1.801 1.446 1.801 prepare_preconditioner 11 7.9 0.000 0.000 1.563 1.603 make_preconditioner 11 8.9 0.000 0.000 1.563 1.603 mp_waitany 12084 13.8 1.416 1.596 1.416 1.596 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.463 1.511 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.463 1.483 grid_collocate_task_list 119 9.7 1.351 1.438 1.351 1.438 acc_transpose_blocks_sync 164592 16.5 1.196 1.432 1.196 1.432 dbcsr_dot_sd 1205 11.9 0.049 0.059 0.855 1.303 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=63.968000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=432.000000, yerr=1.414214 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bbd14a5fe6b002419d127c8ec63e3062d934e058_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 489.242624E+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 956109. MP_SendRecv 16779 37093. MP_ISendRecv 16779 37093. MP_Wait 23539 MP_comm_split 50 MP_ISend 5720 128509. MP_IRecv 5720 128509. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.008 0.026 41.602 41.603 qs_mol_dyn_low 1 2.0 0.003 0.003 41.217 41.384 qs_forces 11 3.9 0.002 0.003 41.152 41.153 qs_energies 11 4.9 0.002 0.002 39.421 39.426 scf_env_do_scf 11 5.9 0.001 0.001 33.724 33.724 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 31.036 31.037 dbcsr_multiply_generic 2286 12.5 0.102 0.104 23.157 23.501 qs_scf_new_mos 108 7.5 0.001 0.001 21.673 21.908 qs_scf_loop_do_ot 108 8.5 0.001 0.001 21.673 21.908 ot_scf_mini 108 9.5 0.002 0.003 20.726 20.894 velocity_verlet 10 3.0 0.001 0.002 19.513 19.514 multiply_cannon 2286 13.5 0.208 0.216 17.599 19.166 multiply_cannon_loop 2286 14.5 1.186 1.251 16.340 17.801 ot_mini 108 10.5 0.001 0.001 12.663 12.896 mp_waitall_1 200699 16.5 6.151 11.486 6.151 11.486 qs_ot_get_derivative 108 11.5 0.001 0.001 10.268 10.438 multiply_cannon_metrocomm3 27432 15.5 0.071 0.073 4.246 9.916 multiply_cannon_multrec 27432 15.5 1.813 4.223 6.362 9.285 rebuild_ks_matrix 119 8.3 0.000 0.000 7.192 7.332 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.022 7.191 7.332 dbcsr_mm_accdrv_process 47894 16.0 3.695 6.100 4.471 6.635 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.326 6.452 qs_ot_get_p 119 10.4 0.001 0.002 5.063 5.288 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.902 4.754 init_scf_run 11 5.9 0.000 0.001 4.478 4.478 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.478 4.478 mp_sum_l 7287 12.8 2.111 4.181 2.111 4.181 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.059 4.090 calculate_rho_elec 119 8.7 0.021 0.024 4.058 4.090 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.030 4.084 apply_single 119 13.6 0.000 0.000 3.030 4.084 sum_up_and_integrate 119 10.3 0.001 0.002 4.077 4.083 integrate_v_rspace 119 11.3 0.002 0.002 4.063 4.069 qs_ot_p2m_diag 50 11.0 0.008 0.012 3.285 3.303 make_m2s 4572 13.5 0.051 0.053 2.992 3.300 make_images 4572 14.5 0.205 0.243 2.902 3.214 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.862 2.862 calculate_first_density_matrix 1 7.0 0.000 0.000 2.823 2.824 init_scf_loop 11 6.9 0.000 0.000 2.662 2.663 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.474 2.564 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.481 2.482 cp_fm_redistribute_end 50 14.0 1.253 2.445 1.259 2.449 density_rs2pw 119 9.7 0.004 0.004 2.345 2.447 cp_fm_diag_elpa_base 50 14.0 1.154 2.336 1.184 2.385 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.346 2.349 ot_diis_step 108 11.5 0.011 0.011 2.344 2.345 calculate_dm_sparse 119 9.5 0.000 0.001 2.209 2.288 acc_transpose_blocks 27432 15.5 0.115 0.122 1.795 2.229 multiply_cannon_sync_h2d 27432 15.5 1.688 2.198 1.688 2.198 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.095 2.135 pw_transfer 1439 11.6 0.065 0.068 2.035 2.075 jit_kernel_multiply 9 16.3 0.718 1.998 0.718 1.998 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.944 1.987 grid_integrate_task_list 119 12.3 1.840 1.951 1.840 1.951 potential_pw2rs 119 12.3 0.006 0.006 1.796 1.810 make_images_data 4572 15.5 0.047 0.054 1.343 1.753 prepare_preconditioner 11 7.9 0.000 0.000 1.714 1.740 make_preconditioner 11 8.9 0.000 0.000 1.714 1.740 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.604 1.664 fft3d_ps 1201 14.6 0.523 0.578 1.576 1.616 hybrid_alltoall_any 4725 16.4 0.053 0.114 1.196 1.611 wfi_extrapolate 11 7.9 0.001 0.001 1.576 1.576 mp_alltoall_d11v 2130 13.8 1.353 1.538 1.353 1.538 fft_wrap_pw1pw2_140 487 13.2 0.127 0.130 1.456 1.494 transfer_rs2pw 487 10.6 0.005 0.005 1.413 1.482 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.375 1.383 grid_collocate_task_list 119 9.7 1.290 1.371 1.290 1.371 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.247 1.295 transfer_pw2rs 487 13.2 0.004 0.005 1.183 1.196 acc_transpose_blocks_kernels 27432 16.5 0.191 0.281 0.839 1.139 mp_allgather_i34 2286 14.5 0.684 1.107 0.684 1.107 make_images_sizes 4572 15.5 0.005 0.005 0.783 1.099 mp_alltoall_i44 4572 16.5 0.778 1.094 0.778 1.094 mp_sum_d 4137 12.0 0.567 1.041 0.567 1.041 acc_transpose_blocks_sync 82296 16.5 0.815 0.942 0.815 0.942 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.924 0.939 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.937 0.938 mp_alltoall_z22v 1201 16.6 0.795 0.895 0.795 0.895 jit_kernel_transpose 5 15.5 0.649 0.859 0.649 0.859 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=41.603000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=465.909091, yerr=1.443137 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bbd14a5fe6b002419d127c8ec63e3062d934e058_performance_tests/05/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 59.051995E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3143552 0.0% 0.0% 100.0% average stack size 0.0 0.0 46.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 522.596352E+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 0.016 0.035 42.114 42.116 qs_mol_dyn_low 1 2.0 0.003 0.003 41.848 41.857 qs_forces 11 3.9 0.002 0.003 41.485 41.486 qs_energies 11 4.9 0.002 0.002 39.714 39.723 scf_env_do_scf 11 5.9 0.001 0.001 34.500 34.500 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 30.741 30.742 velocity_verlet 10 3.0 0.002 0.002 20.625 20.636 qs_scf_new_mos 108 7.5 0.001 0.001 20.223 20.244 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.223 20.244 dbcsr_multiply_generic 2286 12.5 0.101 0.103 19.850 20.240 ot_scf_mini 108 9.5 0.002 0.002 19.266 19.282 multiply_cannon 2286 13.5 0.195 0.199 14.474 15.707 multiply_cannon_loop 2286 14.5 0.867 0.909 13.351 14.612 ot_mini 108 10.5 0.001 0.001 11.323 11.362 qs_ot_get_derivative 108 11.5 0.001 0.001 9.531 9.546 rebuild_ks_matrix 119 8.3 0.000 0.000 7.667 7.781 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 7.666 7.781 multiply_cannon_multrec 18288 15.5 1.841 2.873 6.932 7.266 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.804 6.905 mp_waitall_1 158411 16.6 4.691 6.318 4.691 6.318 dbcsr_mm_accdrv_process 38222 16.0 4.901 5.685 5.000 5.750 qs_ot_get_p 119 10.4 0.001 0.001 5.253 5.294 sum_up_and_integrate 119 10.3 0.001 0.002 5.039 5.043 integrate_v_rspace 119 11.3 0.002 0.003 5.026 5.030 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.766 4.776 calculate_rho_elec 119 8.7 0.031 0.031 4.766 4.775 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.514 4.140 init_scf_run 11 5.9 0.000 0.001 3.909 3.909 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.909 3.909 init_scf_loop 11 6.9 0.000 0.000 3.721 3.722 qs_ot_p2m_diag 50 11.0 0.012 0.013 3.468 3.479 multiply_cannon_metrocomm3 18288 15.5 0.048 0.049 1.833 3.424 density_rs2pw 119 9.7 0.004 0.004 3.254 3.355 cp_dbcsr_syevd 50 12.0 0.003 0.003 3.091 3.092 make_m2s 4572 13.5 0.044 0.045 2.913 3.072 make_images 4572 14.5 0.192 0.206 2.826 2.985 potential_pw2rs 119 12.3 0.007 0.008 2.737 2.769 prepare_preconditioner 11 7.9 0.000 0.000 2.752 2.756 make_preconditioner 11 8.9 0.000 0.000 2.752 2.756 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.577 2.674 pw_transfer 1439 11.6 0.065 0.068 2.620 2.658 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.641 2.650 cp_fm_diag_elpa_base 50 14.0 2.606 2.623 2.638 2.647 mp_sum_l 7287 12.8 2.079 2.574 2.079 2.574 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.527 2.565 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.142 2.562 apply_single 119 13.6 0.000 0.000 2.142 2.562 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.388 2.399 calculate_first_density_matrix 1 7.0 0.000 0.001 2.225 2.225 transfer_rs2pw 487 10.6 0.005 0.005 2.023 2.127 fft3d_ps 1201 14.6 0.556 0.576 2.085 2.124 calculate_dm_sparse 119 9.5 0.000 0.001 2.067 2.080 grid_integrate_task_list 119 12.3 1.791 1.924 1.791 1.924 transfer_pw2rs 487 13.2 0.004 0.004 1.891 1.918 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.839 1.842 acc_transpose_blocks 18288 15.5 0.080 0.083 1.651 1.755 fft_wrap_pw1pw2_140 487 13.2 0.178 0.183 1.682 1.718 ot_diis_step 108 11.5 0.011 0.012 1.685 1.686 wfi_extrapolate 11 7.9 0.001 0.001 1.620 1.620 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.563 1.602 make_images_data 4572 15.5 0.047 0.050 1.265 1.534 cp_fm_cholesky_invert 11 10.9 1.397 1.407 1.397 1.407 hybrid_alltoall_any 4725 16.4 0.058 0.117 1.074 1.384 mp_alltoall_z22v 1201 16.6 1.285 1.375 1.285 1.375 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.299 1.359 grid_collocate_task_list 119 9.7 1.232 1.323 1.232 1.323 mp_alltoall_d11v 2130 13.8 1.147 1.316 1.147 1.316 mp_sendrecv_dv 11067 12.7 1.240 1.270 1.240 1.270 make_images_sizes 4572 15.5 0.005 0.005 0.900 1.251 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.235 1.250 mp_alltoall_i44 4572 16.5 0.895 1.246 0.895 1.246 multiply_cannon_sync_h2d 18288 15.5 0.960 1.148 0.960 1.148 mp_allgather_i34 2286 14.5 0.569 1.019 0.569 1.019 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.975 0.975 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.943 0.955 dbcsr_complete_redistribute 329 12.2 0.092 0.106 0.777 0.893 acc_transpose_blocks_sync 54864 16.5 0.746 0.849 0.746 0.849 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=42.116000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=497.181818, yerr=1.336085 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bbd14a5fe6b002419d127c8ec63e3062d934e058_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 559.235072E+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 0.010 0.029 46.803 46.804 qs_mol_dyn_low 1 2.0 0.003 0.003 46.606 46.614 qs_forces 11 3.9 0.002 0.002 46.538 46.539 qs_energies 11 4.9 0.002 0.002 44.597 44.603 scf_env_do_scf 11 5.9 0.001 0.001 38.918 38.919 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 33.822 33.823 velocity_verlet 10 3.0 0.001 0.002 24.014 24.016 qs_scf_new_mos 108 7.5 0.001 0.001 23.137 23.199 qs_scf_loop_do_ot 108 8.5 0.001 0.001 23.137 23.198 dbcsr_multiply_generic 2286 12.5 0.104 0.106 22.687 22.805 ot_scf_mini 108 9.5 0.002 0.003 21.933 21.983 multiply_cannon 2286 13.5 0.218 0.222 16.780 17.247 multiply_cannon_loop 2286 14.5 1.532 1.610 15.335 15.838 ot_mini 108 10.5 0.001 0.001 12.559 12.628 qs_ot_get_derivative 108 11.5 0.001 0.001 10.522 10.574 multiply_cannon_multrec 27432 15.5 2.444 3.111 8.998 9.240 rebuild_ks_matrix 119 8.3 0.000 0.000 7.913 7.960 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.015 7.913 7.959 dbcsr_mm_accdrv_process 47916 15.9 5.886 7.487 6.454 7.783 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.030 7.071 qs_ot_get_p 119 10.4 0.001 0.002 6.108 6.188 init_scf_loop 11 6.9 0.000 0.000 5.052 5.053 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.832 4.877 calculate_rho_elec 119 8.7 0.040 0.046 4.832 4.877 sum_up_and_integrate 119 10.3 0.001 0.002 4.850 4.860 integrate_v_rspace 119 11.3 0.002 0.003 4.839 4.849 mp_waitall_1 137007 16.6 3.613 4.639 3.613 4.639 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.736 4.264 init_scf_run 11 5.9 0.000 0.001 4.163 4.164 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.163 4.163 qs_ot_p2m_diag 50 11.0 0.015 0.022 3.989 4.000 prepare_preconditioner 11 7.9 0.000 0.000 3.983 3.991 make_preconditioner 11 8.9 0.000 0.000 3.983 3.991 make_m2s 4572 13.5 0.054 0.056 3.750 3.947 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.535 3.891 make_images 4572 14.5 0.271 0.334 3.641 3.842 cp_dbcsr_syevd 50 12.0 0.003 0.003 3.477 3.477 density_rs2pw 119 9.7 0.003 0.004 3.213 3.343 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.993 3.021 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.928 2.937 cp_fm_diag_elpa_base 50 14.0 2.875 2.897 2.923 2.932 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.353 2.780 apply_single 119 13.6 0.000 0.000 2.353 2.780 pw_transfer 1439 11.6 0.065 0.068 2.706 2.741 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.614 2.651 potential_pw2rs 119 12.3 0.009 0.009 2.570 2.590 acc_transpose_blocks 27432 15.5 0.121 0.124 2.412 2.526 calculate_dm_sparse 119 9.5 0.000 0.000 2.329 2.396 mp_sum_l 7287 12.8 1.731 2.364 1.731 2.364 calculate_first_density_matrix 1 7.0 0.000 0.001 2.233 2.234 fft3d_ps 1201 14.6 0.598 0.659 2.104 2.135 transfer_rs2pw 487 10.6 0.004 0.005 1.925 2.044 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.978 1.981 ot_diis_step 108 11.5 0.012 0.012 1.975 1.976 fft_wrap_pw1pw2_140 487 13.2 0.228 0.235 1.902 1.940 grid_integrate_task_list 119 12.3 1.815 1.918 1.815 1.918 make_images_data 4572 15.5 0.048 0.052 1.534 1.888 wfi_extrapolate 11 7.9 0.001 0.001 1.861 1.862 transfer_pw2rs 487 13.2 0.004 0.004 1.733 1.761 multiply_cannon_metrocomm3 27432 15.5 0.040 0.040 0.840 1.682 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.651 1.669 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.588 1.624 hybrid_alltoall_any 4725 16.4 0.065 0.152 1.275 1.588 cp_fm_cholesky_invert 11 10.9 1.572 1.580 1.572 1.580 dbcsr_complete_redistribute 329 12.2 0.122 0.150 1.268 1.554 acc_transpose_blocks_sync 82296 16.5 1.440 1.548 1.440 1.548 jit_kernel_multiply 8 15.9 0.502 1.485 0.502 1.485 make_images_sizes 4572 15.5 0.005 0.005 1.088 1.453 mp_alltoall_i44 4572 16.5 1.083 1.448 1.083 1.448 mp_alltoall_z22v 1201 16.6 1.389 1.419 1.389 1.419 mp_alltoall_d11v 2130 13.8 1.290 1.389 1.290 1.389 grid_collocate_task_list 119 9.7 1.249 1.339 1.249 1.339 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.291 1.303 cp_fm_upper_to_full 72 14.2 0.895 1.288 0.895 1.288 mp_sendrecv_dv 8211 12.7 1.203 1.222 1.203 1.222 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.121 1.122 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.808 1.090 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.984 0.990 cp_fm_cholesky_decompose 22 10.9 0.947 0.952 0.947 0.952 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=46.804000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=531.545455, yerr=3.726164 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bbd14a5fe6b002419d127c8ec63e3062d934e058_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 612.413440E+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 0.009 0.026 29.804 29.805 qs_mol_dyn_low 1 2.0 0.003 0.003 29.587 29.595 qs_forces 11 3.9 0.003 0.003 29.525 29.526 qs_energies 11 4.9 0.002 0.002 27.802 27.805 scf_env_do_scf 11 5.9 0.001 0.001 22.860 22.860 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 20.307 20.308 velocity_verlet 10 3.0 0.002 0.002 15.066 15.069 dbcsr_multiply_generic 2286 12.5 0.094 0.096 13.487 13.609 qs_scf_new_mos 108 7.5 0.001 0.001 12.361 12.384 qs_scf_loop_do_ot 108 8.5 0.001 0.001 12.360 12.383 ot_scf_mini 108 9.5 0.002 0.002 11.637 11.661 multiply_cannon 2286 13.5 0.224 0.230 10.742 11.374 multiply_cannon_loop 2286 14.5 0.639 0.661 9.799 10.029 ot_mini 108 10.5 0.001 0.001 6.743 6.772 multiply_cannon_multrec 9144 15.5 1.740 1.932 6.079 6.257 rebuild_ks_matrix 119 8.3 0.000 0.000 5.941 5.965 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.016 5.940 5.965 qs_ot_get_derivative 108 11.5 0.001 0.001 5.447 5.468 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.293 5.315 dbcsr_mm_accdrv_process 12550 15.8 3.625 4.245 4.229 4.298 sum_up_and_integrate 119 10.3 0.001 0.002 3.555 3.561 integrate_v_rspace 119 11.3 0.003 0.003 3.545 3.550 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.528 3.534 calculate_rho_elec 119 8.7 0.060 0.061 3.527 3.534 init_scf_run 11 5.9 0.000 0.001 3.494 3.494 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.494 3.494 qs_ot_get_p 119 10.4 0.001 0.002 2.970 3.003 init_scf_loop 11 6.9 0.000 0.000 2.532 2.533 calculate_first_density_matrix 1 7.0 0.006 0.043 2.314 2.316 pw_transfer 1439 11.6 0.065 0.067 2.211 2.223 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.118 2.131 mp_waitall_1 115863 16.7 1.634 2.101 1.634 2.101 make_m2s 4572 13.5 0.033 0.034 1.913 2.077 make_images 4572 14.5 0.269 0.299 1.823 1.986 density_rs2pw 119 9.7 0.003 0.003 1.908 1.970 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.948 1.949 grid_integrate_task_list 119 12.3 1.870 1.944 1.870 1.944 calculate_dm_sparse 119 9.5 0.000 0.000 1.796 1.818 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.799 1.815 prepare_preconditioner 11 7.9 0.000 0.000 1.778 1.783 make_preconditioner 11 8.9 0.000 0.000 1.778 1.783 fft_wrap_pw1pw2_140 487 13.2 0.325 0.333 1.724 1.740 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.700 1.701 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.665 1.694 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.645 1.646 acc_transpose_blocks 9144 15.5 0.043 0.044 1.452 1.488 fft3d_ps 1201 14.6 0.652 0.668 1.470 1.479 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.432 1.445 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.434 1.444 jit_kernel_multiply 8 15.7 0.564 1.431 0.564 1.431 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.411 1.419 cp_fm_diag_elpa_base 50 14.0 1.382 1.401 1.409 1.417 grid_collocate_task_list 119 9.7 1.299 1.376 1.299 1.376 potential_pw2rs 119 12.3 0.010 0.010 1.348 1.352 ot_diis_step 108 11.5 0.013 0.013 1.279 1.280 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.216 1.254 apply_single 119 13.6 0.000 0.000 1.216 1.254 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.229 1.230 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.173 1.184 wfi_extrapolate 11 7.9 0.001 0.001 1.122 1.122 hybrid_alltoall_any 4725 16.4 0.064 0.175 0.810 1.060 make_images_data 4572 15.5 0.041 0.045 0.844 1.042 transfer_rs2pw 487 10.6 0.004 0.004 0.891 0.970 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.876 0.931 cp_fm_cholesky_invert 11 10.9 0.884 0.886 0.884 0.886 mp_alltoall_d11v 2130 13.8 0.811 0.879 0.811 0.879 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.837 0.844 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.769 0.773 acc_transpose_blocks_sync 27432 16.5 0.711 0.744 0.711 0.744 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.675 0.725 mp_alltoall_z22v 1201 16.6 0.691 0.721 0.691 0.721 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 0.405 0.716 mp_allgather_i34 2286 14.5 0.259 0.708 0.259 0.708 transfer_pw2rs 487 13.2 0.003 0.003 0.692 0.695 acc_transpose_blocks_kernels 9144 16.5 0.119 0.121 0.682 0.684 dbcsr_complete_redistribute 329 12.2 0.209 0.219 0.602 0.643 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=29.805000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=579.818182, yerr=6.421812 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bbd14a5fe6b002419d127c8ec63e3062d934e058_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 785.928192E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63723. MP_Allreduce 10154 429. MP_Sync 54 MP_Alltoall 1582 7383731. MP_SendRecv 2499 189067. MP_ISendRecv 2499 189067. MP_Wait 6399 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.033 50.576 50.577 qs_mol_dyn_low 1 2.0 0.003 0.003 50.304 50.312 qs_forces 11 3.9 0.002 0.002 50.209 50.210 qs_energies 11 4.9 0.002 0.002 48.019 48.026 scf_env_do_scf 11 5.9 0.001 0.001 41.858 41.859 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 32.536 32.538 velocity_verlet 10 3.0 0.002 0.002 28.402 28.411 dbcsr_multiply_generic 2286 12.5 0.106 0.108 21.386 21.699 qs_scf_new_mos 108 7.5 0.001 0.001 21.026 21.122 qs_scf_loop_do_ot 108 8.5 0.001 0.001 21.025 21.121 ot_scf_mini 108 9.5 0.002 0.002 19.718 19.818 multiply_cannon 2286 13.5 0.296 0.300 15.758 17.230 multiply_cannon_loop 2286 14.5 0.859 0.872 14.344 15.433 ot_mini 108 10.5 0.001 0.001 11.636 11.778 qs_ot_get_derivative 108 11.5 0.001 0.001 9.323 9.427 init_scf_loop 11 6.9 0.000 0.000 9.287 9.291 multiply_cannon_multrec 9144 15.5 3.426 4.858 8.912 9.038 rebuild_ks_matrix 119 8.3 0.000 0.001 8.413 8.606 qs_ks_build_kohn_sham_matrix 119 9.3 0.014 0.014 8.412 8.605 prepare_preconditioner 11 7.9 0.000 0.000 8.195 8.210 make_preconditioner 11 8.9 0.000 0.000 8.195 8.210 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.687 8.071 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.559 7.734 dbcsr_mm_accdrv_process 12550 15.8 4.810 6.546 5.350 6.579 qs_rho_update_rho_low 119 7.7 0.001 0.001 5.151 5.157 calculate_rho_elec 119 8.7 0.117 0.120 5.151 5.157 qs_ot_get_p 119 10.4 0.002 0.002 4.775 4.942 mp_waitall_1 94719 16.7 3.832 4.937 3.832 4.937 sum_up_and_integrate 119 10.3 0.001 0.001 4.674 4.683 integrate_v_rspace 119 11.3 0.003 0.003 4.664 4.672 cp_fm_upper_to_full 72 14.2 3.202 4.653 3.202 4.653 init_scf_run 11 5.9 0.000 0.001 3.994 3.994 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.994 3.994 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.271 3.724 pw_transfer 1439 11.6 0.068 0.068 3.699 3.710 make_m2s 4572 13.5 0.037 0.037 3.464 3.696 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 3.600 3.612 make_images 4572 14.5 0.350 0.385 3.345 3.577 dbcsr_complete_redistribute 329 12.2 0.290 0.299 2.228 3.089 density_rs2pw 119 9.7 0.003 0.004 3.014 3.025 fft_wrap_pw1pw2_140 487 13.2 0.649 0.653 2.934 2.945 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.920 2.926 multiply_cannon_metrocomm3 9144 15.5 0.021 0.021 1.741 2.871 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.828 2.693 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.389 2.623 apply_single 119 13.6 0.000 0.000 2.389 2.622 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.536 2.590 mp_sum_l 7287 12.8 1.634 2.585 1.634 2.585 fft3d_ps 1201 14.6 0.873 0.892 2.522 2.534 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.517 2.520 calculate_dm_sparse 119 9.5 0.000 0.000 2.456 2.471 mp_alltoall_i22 627 13.8 1.522 2.447 1.522 2.447 cp_fm_cholesky_invert 11 10.9 2.353 2.357 2.353 2.357 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.502 2.351 ot_diis_step 108 11.5 0.014 0.015 2.249 2.249 calculate_first_density_matrix 1 7.0 0.000 0.000 2.209 2.211 grid_integrate_task_list 119 12.3 2.170 2.187 2.170 2.187 make_images_data 4572 15.5 0.046 0.049 1.727 2.131 hybrid_alltoall_any 4725 16.4 0.090 0.150 1.666 2.107 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.089 2.090 cp_fm_diag_elpa_base 50 14.0 1.924 1.986 2.087 2.087 potential_pw2rs 119 12.3 0.014 0.014 2.010 2.014 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.918 1.918 acc_transpose_blocks 9144 15.5 0.045 0.046 1.809 1.831 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.812 1.813 wfi_extrapolate 11 7.9 0.001 0.001 1.711 1.711 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.603 1.658 grid_collocate_task_list 119 9.7 1.585 1.594 1.585 1.594 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.527 1.577 mp_alltoall_d11v 2130 13.8 1.541 1.565 1.541 1.565 mp_alltoall_z22v 1201 16.6 1.511 1.539 1.511 1.539 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.338 1.350 transfer_rs2pw 487 10.6 0.005 0.005 1.166 1.193 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.143 1.156 acc_transpose_blocks_sync 27432 16.5 1.089 1.111 1.089 1.111 make_images_sizes 4572 15.5 0.005 0.005 0.639 1.090 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.069 1.086 mp_alltoall_i44 4572 16.5 0.634 1.085 0.634 1.085 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.991 1.023 transfer_pw2rs 487 13.2 0.003 0.003 1.009 1.016 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=50.577000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=739.181818, yerr=15.331597 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bbd14a5fe6b002419d127c8ec63e3062d934e058_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 504.725504E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66212. MP_Allreduce 9776 488. MP_Sync 52 MP_Alltoall 1938 1383689. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.050 94.348 94.350 qs_mol_dyn_low 1 2.0 0.003 0.005 93.923 93.944 qs_forces 11 3.9 0.004 0.010 93.837 93.839 qs_energies 11 4.9 0.002 0.004 90.765 90.786 scf_env_do_scf 11 5.9 0.001 0.002 81.048 81.050 scf_env_do_scf_inner_loop 99 6.5 0.002 0.008 73.439 73.440 dbcsr_multiply_generic 2055 12.4 0.109 0.112 53.640 53.922 qs_scf_new_mos 99 7.5 0.000 0.001 53.413 53.543 qs_scf_loop_do_ot 99 8.5 0.000 0.001 53.413 53.543 ot_scf_mini 99 9.5 0.002 0.002 50.844 50.948 velocity_verlet 10 3.0 0.001 0.002 49.373 49.374 multiply_cannon 2055 13.4 0.179 0.184 43.448 44.322 multiply_cannon_loop 2055 14.4 1.797 1.844 42.020 42.905 ot_mini 99 10.5 0.001 0.002 28.225 28.325 qs_ot_get_derivative 99 11.5 0.001 0.005 21.312 21.436 multiply_cannon_multrec 49320 15.4 11.196 11.909 17.247 17.985 rebuild_ks_matrix 110 8.3 0.000 0.001 16.339 16.479 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.016 16.339 16.478 qs_ks_update_qs_env 110 7.6 0.001 0.001 14.417 14.546 qs_ot_get_p 110 10.4 0.001 0.003 13.579 13.688 mp_waitall_1 220248 16.4 12.774 13.585 12.774 13.585 multiply_cannon_sync_h2d 49320 15.4 9.365 10.050 9.365 10.050 qs_ot_p2m_diag 48 11.0 0.013 0.021 9.442 9.479 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 8.190 8.748 sum_up_and_integrate 110 10.3 0.002 0.004 8.427 8.439 integrate_v_rspace 110 11.3 0.003 0.004 8.401 8.420 cp_dbcsr_syevd 48 12.0 0.002 0.003 8.410 8.411 qs_rho_update_rho_low 110 7.6 0.001 0.002 8.020 8.176 calculate_rho_elec 110 8.6 0.021 0.025 8.020 8.175 multiply_cannon_metrocomm3 49320 15.4 0.081 0.084 6.724 8.008 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.254 7.750 apply_single 110 13.6 0.000 0.001 7.254 7.750 init_scf_loop 11 6.9 0.000 0.000 7.567 7.567 cp_fm_diag_elpa 48 13.0 0.000 0.000 7.513 7.537 cp_fm_diag_elpa_base 48 14.0 7.486 7.514 7.510 7.535 init_scf_run 11 5.9 0.000 0.001 7.418 7.418 scf_env_initial_rho_setup 11 6.9 0.001 0.002 7.417 7.418 ot_diis_step 99 11.5 0.006 0.008 6.693 6.695 dbcsr_mm_accdrv_process 87628 16.1 3.101 3.207 5.922 6.183 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.885 5.949 make_m2s 4110 13.4 0.062 0.065 5.251 5.391 mp_sum_l 6594 12.7 4.498 5.323 4.498 5.323 make_images 4110 14.4 0.179 0.192 5.150 5.292 density_rs2pw 110 9.6 0.004 0.004 5.006 5.231 prepare_preconditioner 11 7.9 0.000 0.000 5.118 5.141 make_preconditioner 11 8.9 0.000 0.000 5.118 5.140 make_full_inverse_cholesky 11 9.9 0.000 0.000 4.860 4.925 pw_transfer 1331 11.6 0.055 0.065 4.729 4.851 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.641 4.766 wfi_extrapolate 11 7.9 0.001 0.004 4.676 4.677 multiply_cannon_metrocomm1 49320 15.4 0.064 0.067 3.006 4.324 calculate_dm_sparse 110 9.5 0.001 0.001 3.980 4.071 fft3d_ps 1111 14.6 0.801 0.897 3.941 4.051 potential_pw2rs 110 12.3 0.005 0.006 3.966 4.025 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.669 3.739 fft_wrap_pw1pw2_140 451 13.1 0.307 0.329 3.595 3.722 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.687 3.692 grid_integrate_task_list 110 12.3 3.224 3.434 3.224 3.434 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.221 3.292 transfer_rs2pw 451 10.6 0.005 0.006 2.814 3.022 mp_alltoall_d11v 2046 13.8 2.478 2.890 2.478 2.890 cp_fm_cholesky_invert 11 10.9 2.831 2.836 2.831 2.836 mp_alltoall_z22v 1111 16.6 2.610 2.799 2.610 2.799 acc_transpose_blocks 49320 15.4 0.213 0.224 2.576 2.647 transfer_pw2rs 451 13.1 0.006 0.006 2.583 2.642 calculate_first_density_matrix 1 7.0 0.001 0.002 2.607 2.610 jit_kernel_multiply 13 15.9 2.541 2.554 2.541 2.554 make_images_data 4110 15.4 0.044 0.048 2.177 2.393 grid_collocate_task_list 110 9.6 2.161 2.261 2.161 2.261 hybrid_alltoall_any 4261 16.3 0.083 0.473 1.863 2.179 mp_waitany 14300 13.8 1.911 2.116 1.911 2.116 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.060 2.079 mp_sum_d 3891 11.9 1.538 2.001 1.538 2.001 make_images_sizes 4110 15.4 0.004 0.005 1.487 1.961 mp_alltoall_i44 4110 16.4 1.483 1.956 1.483 1.956 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=94.350000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=478.090909, yerr=3.752685 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bbd14a5fe6b002419d127c8ec63e3062d934e058_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 594.952192E+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 2509598. 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 0.024 0.050 78.120 78.122 qs_mol_dyn_low 1 2.0 0.004 0.008 77.725 77.734 qs_forces 11 3.9 0.006 0.008 77.628 77.630 qs_energies 11 4.9 0.004 0.010 74.058 74.064 scf_env_do_scf 11 5.9 0.001 0.001 64.868 64.871 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 55.536 55.538 velocity_verlet 10 3.0 0.003 0.018 40.496 40.498 dbcsr_multiply_generic 2055 12.4 0.117 0.121 39.560 39.776 qs_scf_new_mos 99 7.5 0.001 0.001 37.360 37.506 qs_scf_loop_do_ot 99 8.5 0.001 0.001 37.359 37.506 ot_scf_mini 99 9.5 0.003 0.004 35.624 35.753 multiply_cannon 2055 13.4 0.224 0.245 31.699 32.907 multiply_cannon_loop 2055 14.4 1.160 1.191 30.168 31.246 ot_mini 99 10.5 0.001 0.001 19.652 19.793 multiply_cannon_multrec 24660 15.4 6.956 8.744 14.047 15.928 rebuild_ks_matrix 110 8.3 0.000 0.000 14.955 15.130 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.019 14.955 15.130 qs_ot_get_derivative 99 11.5 0.001 0.001 13.784 13.915 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.163 13.317 mp_waitall_1 176588 16.5 8.626 11.189 8.626 11.189 init_scf_loop 11 6.9 0.001 0.003 9.283 9.288 qs_ot_get_p 110 10.4 0.001 0.003 9.082 9.234 multiply_cannon_metrocomm3 24660 15.4 0.074 0.075 5.131 7.766 sum_up_and_integrate 110 10.3 0.002 0.003 7.604 7.615 integrate_v_rspace 110 11.3 0.002 0.003 7.576 7.589 multiply_cannon_sync_h2d 24660 15.4 6.439 7.530 6.439 7.530 dbcsr_mm_accdrv_process 52282 16.1 5.418 6.267 6.923 7.241 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.187 7.200 calculate_rho_elec 110 8.6 0.039 0.047 7.187 7.200 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.511 7.121 apply_single 110 13.6 0.000 0.001 6.511 7.121 prepare_preconditioner 11 7.9 0.000 0.000 7.087 7.115 make_preconditioner 11 8.9 0.001 0.003 7.087 7.115 make_full_inverse_cholesky 11 9.9 0.000 0.001 6.618 6.820 qs_ot_p2m_diag 48 11.0 0.029 0.044 6.667 6.690 init_scf_run 11 5.9 0.000 0.001 6.629 6.630 scf_env_initial_rho_setup 11 6.9 0.001 0.002 6.629 6.630 cp_dbcsr_syevd 48 12.0 0.003 0.003 6.135 6.136 ot_diis_step 99 11.5 0.010 0.010 5.816 5.817 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 5.041 5.748 make_m2s 4110 13.4 0.056 0.059 5.207 5.686 make_images 4110 14.4 0.404 0.460 5.093 5.568 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.315 5.330 cp_fm_diag_elpa_base 48 14.0 5.244 5.278 5.311 5.327 pw_transfer 1331 11.6 0.066 0.073 4.514 4.696 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 4.407 4.591 density_rs2pw 110 9.6 0.004 0.004 4.352 4.525 wfi_extrapolate 11 7.9 0.001 0.001 3.955 3.955 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.714 3.785 fft3d_ps 1111 14.6 1.119 1.358 3.505 3.710 fft_wrap_pw1pw2_140 451 13.1 0.358 0.379 3.318 3.506 potential_pw2rs 110 12.3 0.008 0.009 3.341 3.405 cp_fm_cholesky_invert 11 10.9 3.394 3.402 3.394 3.402 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.392 3.398 grid_integrate_task_list 110 12.3 3.130 3.363 3.130 3.363 make_images_data 4110 15.4 0.048 0.053 2.697 3.241 hybrid_alltoall_any 4261 16.3 0.104 0.448 2.352 3.170 calculate_dm_sparse 110 9.5 0.001 0.001 3.056 3.083 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.954 3.015 mp_sum_l 6594 12.7 2.050 2.812 2.050 2.812 mp_alltoall_d11v 2046 13.8 2.294 2.671 2.294 2.671 calculate_first_density_matrix 1 7.0 0.002 0.010 2.560 2.564 transfer_rs2pw 451 10.6 0.006 0.007 2.216 2.394 grid_collocate_task_list 110 9.6 2.168 2.303 2.168 2.303 mp_alltoall_z22v 1111 16.6 2.042 2.162 2.042 2.162 qs_ot_get_orbitals 99 10.5 0.001 0.001 2.099 2.110 acc_transpose_blocks 24660 15.4 0.116 0.119 1.986 2.014 transfer_pw2rs 451 13.1 0.005 0.006 1.950 2.005 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.961 1.979 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.953 1.957 jit_kernel_multiply 10 16.3 1.136 1.950 1.136 1.950 cp_fm_cholesky_decompose 22 10.9 1.925 1.934 1.925 1.934 mp_allgather_i34 2055 14.4 0.840 1.918 0.840 1.918 dbcsr_complete_redistribute 325 12.2 0.230 0.293 1.490 1.783 mp_waitany 10164 13.8 1.374 1.592 1.374 1.592 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.582 1.592 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=78.122000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=561.545455, yerr=6.919848 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bbd14a5fe6b002419d127c8ec63e3062d934e058_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 667.840512E+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 0.009 0.028 61.313 61.314 qs_mol_dyn_low 1 2.0 0.003 0.003 61.030 61.041 qs_forces 11 3.9 0.003 0.003 60.961 60.961 qs_energies 11 4.9 0.002 0.002 57.750 57.754 scf_env_do_scf 11 5.9 0.001 0.001 49.730 49.730 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 40.880 40.881 velocity_verlet 10 3.0 0.001 0.001 33.244 33.246 dbcsr_multiply_generic 2055 12.4 0.110 0.114 28.991 29.211 qs_scf_new_mos 99 7.5 0.001 0.001 25.933 26.012 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.933 26.012 ot_scf_mini 99 9.5 0.002 0.003 24.672 24.779 multiply_cannon 2055 13.4 0.211 0.219 22.564 23.739 multiply_cannon_loop 2055 14.4 0.815 0.838 21.334 22.330 ot_mini 99 10.5 0.001 0.001 13.924 14.036 rebuild_ks_matrix 110 8.3 0.000 0.000 12.340 12.452 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.015 12.339 12.451 multiply_cannon_multrec 16440 15.4 3.764 5.044 9.992 11.169 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.856 10.955 mp_waitall_1 139946 16.5 6.952 10.057 6.952 10.057 qs_ot_get_derivative 99 11.5 0.001 0.001 9.508 9.619 init_scf_loop 11 6.9 0.000 0.000 8.813 8.814 multiply_cannon_metrocomm3 16440 15.4 0.045 0.047 4.174 7.110 prepare_preconditioner 11 7.9 0.000 0.000 7.044 7.059 make_preconditioner 11 8.9 0.000 0.000 7.044 7.059 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.366 6.723 sum_up_and_integrate 110 10.3 0.002 0.003 6.449 6.465 integrate_v_rspace 110 11.3 0.003 0.003 6.423 6.440 dbcsr_mm_accdrv_process 34862 16.1 5.323 5.768 6.073 6.196 qs_ot_get_p 110 10.4 0.001 0.002 5.879 6.052 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.935 5.943 calculate_rho_elec 110 8.6 0.059 0.059 5.934 5.943 init_scf_run 11 5.9 0.000 0.001 5.592 5.592 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.592 5.592 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.857 5.360 apply_single 110 13.6 0.000 0.000 4.856 5.360 make_m2s 4110 13.4 0.049 0.051 4.353 4.753 make_images 4110 14.4 0.399 0.520 4.236 4.636 ot_diis_step 99 11.5 0.011 0.011 4.387 4.388 qs_ot_p2m_diag 48 11.0 0.042 0.044 4.138 4.142 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.308 3.895 multiply_cannon_sync_h2d 16440 15.4 3.239 3.844 3.239 3.844 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.742 3.742 pw_transfer 1331 11.6 0.065 0.072 3.730 3.739 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.623 3.635 grid_integrate_task_list 110 12.3 3.191 3.365 3.191 3.365 density_rs2pw 110 9.6 0.004 0.004 3.153 3.306 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.175 3.184 cp_fm_diag_elpa_base 48 14.0 3.105 3.141 3.173 3.182 fft_wrap_pw1pw2_140 451 13.1 0.460 0.469 3.081 3.093 wfi_extrapolate 11 7.9 0.001 0.001 3.067 3.067 make_images_data 4110 15.4 0.046 0.051 2.398 2.937 hybrid_alltoall_any 4261 16.3 0.107 0.380 2.087 2.858 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.844 2.845 cp_fm_cholesky_invert 11 10.9 2.837 2.844 2.837 2.844 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.514 2.581 calculate_dm_sparse 110 9.5 0.001 0.001 2.536 2.568 fft3d_ps 1111 14.6 1.120 1.133 2.542 2.556 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.438 2.486 calculate_first_density_matrix 1 7.0 0.000 0.001 2.434 2.435 grid_collocate_task_list 110 9.6 2.219 2.395 2.219 2.395 multiply_cannon_metrocomm4 14385 15.4 0.048 0.051 0.863 2.359 potential_pw2rs 110 12.3 0.011 0.011 2.263 2.273 mp_irecv_dv 48980 15.7 0.788 2.225 0.788 2.225 mp_sum_l 6594 12.7 1.447 2.101 1.447 2.101 mp_alltoall_d11v 2046 13.8 1.836 2.058 1.836 2.058 dbcsr_complete_redistribute 325 12.2 0.390 0.415 1.559 2.027 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.962 1.963 cp_fm_upper_to_full 70 14.2 1.418 1.806 1.418 1.806 cp_fm_cholesky_decompose 22 10.9 1.689 1.710 1.689 1.710 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.677 1.692 acc_transpose_blocks 16440 15.4 0.079 0.081 1.588 1.647 mp_allgather_i34 2055 14.4 0.543 1.622 0.543 1.622 transfer_rs2pw 451 10.6 0.005 0.006 1.372 1.559 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.075 1.530 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.474 1.487 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.366 1.467 mp_waitany 17072 13.8 1.149 1.385 1.149 1.385 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.307 1.317 transfer_pw2rs 451 13.1 0.005 0.005 1.264 1.271 mp_alltoall_z22v 1111 16.6 1.209 1.233 1.209 1.233 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=61.314000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=632.727273, yerr=7.349594 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bbd14a5fe6b002419d127c8ec63e3062d934e058_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 740.552704E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66419. MP_Allreduce 9774 603. MP_Sync 52 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.009 0.028 66.512 66.512 qs_mol_dyn_low 1 2.0 0.003 0.003 66.275 66.284 qs_forces 11 3.9 0.003 0.003 66.203 66.203 qs_energies 11 4.9 0.002 0.002 62.804 62.808 scf_env_do_scf 11 5.9 0.001 0.001 54.397 54.400 scf_env_do_scf_inner_loop 99 6.5 0.003 0.006 42.490 42.491 velocity_verlet 10 3.0 0.002 0.002 37.875 37.877 dbcsr_multiply_generic 2055 12.4 0.118 0.121 30.802 31.004 qs_scf_new_mos 99 7.5 0.001 0.001 27.740 27.858 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.740 27.858 ot_scf_mini 99 9.5 0.002 0.003 26.046 26.157 multiply_cannon 2055 13.4 0.236 0.249 23.667 24.348 multiply_cannon_loop 2055 14.4 1.423 1.486 22.129 22.687 ot_mini 99 10.5 0.001 0.001 14.890 15.026 multiply_cannon_multrec 24660 15.4 4.109 6.876 13.072 14.181 rebuild_ks_matrix 110 8.3 0.000 0.000 12.059 12.156 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.015 12.058 12.155 init_scf_loop 11 6.9 0.000 0.000 11.866 11.867 qs_ot_get_derivative 99 11.5 0.001 0.001 10.728 10.846 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.658 10.743 prepare_preconditioner 11 7.9 0.000 0.000 10.113 10.132 make_preconditioner 11 8.9 0.000 0.000 10.113 10.132 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.358 9.788 dbcsr_mm_accdrv_process 52304 16.0 7.793 9.113 8.807 9.740 sum_up_and_integrate 110 10.3 0.002 0.002 6.305 6.318 integrate_v_rspace 110 11.3 0.003 0.003 6.279 6.291 qs_ot_get_p 110 10.4 0.002 0.002 6.021 6.188 mp_waitall_1 121746 16.5 4.133 5.952 4.133 5.952 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.925 5.933 calculate_rho_elec 110 8.6 0.077 0.081 5.924 5.932 make_m2s 4110 13.4 0.059 0.061 5.446 5.894 make_images 4110 14.4 0.574 0.695 5.306 5.750 init_scf_run 11 5.9 0.000 0.001 5.666 5.666 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.665 5.666 cp_fm_upper_to_full 70 14.2 3.342 4.830 3.342 4.830 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.050 4.146 apply_single 110 13.6 0.000 0.000 4.049 4.146 ot_diis_step 99 11.5 0.011 0.012 4.125 4.125 qs_ot_p2m_diag 48 11.0 0.055 0.064 4.098 4.115 pw_transfer 1331 11.6 0.065 0.072 3.833 3.866 dbcsr_complete_redistribute 325 12.2 0.421 0.457 2.674 3.793 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.726 3.763 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.569 3.625 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.530 3.531 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.386 3.450 grid_integrate_task_list 110 12.3 3.298 3.420 3.298 3.420 density_rs2pw 110 9.6 0.004 0.004 3.051 3.261 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.125 3.233 fft_wrap_pw1pw2_140 451 13.1 0.531 0.544 3.173 3.215 make_images_data 4110 15.4 0.048 0.052 2.659 3.114 wfi_extrapolate 11 7.9 0.001 0.001 3.105 3.106 hybrid_alltoall_any 4261 16.3 0.123 0.463 2.316 3.104 calculate_dm_sparse 110 9.5 0.001 0.001 2.995 3.029 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.984 2.995 cp_fm_diag_elpa_base 48 14.0 2.837 2.898 2.982 2.993 multiply_cannon_metrocomm3 24660 15.4 0.037 0.038 1.294 2.926 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.903 2.906 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.747 2.839 cp_fm_cholesky_invert 11 10.9 2.818 2.826 2.818 2.826 mp_alltoall_i22 605 13.7 1.638 2.795 1.638 2.795 acc_transpose_blocks 24660 15.4 0.113 0.118 2.423 2.583 fft3d_ps 1111 14.6 1.153 1.194 2.552 2.569 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.512 2.545 multiply_cannon_sync_h2d 24660 15.4 2.390 2.518 2.390 2.518 calculate_first_density_matrix 1 7.0 0.000 0.001 2.463 2.466 grid_collocate_task_list 110 9.6 2.263 2.410 2.263 2.410 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 2.244 2.245 potential_pw2rs 110 12.3 0.013 0.013 2.101 2.110 mp_alltoall_d11v 2046 13.8 1.792 1.981 1.792 1.981 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.810 1.845 cp_fm_cholesky_decompose 22 10.9 1.790 1.839 1.790 1.839 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.613 1.707 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.641 1.652 jit_kernel_multiply 9 15.8 0.665 1.632 0.665 1.632 mp_sum_l 6594 12.7 0.996 1.602 0.996 1.602 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.542 1.558 acc_transpose_blocks_sync 73980 16.4 1.407 1.551 1.407 1.551 transfer_rs2pw 451 10.6 0.005 0.006 1.213 1.449 multiply_cannon_metrocomm4 20550 15.4 0.060 0.062 0.830 1.442 mp_irecv_dv 62702 16.1 0.728 1.358 0.728 1.358 mp_waitany 13376 13.8 1.022 1.342 1.022 1.342 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=66.512000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=699.818182, yerr=11.722691 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bbd14a5fe6b002419d127c8ec63e3062d934e058_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 856.178688E+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 0.010 0.029 56.699 56.700 qs_mol_dyn_low 1 2.0 0.003 0.003 56.073 56.451 qs_forces 11 3.9 0.003 0.003 55.857 55.858 qs_energies 11 4.9 0.013 0.015 52.190 52.196 scf_env_do_scf 11 5.9 0.000 0.001 43.862 43.862 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 35.859 35.860 velocity_verlet 10 3.0 0.042 0.133 31.795 31.798 dbcsr_multiply_generic 2055 12.4 0.106 0.108 23.481 23.666 qs_scf_new_mos 99 7.5 0.001 0.001 21.113 21.155 qs_scf_loop_do_ot 99 8.5 0.001 0.001 21.112 21.154 ot_scf_mini 99 9.5 0.002 0.002 19.847 19.871 multiply_cannon 2055 13.4 0.238 0.249 17.805 19.090 multiply_cannon_loop 2055 14.4 0.604 0.626 16.518 16.874 rebuild_ks_matrix 110 8.3 0.000 0.000 11.778 11.801 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.778 11.801 ot_mini 99 10.5 0.001 0.001 10.939 10.955 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.480 10.501 multiply_cannon_multrec 8220 15.4 3.188 4.293 7.683 8.572 init_scf_loop 11 6.9 0.000 0.000 7.957 7.959 mp_waitall_1 103326 16.6 5.966 7.554 5.966 7.554 qs_ot_get_derivative 99 11.5 0.001 0.001 7.250 7.272 sum_up_and_integrate 110 10.3 0.001 0.002 6.303 6.315 prepare_preconditioner 11 7.9 0.000 0.000 6.286 6.289 make_preconditioner 11 8.9 0.000 0.000 6.286 6.289 integrate_v_rspace 110 11.3 0.003 0.003 6.276 6.288 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.049 6.061 calculate_rho_elec 110 8.6 0.112 0.113 6.048 6.061 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.850 5.923 dbcsr_mm_accdrv_process 17442 15.9 3.156 4.142 4.356 5.292 init_scf_run 11 5.9 0.000 0.001 5.175 5.175 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.175 5.175 qs_ot_get_p 110 10.4 0.002 0.002 4.929 4.955 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 3.015 4.594 make_m2s 4110 13.4 0.037 0.038 4.284 4.565 make_images 4110 14.4 0.645 0.706 4.154 4.436 pw_transfer 1331 11.6 0.065 0.070 4.161 4.174 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.054 4.070 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.720 3.763 apply_single 110 13.6 0.000 0.000 3.720 3.763 ot_diis_step 99 11.5 0.012 0.012 3.666 3.667 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.513 3.517 fft_wrap_pw1pw2_140 451 13.1 0.717 0.731 3.499 3.513 grid_integrate_task_list 110 12.3 3.373 3.506 3.373 3.506 density_rs2pw 110 9.6 0.004 0.004 3.082 3.290 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.166 3.167 cp_fm_cholesky_invert 11 10.9 3.042 3.046 3.042 3.046 wfi_extrapolate 11 7.9 0.001 0.001 2.749 2.749 hybrid_alltoall_any 4261 16.3 0.199 0.847 2.251 2.714 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.682 2.683 make_images_data 4110 15.4 0.040 0.046 2.301 2.669 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.641 2.648 cp_fm_diag_elpa_base 48 14.0 2.584 2.607 2.639 2.646 fft3d_ps 1111 14.6 1.292 1.302 2.601 2.613 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.581 2.582 calculate_dm_sparse 110 9.5 0.001 0.001 2.486 2.530 grid_collocate_task_list 110 9.6 2.363 2.473 2.363 2.473 multiply_cannon_sync_h2d 8220 15.4 2.372 2.440 2.372 2.440 calculate_first_density_matrix 1 7.0 0.000 0.000 2.329 2.331 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.208 2.222 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.083 2.111 potential_pw2rs 110 12.3 0.015 0.016 2.088 2.091 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.793 1.996 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.951 1.958 mp_alltoall_d11v 2046 13.8 1.666 1.849 1.666 1.849 cp_fm_cholesky_decompose 22 10.9 1.789 1.803 1.789 1.803 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.628 1.630 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.507 1.626 mp_allgather_i34 2055 14.4 0.457 1.568 0.457 1.568 dbcsr_complete_redistribute 325 12.2 0.564 0.594 1.464 1.561 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.425 1.441 acc_transpose_blocks 8220 15.4 0.041 0.042 1.368 1.394 transfer_rs2pw 451 10.6 0.005 0.005 1.134 1.370 qs_create_task_list 11 7.9 0.000 0.001 1.215 1.318 generate_qs_task_list 11 8.9 0.375 0.442 1.214 1.317 mp_waitany 9240 13.8 1.061 1.306 1.061 1.306 multiply_cannon_metrocomm1 8220 15.4 0.022 0.023 0.790 1.295 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.147 1.163 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=56.700000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=808.272727, yerr=11.947544 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bbd14a5fe6b002419d127c8ec63e3062d934e058_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 1.408066E+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 0.014 0.034 89.519 89.520 qs_mol_dyn_low 1 2.0 0.003 0.003 89.217 89.227 qs_forces 11 3.9 0.003 0.004 89.133 89.134 qs_energies 11 4.9 0.002 0.002 84.969 84.971 scf_env_do_scf 11 5.9 0.000 0.001 74.826 74.827 velocity_verlet 10 3.0 0.002 0.002 57.239 57.247 scf_env_do_scf_inner_loop 99 6.5 0.004 0.007 45.724 45.727 dbcsr_multiply_generic 2055 12.4 0.121 0.125 29.869 29.968 init_scf_loop 11 6.9 0.000 0.000 29.027 29.031 qs_scf_new_mos 99 7.5 0.001 0.001 27.430 27.473 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.429 27.473 prepare_preconditioner 11 7.9 0.000 0.000 26.965 26.976 make_preconditioner 11 8.9 0.000 0.000 26.965 26.976 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.875 26.450 ot_scf_mini 99 9.5 0.002 0.002 25.604 25.642 multiply_cannon 2055 13.4 0.337 0.370 22.678 23.422 multiply_cannon_loop 2055 14.4 0.815 0.824 20.868 21.137 cp_fm_upper_to_full 70 14.2 12.811 18.654 12.811 18.654 ot_mini 99 10.5 0.001 0.001 14.376 14.413 rebuild_ks_matrix 110 8.3 0.000 0.001 14.214 14.265 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 14.213 14.265 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.897 12.944 dbcsr_complete_redistribute 325 12.2 1.066 1.084 7.753 11.113 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.716 10.080 multiply_cannon_multrec 8220 15.4 4.055 4.215 9.737 9.843 qs_ot_get_derivative 99 11.5 0.001 0.001 9.712 9.749 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.075 9.401 mp_alltoall_i22 605 13.7 5.718 9.110 5.718 9.110 mp_waitall_1 84994 16.7 7.794 8.696 7.794 8.696 qs_rho_update_rho_low 110 7.6 0.000 0.001 7.642 7.676 calculate_rho_elec 110 8.6 0.222 0.223 7.641 7.675 sum_up_and_integrate 110 10.3 0.002 0.002 7.102 7.113 integrate_v_rspace 110 11.3 0.003 0.003 7.074 7.085 pw_transfer 1331 11.6 0.074 0.075 5.919 5.929 make_m2s 4110 13.4 0.042 0.043 5.391 5.889 qs_ot_get_p 110 10.4 0.002 0.002 5.840 5.889 init_scf_run 11 5.9 0.000 0.001 5.847 5.847 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.846 5.847 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 5.802 5.812 dbcsr_mm_accdrv_process 11614 15.7 3.888 4.115 5.534 5.731 make_images 4110 14.4 0.883 0.936 5.202 5.699 cp_fm_cholesky_invert 11 10.9 5.473 5.477 5.473 5.477 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 5.101 5.383 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.801 5.270 apply_single 110 13.6 0.000 0.000 4.800 5.270 fft_wrap_pw1pw2_140 451 13.1 1.366 1.369 5.074 5.087 ot_diis_step 99 11.5 0.015 0.015 4.641 4.641 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.189 4.197 density_rs2pw 110 9.6 0.004 0.004 4.052 4.082 grid_integrate_task_list 110 12.3 3.724 3.822 3.724 3.822 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.680 3.681 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.667 3.667 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.094 3.552 hybrid_alltoall_any 4261 16.3 0.262 0.559 2.884 3.543 fft3d_ps 1111 14.6 1.881 1.891 3.509 3.518 make_images_data 4110 15.4 0.045 0.047 2.815 3.473 wfi_extrapolate 11 7.9 0.001 0.001 3.410 3.410 calculate_dm_sparse 110 9.5 0.001 0.001 3.198 3.229 multiply_cannon_sync_h2d 8220 15.4 3.141 3.156 3.141 3.156 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.070 3.071 cp_fm_diag_elpa_base 48 14.0 2.522 2.734 3.068 3.069 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.970 2.974 grid_collocate_task_list 110 9.6 2.703 2.738 2.703 2.738 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.707 2.726 potential_pw2rs 110 12.3 0.021 0.022 2.517 2.521 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.396 2.414 calculate_first_density_matrix 1 7.0 0.000 0.000 2.317 2.318 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.200 2.264 mp_alltoall_d11v 2046 13.8 2.162 2.257 2.162 2.257 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.102 2.175 cp_fm_cholesky_decompose 22 10.9 2.088 2.098 2.088 2.098 qs_create_task_list 11 7.9 0.001 0.001 1.880 1.928 generate_qs_task_list 11 8.9 0.735 0.788 1.880 1.927 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.916 1.921 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.786 1.819 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=89.520000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1267.909091, yerr=62.593137 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bbd14a5fe6b002419d127c8ec63e3062d934e058_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 634.155008E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175954870160 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 57905. MP_Allreduce 11059 797. MP_Sync 87 MP_Alltoall 2226 2529110. MP_SendRecv 24320 18752. MP_ISendRecv 24320 18752. MP_Wait 42476 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.030 211.263 211.264 qs_mol_dyn_low 1 2.0 0.003 0.003 210.826 210.840 qs_forces 11 3.9 0.005 0.005 210.736 210.737 qs_energies 11 4.9 0.002 0.003 205.021 205.041 scf_env_do_scf 11 5.9 0.001 0.002 188.189 188.193 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 167.039 167.041 dbcsr_multiply_generic 2507 12.6 0.178 0.182 126.899 127.653 qs_scf_new_mos 117 7.6 0.001 0.001 127.104 127.436 qs_scf_loop_do_ot 117 8.6 0.001 0.001 127.103 127.435 velocity_verlet 10 3.0 0.001 0.002 126.447 126.448 ot_scf_mini 117 9.6 0.003 0.003 120.372 120.710 multiply_cannon 2507 13.6 0.236 0.245 102.196 103.636 multiply_cannon_loop 2507 14.6 2.412 2.467 99.898 101.447 ot_mini 117 10.6 0.001 0.001 67.080 67.422 multiply_cannon_multrec 60168 15.6 31.663 33.876 41.493 44.039 qs_ot_get_derivative 117 11.6 0.001 0.001 42.232 42.541 rebuild_ks_matrix 128 8.3 0.001 0.001 33.865 34.251 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.019 33.865 34.251 mp_waitall_1 267128 16.5 29.871 32.419 29.871 32.419 qs_ot_get_p 128 10.4 0.001 0.002 30.907 31.205 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.395 30.764 multiply_cannon_sync_h2d 60168 15.6 26.156 28.450 26.156 28.450 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.356 25.238 apply_single 128 13.6 0.001 0.001 24.355 25.238 ot_diis_step 117 11.6 0.008 0.008 24.591 24.593 qs_ot_p2m_diag 83 11.4 0.079 0.091 24.095 24.164 cp_dbcsr_syevd 83 12.4 0.005 0.005 21.165 21.166 init_scf_loop 11 6.9 0.000 0.001 21.064 21.065 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 20.433 20.676 multiply_cannon_metrocomm3 60168 15.6 0.115 0.120 16.507 18.625 cp_fm_diag_elpa 83 13.4 0.000 0.000 18.112 18.143 cp_fm_diag_elpa_base 83 14.4 18.029 18.076 18.107 18.140 prepare_preconditioner 11 7.9 0.000 0.000 16.443 16.490 make_preconditioner 11 8.9 0.000 0.000 16.443 16.490 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.677 15.849 make_m2s 5014 13.6 0.104 0.112 14.012 14.366 sum_up_and_integrate 128 10.3 0.002 0.005 14.280 14.294 integrate_v_rspace 128 11.3 0.003 0.004 14.222 14.236 make_images 5014 14.6 0.399 0.419 13.827 14.194 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.457 13.565 calculate_rho_elec 128 8.7 0.045 0.064 13.456 13.564 init_scf_run 11 5.9 0.000 0.001 12.656 12.657 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.656 12.656 mp_sum_l 7950 12.9 9.432 10.752 9.432 10.752 dbcsr_mm_accdrv_process 124484 16.2 4.767 4.917 9.401 9.895 cp_fm_cholesky_invert 11 10.9 9.453 9.461 9.453 9.461 wfi_extrapolate 11 7.9 0.001 0.001 9.177 9.177 calculate_dm_sparse 128 9.5 0.001 0.001 8.775 8.864 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.312 8.445 qs_ot_get_orbitals 117 10.6 0.001 0.001 8.308 8.390 multiply_cannon_metrocomm1 60168 15.6 0.092 0.096 6.588 8.300 pw_transfer 1547 11.6 0.073 0.093 7.867 8.065 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.667 7.867 make_images_data 5014 15.6 0.066 0.071 6.877 7.821 density_rs2pw 128 9.7 0.006 0.007 7.080 7.503 grid_integrate_task_list 128 12.3 7.046 7.495 7.046 7.495 hybrid_alltoall_any 5200 16.5 0.292 2.261 6.024 7.327 fft_wrap_pw1pw2_140 523 13.2 0.856 0.895 6.638 6.829 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.675 6.687 fft3d_ps 1291 14.7 2.196 2.843 5.676 6.011 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.747 5.861 mp_alltoall_d11v 2415 14.1 4.486 5.402 4.486 5.402 grid_collocate_task_list 128 9.7 4.841 5.343 4.841 5.343 cp_fm_cholesky_decompose 22 10.9 4.769 4.782 4.769 4.782 potential_pw2rs 128 12.3 0.009 0.010 4.742 4.770 mp_sum_d 4467 12.1 3.495 4.293 3.495 4.293 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=211.264000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=599.272727, yerr=6.876718 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bbd14a5fe6b002419d127c8ec63e3062d934e058_performance_tests/16/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 2.200017E+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 6024768 0.0% 0.0% 100.0% average stack size 0.0 0.0 1129.7 marketing flops 145.651870E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 842.657792E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2427840 MPI messages size (bytes): total size 4.132588E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.702167E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 71532 2339635200 32768 < size <= 131072 729952 56049532928 131072 < size <= 4194304 1387568 1410045313024 4194304 < size <= 16777216 155760 1473828901424 16777216 < size 68112 1190343475200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 57636. MP_Allreduce 11185 956. MP_Sync 88 MP_Alltoall 1983 5733331. MP_SendRecv 12126 47072. MP_ISendRecv 12126 47072. MP_Wait 26114 MP_ISend 11836 212447. MP_IRecv 11836 212447. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.030 211.924 211.925 qs_mol_dyn_low 1 2.0 0.003 0.003 211.551 211.566 qs_forces 11 3.9 0.004 0.005 211.462 211.463 qs_energies 11 4.9 0.002 0.004 204.501 204.516 scf_env_do_scf 11 5.9 0.001 0.002 187.497 187.507 scf_env_do_scf_inner_loop 118 6.6 0.004 0.012 149.298 149.302 velocity_verlet 10 3.0 0.001 0.002 133.388 133.390 qs_scf_new_mos 118 7.6 0.001 0.001 107.615 108.150 qs_scf_loop_do_ot 118 8.6 0.001 0.001 107.614 108.149 dbcsr_multiply_generic 2529 12.6 0.193 0.200 101.735 103.537 ot_scf_mini 118 9.6 0.004 0.004 102.630 103.332 multiply_cannon 2529 13.6 0.509 0.571 78.756 85.092 multiply_cannon_loop 2529 14.6 1.579 1.655 74.910 79.987 ot_mini 118 10.6 0.001 0.001 53.882 54.524 mp_waitall_1 216598 16.6 27.072 41.119 27.072 41.119 init_scf_loop 11 6.9 0.000 0.000 38.095 38.097 multiply_cannon_multrec 30348 15.6 20.663 25.072 31.322 35.969 rebuild_ks_matrix 129 8.3 0.001 0.001 34.070 34.739 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.021 34.069 34.738 prepare_preconditioner 11 7.9 0.000 0.000 33.596 33.662 make_preconditioner 11 8.9 0.000 0.000 33.596 33.662 make_full_inverse_cholesky 11 9.9 0.000 0.000 32.230 32.827 multiply_cannon_metrocomm3 30348 15.6 0.099 0.109 16.970 32.371 qs_ot_get_derivative 118 11.6 0.001 0.002 31.292 31.986 qs_ks_update_qs_env 129 7.6 0.001 0.001 30.705 31.304 qs_ot_get_p 129 10.4 0.001 0.003 30.221 30.879 qs_ot_p2m_diag 84 11.4 0.191 0.220 24.508 24.609 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 22.519 23.904 apply_single 129 13.6 0.001 0.001 22.519 23.903 cp_dbcsr_syevd 84 12.4 0.006 0.006 23.048 23.050 ot_diis_step 118 11.6 0.014 0.015 22.350 22.352 cp_fm_cholesky_invert 11 10.9 19.990 20.006 19.990 20.006 multiply_cannon_sync_h2d 30348 15.6 17.727 19.891 17.727 19.891 cp_fm_diag_elpa 84 13.4 0.000 0.000 19.158 19.192 cp_fm_diag_elpa_base 84 14.4 18.846 18.980 19.151 19.179 make_m2s 5058 13.6 0.086 0.092 15.571 17.222 make_images 5058 14.6 1.174 1.381 15.353 17.005 sum_up_and_integrate 129 10.3 0.002 0.005 15.489 15.520 integrate_v_rspace 129 11.3 0.003 0.003 15.428 15.462 qs_rho_update_rho_low 129 7.7 0.001 0.001 15.006 15.033 calculate_rho_elec 129 8.7 0.087 0.104 15.005 15.033 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 12.799 13.268 init_scf_run 11 5.9 0.000 0.001 12.067 12.069 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.067 12.068 multiply_cannon_metrocomm4 27819 15.6 0.102 0.117 3.977 11.143 make_images_data 5058 15.6 0.066 0.076 9.026 10.934 mp_irecv_dv 70084 16.3 3.776 10.734 3.776 10.734 dbcsr_mm_accdrv_process 62758 16.2 5.462 6.410 10.133 10.707 hybrid_alltoall_any 5245 16.5 0.352 1.558 7.698 10.292 pw_transfer 1559 11.6 0.085 0.099 9.377 9.441 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 9.153 9.219 density_rs2pw 129 9.7 0.006 0.006 8.589 8.986 wfi_extrapolate 11 7.9 0.001 0.001 8.889 8.889 cp_fm_cholesky_decompose 22 10.9 8.671 8.758 8.671 8.758 mp_sum_l 8016 12.9 5.609 7.757 5.609 7.757 fft_wrap_pw1pw2_140 527 13.2 0.939 0.965 7.665 7.738 grid_integrate_task_list 129 12.3 7.185 7.643 7.185 7.643 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.762 7.596 calculate_dm_sparse 129 9.5 0.001 0.001 6.664 6.810 fft3d_ps 1301 14.7 2.849 3.041 6.701 6.773 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.200 6.209 qs_ot_get_orbitals 118 10.6 0.001 0.001 5.782 5.885 potential_pw2rs 129 12.3 0.016 0.019 5.827 5.865 mp_alltoall_d11v 2429 14.1 4.658 5.701 4.658 5.701 grid_collocate_task_list 129 9.7 5.052 5.593 5.052 5.593 mp_allgather_i34 2529 14.6 2.354 5.521 2.354 5.521 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.225 5.408 dbcsr_complete_redistribute 397 12.7 0.768 0.865 3.543 4.469 transfer_rs2pw 527 10.6 0.007 0.007 4.000 4.397 mp_sum_d 4511 12.1 2.899 4.251 2.899 4.251 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=211.925000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=802.636364, yerr=3.255003 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bbd14a5fe6b002419d127c8ec63e3062d934e058_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430458527744 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613083000832 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228655E+12 0.0% 0.0% 100.0% flops max/rank 2.950886E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806314816 0.0% 0.0% 100.0% number of processed stacks 4015680 0.0% 0.0% 100.0% average stack size 0.0 0.0 1694.9 marketing flops 145.646636E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 964.063232E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1051232 MPI messages size (bytes): total size 2.737039E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.603648E+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 284160 37245419520 131072 < size <= 4194304 665408 1004401590272 4194304 < size <= 16777216 66080 937889582784 16777216 < size 28896 757491302400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4040 57672. MP_Allreduce 11177 996. MP_Sync 87 MP_Alltoall 1724 9394160. MP_SendRecv 7998 75008. MP_ISendRecv 7998 75008. MP_Wait 21986 MP_ISend 11836 275177. MP_IRecv 11836 275177. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.049 0.127 176.255 176.257 qs_mol_dyn_low 1 2.0 0.003 0.003 175.714 175.729 qs_forces 11 3.9 0.004 0.005 175.620 175.623 qs_energies 11 4.9 0.002 0.007 169.116 169.125 scf_env_do_scf 11 5.9 0.001 0.002 153.569 153.571 scf_env_do_scf_inner_loop 118 6.6 0.007 0.020 118.543 118.544 velocity_verlet 10 3.0 0.002 0.002 113.227 113.229 dbcsr_multiply_generic 2527 12.6 0.186 0.191 82.763 83.837 qs_scf_new_mos 118 7.6 0.001 0.001 81.066 81.450 qs_scf_loop_do_ot 118 8.6 0.001 0.001 81.066 81.449 ot_scf_mini 118 9.6 0.004 0.004 76.751 77.127 multiply_cannon 2527 13.6 0.510 0.535 62.875 67.717 multiply_cannon_loop 2527 14.6 1.137 1.214 59.784 62.716 ot_mini 118 10.6 0.001 0.001 42.878 43.262 init_scf_loop 11 6.9 0.001 0.004 34.929 34.930 mp_waitall_1 171878 16.6 24.934 34.140 24.934 34.140 rebuild_ks_matrix 129 8.3 0.001 0.001 30.507 30.952 qs_ks_build_kohn_sham_matrix 129 9.3 0.016 0.018 30.506 30.951 prepare_preconditioner 11 7.9 0.000 0.000 30.897 30.948 make_preconditioner 11 8.9 0.000 0.001 30.897 30.948 make_full_inverse_cholesky 11 9.9 0.000 0.002 28.557 29.928 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.536 27.941 multiply_cannon_multrec 20216 15.6 13.160 15.999 22.577 25.273 multiply_cannon_metrocomm3 20216 15.6 0.061 0.064 15.514 25.074 qs_ot_get_derivative 118 11.6 0.001 0.002 23.186 23.560 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.827 20.839 apply_single 129 13.6 0.001 0.001 19.827 20.838 ot_diis_step 118 11.6 0.018 0.019 19.576 19.576 qs_ot_get_p 129 10.4 0.002 0.003 19.022 19.469 make_m2s 5054 13.6 0.078 0.083 14.813 16.107 make_images 5054 14.6 1.161 1.264 14.574 15.861 multiply_cannon_sync_h2d 20216 15.6 13.803 15.448 13.803 15.448 qs_ot_p2m_diag 83 11.4 0.265 0.273 14.680 14.688 cp_fm_cholesky_invert 11 10.9 14.447 14.455 14.447 14.455 sum_up_and_integrate 129 10.3 0.002 0.003 14.017 14.040 integrate_v_rspace 129 11.3 0.003 0.004 13.957 13.981 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.665 13.707 calculate_rho_elec 129 8.7 0.130 0.144 13.664 13.706 cp_dbcsr_syevd 83 12.4 0.006 0.009 13.629 13.630 make_images_data 5054 15.6 0.064 0.074 8.881 10.562 cp_fm_diag_elpa 83 13.4 0.000 0.001 10.471 10.494 cp_fm_diag_elpa_base 83 14.4 10.058 10.231 10.465 10.487 init_scf_run 11 5.9 0.000 0.001 10.475 10.475 scf_env_initial_rho_setup 11 6.9 0.001 0.002 10.475 10.475 hybrid_alltoall_any 5240 16.5 0.449 2.048 7.758 10.161 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.066 9.360 multiply_cannon_metrocomm4 17689 15.6 0.066 0.075 3.520 9.235 dbcsr_mm_accdrv_process 41830 16.2 5.629 5.945 8.869 9.014 mp_irecv_dv 50625 16.2 3.390 8.977 3.390 8.977 pw_transfer 1559 11.6 0.085 0.103 8.852 8.950 fft_wrap_pw1pw2 1301 12.7 0.010 0.010 8.628 8.732 grid_integrate_task_list 129 12.3 7.333 7.794 7.333 7.794 fft_wrap_pw1pw2_140 527 13.2 1.069 1.107 7.642 7.752 wfi_extrapolate 11 7.9 0.001 0.001 7.465 7.466 cp_fm_cholesky_decompose 22 10.9 7.313 7.370 7.313 7.370 density_rs2pw 129 9.7 0.005 0.006 6.981 7.290 cp_fm_upper_to_full 105 14.8 5.748 7.230 5.748 7.230 dbcsr_complete_redistribute 395 12.7 1.175 1.209 4.532 6.287 fft3d_ps 1301 14.7 2.847 3.076 5.923 6.016 calculate_dm_sparse 129 9.5 0.001 0.001 5.909 6.015 grid_collocate_task_list 129 9.7 5.197 5.561 5.197 5.561 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.463 5.469 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 4.748 5.440 copy_fm_to_dbcsr 209 11.7 0.001 0.002 3.384 5.131 mp_alltoall_d11v 2423 14.1 4.284 5.130 4.284 5.130 mp_allgather_i34 2527 14.6 1.578 4.949 1.578 4.949 mp_sum_l 8010 12.9 3.249 4.880 3.249 4.880 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.576 4.706 potential_pw2rs 129 12.3 0.021 0.023 4.400 4.414 transfer_fm_to_dbcsr 11 9.9 0.018 0.026 2.320 4.062 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.965 4.002 mp_alltoall_i22 718 14.1 1.929 3.849 1.929 3.849 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.735 3.735 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=176.257000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=912.272727, yerr=8.180808 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bbd14a5fe6b002419d127c8ec63e3062d934e058_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 4.353788E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 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 1.142665E+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 1.770319E+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 665379241840 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58348. MP_Allreduce 11057 1083. MP_Sync 87 MP_Alltoall 1712 12503084. 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 0.014 0.032 205.429 205.430 qs_mol_dyn_low 1 2.0 0.003 0.003 205.030 205.044 qs_forces 11 3.9 0.004 0.005 204.912 204.920 qs_energies 11 4.9 0.002 0.003 197.644 197.653 scf_env_do_scf 11 5.9 0.001 0.001 180.120 180.130 velocity_verlet 10 3.0 0.002 0.002 134.893 134.895 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 129.578 129.581 qs_scf_new_mos 117 7.6 0.001 0.001 91.266 91.580 qs_scf_loop_do_ot 117 8.6 0.001 0.001 91.265 91.580 ot_scf_mini 117 9.6 0.003 0.003 86.641 86.931 dbcsr_multiply_generic 2507 12.6 0.192 0.198 82.236 83.047 multiply_cannon 2507 13.6 0.541 0.565 56.265 58.771 multiply_cannon_loop 2507 14.6 1.834 1.919 52.182 54.166 init_scf_loop 11 6.9 0.000 0.000 50.409 50.410 prepare_preconditioner 11 7.9 0.000 0.000 46.116 46.163 make_preconditioner 11 8.9 0.000 0.000 46.116 46.163 make_full_inverse_cholesky 11 9.9 0.012 0.020 39.684 44.721 ot_mini 117 10.6 0.001 0.001 44.429 44.716 multiply_cannon_multrec 30084 15.6 13.762 18.714 26.686 31.234 rebuild_ks_matrix 128 8.3 0.001 0.001 30.500 30.793 qs_ks_build_kohn_sham_matrix 128 9.3 0.050 0.059 30.500 30.793 mp_waitall_1 147882 16.7 18.249 29.095 18.249 29.095 qs_ot_get_p 128 10.4 0.002 0.002 28.175 28.483 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.520 27.782 qs_ot_get_derivative 117 11.6 0.001 0.002 24.776 25.066 qs_ot_p2m_diag 83 11.4 0.343 0.403 23.356 23.407 make_m2s 5014 13.6 0.092 0.097 21.261 22.314 make_images 5014 14.6 1.939 2.162 20.951 22.003 cp_dbcsr_syevd 83 12.4 0.005 0.006 21.840 21.843 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 19.023 19.570 apply_single 128 13.6 0.001 0.001 19.022 19.570 ot_diis_step 117 11.6 0.018 0.019 19.516 19.518 cp_fm_cholesky_invert 11 10.9 18.514 18.523 18.514 18.523 cp_fm_diag_elpa 83 13.4 0.000 0.000 18.003 18.020 cp_fm_diag_elpa_base 83 14.4 16.939 17.291 17.996 18.011 cp_fm_upper_to_full 105 14.8 11.586 17.082 11.586 17.082 multiply_cannon_metrocomm3 30084 15.6 0.048 0.051 6.108 15.278 sum_up_and_integrate 128 10.3 0.002 0.003 15.012 15.040 integrate_v_rspace 128 11.3 0.003 0.004 14.953 14.982 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.929 14.974 calculate_rho_elec 128 8.7 0.169 0.184 14.928 14.974 dbcsr_complete_redistribute 395 12.7 1.498 1.613 9.480 13.346 make_images_data 5014 15.6 0.064 0.071 11.376 13.273 dbcsr_mm_accdrv_process 62264 16.2 8.319 8.846 12.500 12.998 hybrid_alltoall_any 5200 16.5 0.522 2.173 10.046 12.198 copy_fm_to_dbcsr 209 11.7 0.001 0.002 7.937 11.810 init_scf_run 11 5.9 0.000 0.001 11.452 11.453 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.452 11.453 multiply_cannon_sync_h2d 30084 15.6 10.422 11.298 10.422 11.298 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 10.485 10.702 transfer_fm_to_dbcsr 11 9.9 0.002 0.008 6.409 10.167 pw_transfer 1547 11.6 0.085 0.101 9.964 10.051 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 9.738 9.830 mp_alltoall_i22 716 14.1 5.683 9.610 5.683 9.610 cp_fm_cholesky_decompose 22 10.9 9.010 9.099 9.010 9.099 fft_wrap_pw1pw2_140 523 13.2 1.248 1.289 8.444 8.557 density_rs2pw 128 9.7 0.005 0.006 7.949 8.331 wfi_extrapolate 11 7.9 0.001 0.001 8.171 8.171 grid_integrate_task_list 128 12.3 7.496 7.962 7.496 7.962 fft3d_ps 1291 14.7 2.999 3.052 6.791 6.835 calculate_dm_sparse 128 9.5 0.001 0.001 6.316 6.417 mp_alltoall_d11v 2415 14.1 5.401 6.400 5.401 6.400 multiply_cannon_metrocomm4 25070 15.6 0.080 0.088 2.752 6.358 mp_irecv_dv 76098 16.2 2.604 6.096 2.604 6.096 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.674 5.731 grid_collocate_task_list 128 9.7 5.281 5.616 5.281 5.616 potential_pw2rs 128 12.3 0.022 0.023 5.184 5.233 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.734 4.811 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.483 4.591 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 4.476 4.477 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.378 4.452 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=205.430000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1081.000000, yerr=15.948782 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bbd14a5fe6b002419d127c8ec63e3062d934e058_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963542011904 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019182452736 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019182452736 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796564E+12 0.0% 0.0% 100.0% flops max/rank 5.820057E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499488 0.0% 0.0% 100.0% number of processed stacks 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 1.566867E+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 716108490000 ------------------------------------------------------------------------------- - - - 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 3989 58388. MP_Allreduce 11050 1168. MP_Sync 86 MP_Alltoall 1700 18828174. 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 0.015 0.032 190.922 190.923 qs_mol_dyn_low 1 2.0 0.003 0.003 190.512 190.525 qs_forces 11 3.9 0.004 0.004 190.400 190.404 qs_energies 11 4.9 0.003 0.003 182.686 182.694 scf_env_do_scf 11 5.9 0.001 0.001 164.592 164.604 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 123.569 123.570 velocity_verlet 10 3.0 0.002 0.002 122.985 122.989 qs_scf_new_mos 116 7.6 0.001 0.001 84.964 85.049 qs_scf_loop_do_ot 116 8.6 0.001 0.001 84.963 85.048 ot_scf_mini 116 9.6 0.003 0.004 80.523 80.574 dbcsr_multiply_generic 2485 12.5 0.185 0.189 73.028 73.262 multiply_cannon 2485 13.5 0.553 0.583 53.200 57.155 multiply_cannon_loop 2485 14.5 0.799 0.836 50.086 50.825 init_scf_loop 11 6.9 0.000 0.000 40.857 40.859 ot_mini 116 10.6 0.001 0.001 38.879 38.935 prepare_preconditioner 11 7.9 0.000 0.000 36.900 36.924 make_preconditioner 11 8.9 0.000 0.000 36.900 36.924 make_full_inverse_cholesky 11 9.9 0.018 0.025 34.673 34.994 mp_waitall_1 124680 16.7 25.310 31.155 25.310 31.155 rebuild_ks_matrix 127 8.3 0.001 0.001 30.030 30.095 qs_ks_build_kohn_sham_matrix 127 9.3 0.018 0.022 30.029 30.095 qs_ot_get_p 127 10.4 0.002 0.002 28.933 29.020 qs_ks_update_qs_env 127 7.6 0.001 0.001 27.275 27.329 qs_ot_p2m_diag 82 11.4 0.489 0.495 24.511 24.536 cp_dbcsr_syevd 82 12.4 0.005 0.005 23.156 23.158 qs_ot_get_derivative 116 11.6 0.002 0.002 22.190 22.247 cp_fm_cholesky_invert 11 10.9 21.350 21.358 21.350 21.358 multiply_cannon_multrec 9940 15.5 10.347 14.460 17.901 20.652 cp_fm_diag_elpa 82 13.4 0.000 0.000 19.500 19.509 cp_fm_diag_elpa_base 82 14.4 19.192 19.330 19.494 19.502 multiply_cannon_metrocomm3 9940 15.5 0.024 0.025 11.882 18.421 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 16.803 17.094 apply_single 127 13.6 0.001 0.001 16.803 17.093 ot_diis_step 116 11.6 0.020 0.020 16.607 16.608 make_m2s 4970 13.5 0.063 0.069 15.522 16.302 make_images 4970 14.5 2.141 2.543 15.217 15.997 qs_rho_update_rho_low 127 7.7 0.001 0.001 15.280 15.296 calculate_rho_elec 127 8.7 0.251 0.261 15.279 15.295 sum_up_and_integrate 127 10.3 0.002 0.003 15.061 15.113 integrate_v_rspace 127 11.3 0.004 0.004 15.002 15.055 init_scf_run 11 5.9 0.000 0.001 11.052 11.052 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.052 11.052 multiply_cannon_sync_h2d 9940 15.5 10.749 11.035 10.749 11.035 pw_transfer 1535 11.6 0.085 0.095 10.537 10.593 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 10.316 10.377 make_images_data 4970 15.5 0.053 0.065 8.711 10.032 cp_fm_cholesky_decompose 22 10.9 9.684 9.844 9.684 9.844 hybrid_alltoall_any 5155 16.4 0.832 3.743 8.372 9.362 fft_wrap_pw1pw2_140 519 13.2 1.663 1.696 8.856 8.926 qs_ot_get_derivative_diag 76 12.4 0.002 0.003 8.818 8.856 density_rs2pw 127 9.7 0.005 0.005 7.906 8.235 grid_integrate_task_list 127 12.3 7.685 8.048 7.685 8.048 wfi_extrapolate 11 7.9 0.001 0.001 7.929 7.929 dbcsr_mm_accdrv_process 20590 16.1 3.113 4.450 7.209 7.918 multiply_cannon_metrocomm1 9940 15.5 0.030 0.030 5.001 7.599 fft3d_ps 1281 14.7 3.115 3.225 6.805 6.843 mp_alltoall_d11v 2401 14.1 5.361 6.265 5.361 6.265 dbcsr_complete_redistribute 393 12.7 2.113 2.247 5.770 6.235 calculate_dm_sparse 127 9.5 0.001 0.001 6.107 6.179 grid_collocate_task_list 127 9.7 5.530 5.834 5.530 5.834 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 5.309 5.310 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.250 5.269 potential_pw2rs 127 12.3 0.026 0.027 5.090 5.117 mp_allgather_i34 2485 14.5 1.323 4.531 1.323 4.531 copy_fm_to_dbcsr 208 11.6 0.001 0.002 3.698 4.093 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.027 4.069 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.855 3.937 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.595 3.909 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.784 3.877 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=190.923000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1489.181818, yerr=18.398706 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bbd14a5fe6b002419d127c8ec63e3062d934e058_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456868864 0.0% 0.0% 100.0% flops 32 x 32 x 32 1967095021568 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986252263424 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992007581696 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753954734080 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 11613074706432 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176888320 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176888320 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911117062144 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.237301E+12 0.0% 0.0% 100.0% flops max/rank 11.786605E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806449600 0.0% 0.0% 100.0% number of processed stacks 1981024 0.0% 0.0% 100.0% average stack size 0.0 0.0 3435.8 marketing flops 145.655226E+12 ------------------------------------------------------------------------------- # multiplications 2531 max memory usage/rank 3.165299E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101240 MPI messages size (bytes): total size 1.145036E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.310119E+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 45728 35500589056 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592330608 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4080 58046. MP_Allreduce 11277 1498. MP_Sync 87 MP_Alltoall 1724 36993608. 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 0.030 0.051 297.886 297.888 qs_mol_dyn_low 1 2.0 0.003 0.003 297.085 297.345 qs_forces 11 3.9 0.004 0.005 296.947 296.950 qs_energies 11 4.9 0.002 0.002 288.011 288.014 scf_env_do_scf 11 5.9 0.001 0.001 265.249 265.258 velocity_verlet 10 3.0 0.002 0.002 215.246 215.254 scf_env_do_scf_inner_loop 118 6.6 0.005 0.012 139.423 139.425 init_scf_loop 11 6.9 0.000 0.000 125.548 125.552 prepare_preconditioner 11 7.9 0.000 0.000 120.588 120.606 make_preconditioner 11 8.9 0.000 0.000 120.588 120.606 make_full_inverse_cholesky 11 9.9 0.038 0.039 96.601 117.729 qs_scf_new_mos 118 7.6 0.001 0.001 93.385 93.460 qs_scf_loop_do_ot 118 8.6 0.001 0.001 93.384 93.459 ot_scf_mini 118 9.6 0.004 0.004 88.420 88.449 dbcsr_multiply_generic 2531 12.6 0.218 0.228 83.227 83.557 cp_fm_upper_to_full 105 14.8 52.728 75.702 52.728 75.702 multiply_cannon 2531 13.6 0.677 0.725 59.456 61.336 multiply_cannon_loop 2531 14.6 1.058 1.088 55.291 56.258 ot_mini 118 10.6 0.001 0.001 44.952 44.995 dbcsr_complete_redistribute 395 12.7 4.007 4.056 30.087 43.059 copy_fm_to_dbcsr 209 11.7 0.001 0.002 26.490 39.455 transfer_fm_to_dbcsr 11 9.9 0.030 0.030 23.945 36.831 rebuild_ks_matrix 129 8.3 0.001 0.001 35.588 35.643 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.018 35.588 35.643 mp_alltoall_i22 718 14.1 21.722 34.559 21.722 34.559 cp_fm_cholesky_invert 11 10.9 34.216 34.223 34.216 34.223 qs_ks_update_qs_env 129 7.6 0.001 0.001 33.065 33.127 mp_waitall_1 104660 16.8 28.443 32.758 28.443 32.758 qs_ot_get_p 129 10.4 0.002 0.002 27.976 28.069 qs_ot_get_derivative 118 11.6 0.002 0.002 25.193 25.220 qs_ot_p2m_diag 83 11.4 0.880 0.886 23.648 23.673 cp_dbcsr_syevd 83 12.4 0.006 0.006 21.827 21.828 make_m2s 5062 13.6 0.074 0.076 19.023 20.509 make_images 5062 14.6 3.083 3.264 18.537 20.023 multiply_cannon_metrocomm3 10124 15.6 0.025 0.025 18.958 19.901 ot_diis_step 118 11.6 0.022 0.023 19.718 19.718 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 18.942 19.124 apply_single 129 13.6 0.001 0.001 18.942 19.124 cp_fm_diag_elpa 83 13.4 0.000 0.000 18.495 18.496 cp_fm_diag_elpa_base 83 14.4 14.074 15.680 18.489 18.490 multiply_cannon_multrec 10124 15.6 10.419 12.243 18.273 18.336 qs_rho_update_rho_low 129 7.7 0.001 0.001 17.798 17.811 calculate_rho_elec 129 8.7 0.481 0.482 17.797 17.811 sum_up_and_integrate 129 10.3 0.002 0.002 16.915 17.007 integrate_v_rspace 129 11.3 0.004 0.004 16.852 16.944 multiply_cannon_sync_h2d 10124 15.6 14.412 14.444 14.412 14.444 pw_transfer 1559 11.6 0.096 0.097 13.689 13.699 fft_wrap_pw1pw2 1301 12.7 0.012 0.012 13.450 13.459 hybrid_alltoall_any 5248 16.5 1.322 3.077 10.903 13.358 make_images_data 5062 15.6 0.063 0.068 10.705 12.978 init_scf_run 11 5.9 0.000 0.001 12.359 12.360 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.359 12.360 fft_wrap_pw1pw2_140 527 13.2 3.223 3.250 11.884 11.895 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 9.836 9.852 dbcsr_mm_accdrv_process 20942 16.1 4.256 6.194 7.602 9.496 cp_fm_cholesky_decompose 22 10.9 9.470 9.493 9.470 9.493 wfi_extrapolate 11 7.9 0.001 0.001 9.204 9.205 grid_integrate_task_list 129 12.3 8.619 8.811 8.619 8.811 density_rs2pw 129 9.7 0.005 0.005 8.686 8.732 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 8.242 8.243 fft3d_ps 1301 14.7 3.998 4.015 7.895 7.913 mp_alltoall_d11v 2423 14.1 7.218 7.347 7.218 7.347 calculate_dm_sparse 129 9.5 0.001 0.001 6.831 6.863 grid_collocate_task_list 129 9.7 6.466 6.539 6.466 6.539 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.452 6.483 copy_dbcsr_to_fm 186 11.8 0.004 0.004 6.268 6.331 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=297.888000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2848.545455, yerr=170.083062 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bbd14a5fe6b002419d127c8ec63e3062d934e058_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 1.260585E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255646. MP_Allreduce 3139 6114. MP_Sync 4 MP_Alltoall 54 MP_SendRecv 285 19200. MP_ISendRecv 285 19200. MP_Wait 1017 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.050 85.053 85.073 qs_energies 1 2.0 0.000 0.000 84.554 84.567 ls_scf 1 3.0 0.000 0.000 83.659 83.672 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.519 72.690 multiply_cannon 111 7.7 0.017 0.020 55.814 57.266 multiply_cannon_loop 111 8.7 0.228 0.241 52.401 53.843 ls_scf_main 1 4.0 0.000 0.000 52.452 52.453 density_matrix_trs4 2 5.0 0.002 0.003 46.924 46.991 ls_scf_init_scf 1 4.0 0.000 0.000 28.193 28.194 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.060 27.119 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 24.925 24.942 mp_waitall_1 11031 10.9 22.404 24.376 22.404 24.376 multiply_cannon_multrec 2664 9.7 8.140 8.908 15.633 17.476 multiply_cannon_sync_h2d 2664 9.7 13.493 15.324 13.493 15.324 make_m2s 222 7.7 0.008 0.011 13.119 13.592 make_images 222 8.7 0.098 0.109 13.097 13.573 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.621 11.909 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.447 8.272 make_images_data 222 9.7 0.004 0.005 7.667 8.249 hybrid_alltoall_any 227 10.6 0.216 1.844 6.597 8.160 dbcsr_mm_accdrv_process 4760 10.4 0.585 0.689 7.113 8.149 dbcsr_mm_accdrv_process_sort 4760 11.4 6.329 7.297 6.329 7.297 calculate_norms 4752 9.8 5.506 6.135 5.506 6.135 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.011 5.142 mp_sum_l 887 5.1 3.047 4.266 3.047 4.266 make_images_sizes 222 9.7 0.000 0.000 0.770 3.573 mp_alltoall_i44 222 10.7 0.770 3.573 0.770 3.573 multiply_cannon_metrocomm4 2442 9.7 0.012 0.014 2.037 3.374 mp_irecv_dv 6231 10.9 2.020 3.348 2.020 3.348 arnoldi_extremal 4 6.8 0.000 0.000 3.321 3.342 arnoldi_normal_ev 4 7.8 0.001 0.003 3.321 3.342 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.280 3.309 build_subspace 16 8.4 0.009 0.012 3.230 3.232 ls_scf_post 1 4.0 0.000 0.000 3.014 3.027 ls_scf_store_result 1 5.0 0.000 0.000 2.844 2.876 dbcsr_special_finalize 555 9.7 0.005 0.006 2.329 2.806 dbcsr_merge_single_wm 555 10.7 0.463 0.568 2.321 2.798 dbcsr_matrix_vector_mult 304 9.0 0.006 0.014 2.379 2.648 make_images_pack 222 9.7 2.206 2.639 2.208 2.641 dbcsr_matrix_vector_mult_local 304 10.0 2.076 2.525 2.079 2.527 dbcsr_sort_data 658 11.4 2.116 2.518 2.116 2.518 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.308 2.401 buffer_matrices_ensure_size 222 8.7 1.756 2.019 1.756 2.019 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.769 1.770 rebuild_ks_matrix 3 7.3 0.000 0.000 1.760 1.760 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.760 1.760 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.073000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1141.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bbd14a5fe6b002419d127c8ec63e3062d934e058_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 2.139451E+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 0.022 0.039 90.571 90.571 qs_energies 1 2.0 0.000 0.000 90.135 90.138 ls_scf 1 3.0 0.000 0.000 88.797 88.801 dbcsr_multiply_generic 111 6.7 0.015 0.016 74.854 75.147 multiply_cannon 111 7.7 0.026 0.038 52.868 57.856 ls_scf_main 1 4.0 0.000 0.000 55.062 55.067 multiply_cannon_loop 111 8.7 0.135 0.146 49.992 53.480 density_matrix_trs4 2 5.0 0.002 0.003 49.315 49.587 ls_scf_init_scf 1 4.0 0.000 0.000 29.993 29.995 mp_waitall_1 9105 10.9 20.897 29.799 20.897 29.799 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.769 28.877 multiply_cannon_multrec 1332 9.7 13.316 17.038 22.710 27.575 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.394 26.407 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.469 20.245 make_m2s 222 7.7 0.006 0.007 15.310 15.941 make_images 222 8.7 1.370 1.696 15.280 15.911 dbcsr_mm_accdrv_process 4041 10.4 0.333 0.503 8.989 10.580 dbcsr_mm_accdrv_process_sort 4041 11.4 8.509 10.051 8.509 10.051 make_images_data 222 9.7 0.004 0.005 8.918 9.792 hybrid_alltoall_any 227 10.6 0.544 2.551 8.226 9.713 mp_sum_l 887 5.1 5.208 8.445 5.208 8.445 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.213 7.738 mp_irecv_dv 3311 11.0 3.193 7.685 3.193 7.685 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.948 6.723 calculate_norms 2376 9.8 6.044 6.621 6.044 6.621 multiply_cannon_sync_h2d 1332 9.7 4.899 6.214 4.899 6.214 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.985 5.171 arnoldi_extremal 4 6.8 0.000 0.000 4.683 4.701 arnoldi_normal_ev 4 7.8 0.001 0.005 4.683 4.701 build_subspace 16 8.4 0.014 0.021 4.428 4.431 ls_scf_post 1 4.0 0.000 0.000 3.742 3.746 ls_scf_store_result 1 5.0 0.000 0.000 3.427 3.562 dbcsr_matrix_vector_mult 304 9.0 0.010 0.023 3.153 3.388 dbcsr_matrix_vector_mult_local 304 10.0 2.740 3.225 2.742 3.227 mp_allgather_i34 111 8.7 0.863 3.161 0.863 3.161 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.609 2.689 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.156 2.591 dbcsr_data_new 4174 10.1 2.116 2.404 2.116 2.404 make_images_pack 222 9.7 1.821 2.131 1.823 2.133 dbcsr_sort_data 436 11.2 1.842 2.087 1.842 2.087 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.875 1.877 rebuild_ks_matrix 3 7.3 0.000 0.000 1.862 1.864 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.013 1.862 1.864 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=90.571000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1758.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bbd14a5fe6b002419d127c8ec63e3062d934e058_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 2.920010E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265448. MP_Allreduce 3138 10896. MP_Sync 4 MP_Alltoall 47 23526250. MP_SendRecv 93 57600. MP_ISendRecv 93 57600. MP_Wait 639 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.024 0.038 93.244 93.245 qs_energies 1 2.0 0.000 0.000 92.715 92.719 ls_scf 1 3.0 0.000 0.000 91.286 91.291 dbcsr_multiply_generic 111 6.7 0.015 0.016 75.965 76.230 ls_scf_main 1 4.0 0.000 0.000 56.947 56.951 multiply_cannon 111 7.7 0.030 0.069 52.550 56.680 multiply_cannon_loop 111 8.7 0.116 0.129 49.771 53.873 density_matrix_trs4 2 5.0 0.002 0.003 51.024 51.239 mp_waitall_1 7281 11.0 24.041 34.342 24.041 34.342 ls_scf_init_scf 1 4.0 0.000 0.000 30.667 30.671 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.470 29.547 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.099 27.115 multiply_cannon_multrec 888 9.7 12.687 15.483 21.346 24.469 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.408 24.024 make_m2s 222 7.7 0.006 0.007 16.515 17.274 make_images 222 8.7 1.584 1.846 16.477 17.234 make_images_data 222 9.7 0.004 0.004 9.710 10.730 hybrid_alltoall_any 227 10.6 0.640 2.922 9.198 10.397 dbcsr_mm_accdrv_process 3754 10.4 0.309 0.486 8.190 9.417 mp_sum_l 887 5.1 5.212 9.310 5.212 9.310 dbcsr_mm_accdrv_process_sort 3754 11.4 7.757 8.932 7.757 8.932 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.461 7.600 mp_irecv_dv 2335 11.1 2.444 7.546 2.444 7.546 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.045 7.442 multiply_cannon_sync_h2d 888 9.7 6.181 7.315 6.181 7.315 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.557 6.763 arnoldi_extremal 4 6.8 0.000 0.000 5.135 5.156 arnoldi_normal_ev 4 7.8 0.001 0.005 5.135 5.155 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.767 5.004 build_subspace 16 8.4 0.014 0.020 4.823 4.830 calculate_norms 1584 9.8 4.350 4.771 4.350 4.771 mp_allgather_i34 111 8.7 0.901 3.850 0.901 3.850 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.445 3.771 ls_scf_post 1 4.0 0.000 0.000 3.672 3.677 dbcsr_matrix_vector_mult_local 304 10.0 3.019 3.585 3.021 3.586 ls_scf_store_result 1 5.0 0.000 0.000 3.421 3.500 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.854 2.954 dbcsr_data_new 4116 9.9 2.108 2.463 2.108 2.463 dbcsr_sort_data 325 11.1 1.897 2.149 1.897 2.149 make_images_sizes 222 9.7 0.000 0.000 0.938 2.077 mp_alltoall_i44 222 10.7 0.937 2.077 0.937 2.077 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.903 1.905 dbcsr_finalize 304 7.8 0.026 0.032 1.599 1.898 rebuild_ks_matrix 3 7.3 0.000 0.000 1.886 1.887 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.885 1.887 make_images_pack 222 9.7 1.625 1.876 1.628 1.879 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=93.245000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2287.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bbd14a5fe6b002419d127c8ec63e3062d934e058_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 3.332395E+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 0.056 0.120 98.408 98.413 qs_energies 1 2.0 0.002 0.015 97.619 97.699 ls_scf 1 3.0 0.002 0.013 95.933 95.999 dbcsr_multiply_generic 111 6.7 0.016 0.017 79.239 79.562 ls_scf_main 1 4.0 0.004 0.041 59.549 59.551 multiply_cannon 111 7.7 0.052 0.178 51.907 56.353 density_matrix_trs4 2 5.0 0.003 0.017 53.363 53.485 multiply_cannon_loop 111 8.7 0.154 0.169 46.880 50.497 ls_scf_init_scf 1 4.0 0.000 0.003 33.070 33.072 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.545 31.632 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.061 29.074 mp_waitall_1 6369 11.0 22.950 28.541 22.950 28.541 multiply_cannon_multrec 1332 9.7 14.117 17.257 22.134 24.531 make_m2s 222 7.7 0.007 0.008 21.246 22.680 make_images 222 8.7 3.144 3.600 21.196 22.632 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.238 17.601 make_images_data 222 9.7 0.004 0.004 11.840 13.536 hybrid_alltoall_any 227 10.6 0.797 3.795 11.216 12.848 dbcsr_mm_accdrv_process 3641 10.4 0.291 0.500 7.650 9.179 mp_sum_l 887 5.1 4.564 8.734 4.564 8.734 dbcsr_mm_accdrv_process_sort 3641 11.4 7.200 8.679 7.200 8.679 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.497 6.933 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.099 6.051 mp_irecv_dv 3229 10.9 2.076 5.970 2.076 5.970 multiply_cannon_sync_h2d 1332 9.7 5.435 5.840 5.435 5.840 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.702 5.324 arnoldi_extremal 4 6.8 0.000 0.000 5.260 5.275 arnoldi_normal_ev 4 7.8 0.001 0.005 5.260 5.274 build_subspace 16 8.4 0.014 0.021 4.926 4.933 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.586 4.764 calculate_norms 2376 9.8 4.162 4.503 4.162 4.503 mp_allgather_i34 111 8.7 2.143 4.449 2.143 4.449 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.610 3.900 dbcsr_matrix_vector_mult_local 304 10.0 3.192 3.685 3.194 3.687 dbcsr_sort_data 658 11.4 3.068 3.427 3.068 3.427 ls_scf_post 1 4.0 0.005 0.041 3.313 3.369 dbcsr_special_finalize 555 9.7 0.006 0.007 2.823 3.240 dbcsr_merge_single_wm 555 10.7 0.538 0.660 2.815 3.232 ls_scf_store_result 1 5.0 0.000 0.000 3.005 3.100 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.034 3.076 dbcsr_data_release 10477 10.7 1.587 2.379 1.587 2.379 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.162 2.165 rebuild_ks_matrix 3 7.3 0.000 0.000 2.139 2.142 qs_ks_build_kohn_sham_matrix 3 8.3 0.006 0.040 2.139 2.141 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=98.413000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2709.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bbd14a5fe6b002419d127c8ec63e3062d934e058_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 4.760846E+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 0.056 0.081 91.966 91.967 qs_energies 1 2.0 0.000 0.000 91.164 91.165 ls_scf 1 3.0 0.000 0.000 89.186 89.189 dbcsr_multiply_generic 111 6.7 0.017 0.018 70.548 70.779 ls_scf_main 1 4.0 0.000 0.000 56.324 56.337 multiply_cannon 111 7.7 0.084 0.146 52.507 55.541 multiply_cannon_loop 111 8.7 0.088 0.093 49.973 51.133 density_matrix_trs4 2 5.0 0.002 0.003 49.375 49.450 ls_scf_init_scf 1 4.0 0.000 0.000 29.445 29.449 mp_waitall_1 5436 11.0 24.155 28.550 24.155 28.550 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.154 28.191 matrix_sqrt_Newton_Schulz 2 6.5 0.008 0.013 26.017 26.030 multiply_cannon_multrec 444 9.7 13.730 16.368 20.894 22.060 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.800 15.454 make_m2s 222 7.7 0.005 0.005 13.368 14.280 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.035 14.240 make_images 222 8.7 2.039 2.470 13.301 14.211 make_images_data 222 9.7 0.003 0.004 8.136 9.565 hybrid_alltoall_any 227 10.6 0.801 3.813 8.092 9.333 multiply_cannon_sync_h2d 444 9.7 6.780 8.174 6.780 8.174 dbcsr_mm_accdrv_process 3003 10.4 0.350 0.400 6.859 7.995 dbcsr_mm_accdrv_process_sort 3003 11.4 6.498 7.598 6.498 7.598 arnoldi_extremal 4 6.8 0.000 0.000 5.883 5.903 arnoldi_normal_ev 4 7.8 0.004 0.012 5.883 5.903 build_subspace 16 8.4 0.015 0.020 5.481 5.490 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.388 4.641 dbcsr_matrix_vector_mult 304 9.0 0.011 0.022 4.201 4.404 mp_sum_l 887 5.1 2.675 4.403 2.675 4.403 dbcsr_matrix_vector_mult_local 304 10.0 3.706 4.186 3.708 4.188 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.509 3.734 mp_irecv_dv 1241 11.2 1.496 3.702 1.496 3.702 calculate_norms 792 9.8 3.612 3.690 3.612 3.690 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.531 3.590 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.872 3.563 mp_allgather_i34 111 8.7 1.116 3.510 1.116 3.510 ls_scf_post 1 4.0 0.000 0.000 3.417 3.419 make_images_sizes 222 9.7 0.000 0.000 0.841 3.234 mp_alltoall_i44 222 10.7 0.841 3.234 0.841 3.234 ls_scf_store_result 1 5.0 0.000 0.000 3.184 3.218 dbcsr_finalize 304 7.8 0.062 0.076 2.202 2.293 dbcsr_data_new 4608 9.7 1.789 2.255 1.789 2.255 dbcsr_merge_all 275 8.9 0.479 0.530 2.060 2.138 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.077 2.079 rebuild_ks_matrix 3 7.3 0.000 0.000 2.044 2.047 qs_ks_build_kohn_sham_matrix 3 8.3 0.014 0.016 2.044 2.047 qs_energies_init_hamiltonians 1 3.0 0.001 0.002 1.961 1.961 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=91.967000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3716.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bbd14a5fe6b002419d127c8ec63e3062d934e058_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 8.879464E+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 0.071 0.088 106.120 106.121 qs_energies 1 2.0 0.000 0.000 104.777 104.782 ls_scf 1 3.0 0.000 0.000 101.767 101.772 dbcsr_multiply_generic 111 6.7 0.024 0.026 74.485 74.653 ls_scf_main 1 4.0 0.000 0.000 63.919 63.919 density_matrix_trs4 2 5.0 0.002 0.003 54.724 54.786 multiply_cannon 111 7.7 0.133 0.221 47.946 50.177 multiply_cannon_loop 111 8.7 0.097 0.099 45.008 45.276 ls_scf_init_scf 1 4.0 0.000 0.000 34.057 34.058 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.429 32.447 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 29.640 29.647 mp_waitall_1 4527 11.1 21.576 25.258 21.576 25.258 make_m2s 222 7.7 0.005 0.005 22.903 23.934 make_images 222 8.7 3.579 3.886 22.795 23.824 multiply_cannon_multrec 444 9.7 17.800 18.397 22.501 23.071 hybrid_alltoall_any 227 10.6 1.655 3.624 12.902 15.856 make_images_data 222 9.7 0.003 0.004 13.197 15.374 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 9.783 10.258 multiply_cannon_sync_h2d 444 9.7 8.790 8.850 8.790 8.850 arnoldi_extremal 4 6.8 0.000 0.000 7.796 7.806 arnoldi_normal_ev 4 7.8 0.003 0.009 7.796 7.806 build_subspace 16 8.4 0.026 0.036 7.243 7.255 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.697 5.797 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.312 5.405 dbcsr_matrix_vector_mult_local 304 10.0 5.084 5.397 5.086 5.399 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.847 5.112 dbcsr_mm_accdrv_process 1814 10.4 0.267 0.360 4.518 4.640 dbcsr_mm_accdrv_process_sort 1814 11.4 4.182 4.311 4.182 4.311 make_images_sizes 222 9.7 0.000 0.000 1.558 3.897 mp_alltoall_i44 222 10.7 1.558 3.897 1.558 3.897 ls_scf_post 1 4.0 0.000 0.000 3.791 3.797 ls_scf_store_result 1 5.0 0.000 0.000 3.515 3.551 calculate_norms 792 9.8 3.237 3.324 3.237 3.324 mp_allgather_i34 111 8.7 1.050 3.220 1.050 3.220 dbcsr_finalize 304 7.8 0.082 0.089 3.084 3.181 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.979 2.979 dbcsr_merge_all 275 8.9 0.885 0.919 2.867 2.959 dbcsr_complete_redistribute 5 7.6 1.462 1.506 2.839 2.950 matrix_ls_to_qs 2 6.0 0.000 0.000 2.457 2.570 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.559 2.561 dbcsr_sort_data 325 11.1 2.443 2.514 2.443 2.514 rebuild_ks_matrix 3 7.3 0.000 0.000 2.492 2.494 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.002 2.492 2.494 dbcsr_data_new 6591 9.6 1.884 2.420 1.884 2.420 dbcsr_new_transposed 4 7.5 0.256 0.282 2.389 2.399 dbcsr_frobenius_norm 74 6.6 2.055 2.133 2.192 2.227 dbcsr_add_d 103 6.2 0.000 0.000 2.135 2.220 dbcsr_add_anytype 103 7.2 0.859 0.893 2.135 2.220 mp_alltoall_d11v 48 9.2 2.145 2.186 2.145 2.186 dbcsr_data_release 12724 10.6 1.979 2.171 1.979 2.171 dbcsr_redistribute 4 8.5 1.367 1.416 2.106 2.127 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=106.121000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=7079.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bbd14a5fe6b002419d127c8ec63e3062d934e058_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 594.137088E+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 1232397663. 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 0.068 0.197 285.262 285.271 qs_mol_dyn_low 1 2.0 0.005 0.028 283.870 283.958 qs_forces 5 3.8 0.006 0.045 283.734 283.742 qs_energies 5 4.8 0.007 0.055 280.344 280.379 scf_env_do_scf 5 5.8 0.001 0.013 264.342 264.345 scf_env_do_scf_inner_loop 105 6.6 0.003 0.015 230.450 230.461 qs_scf_new_mos 105 7.6 0.000 0.001 181.089 181.335 qs_scf_loop_do_ot 105 8.6 0.001 0.001 181.088 181.334 ot_scf_mini 105 9.6 0.003 0.004 170.965 171.172 dbcsr_multiply_generic 1445 12.2 0.134 0.142 137.540 138.098 multiply_cannon 1445 13.2 0.275 0.285 116.309 118.338 velocity_verlet 4 3.0 0.003 0.021 118.179 118.184 multiply_cannon_loop 1445 14.2 2.843 2.984 113.852 115.548 qs_ot_get_p 112 10.4 0.001 0.002 77.723 77.987 qs_ot_p2m_diag 40 11.0 0.020 0.031 65.223 65.322 ot_mini 105 10.6 0.003 0.032 62.504 62.629 cp_dbcsr_syevd 40 12.0 0.002 0.012 61.687 61.689 cp_fm_syevd 40 13.0 0.001 0.023 55.604 55.762 cp_fm_redistribute_end 40 14.0 25.017 49.903 25.028 49.909 cp_fm_syevd_base 40 14.0 24.864 49.750 24.864 49.750 mp_waitall_1 488190 16.1 39.124 47.027 39.124 47.027 multiply_cannon_multrec 69360 15.2 29.442 34.429 39.639 45.262 qs_ot_get_derivative 55 11.6 0.001 0.001 40.665 40.794 multiply_cannon_metrocomm3 69360 15.2 0.199 0.207 26.669 36.863 init_scf_loop 7 6.6 0.000 0.004 33.836 33.840 multiply_cannon_sync_h2d 69360 15.2 28.753 33.042 28.753 33.042 rebuild_ks_matrix 110 8.4 0.000 0.000 32.755 32.958 qs_ks_build_kohn_sham_matrix 110 9.4 0.013 0.031 32.754 32.957 qs_ks_update_qs_env 112 7.6 0.001 0.001 30.052 30.237 prepare_preconditioner 7 7.6 0.000 0.000 28.263 28.295 make_preconditioner 7 8.6 0.000 0.002 28.263 28.295 qs_rho_update_rho_low 110 7.6 0.000 0.001 23.410 23.835 calculate_rho_elec 110 8.6 0.029 0.032 23.410 23.835 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 23.169 23.425 apply_single 62 13.6 0.000 0.000 23.168 23.424 ot_new_cg_direction 55 11.6 0.001 0.003 21.161 21.163 make_full_inverse_cholesky 7 9.6 0.000 0.001 19.775 19.846 density_rs2pw 110 9.6 0.005 0.006 17.168 17.733 qs_ot_get_orbitals 105 10.6 0.001 0.001 15.407 15.621 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 14.453 14.578 transfer_rs2pw 445 10.6 0.007 0.008 13.392 14.026 mp_sum_l 4764 12.2 12.774 13.753 12.774 13.753 cp_fm_cholesky_invert 7 10.6 13.491 13.508 13.491 13.508 init_scf_run 5 5.8 0.000 0.001 13.149 13.150 scf_env_initial_rho_setup 5 6.8 0.002 0.003 13.148 13.150 pw_transfer 1645 12.4 0.080 0.104 12.766 13.036 fft_wrap_pw1pw2 1425 13.5 0.012 0.015 12.626 12.895 calculate_dm_sparse 110 9.5 0.000 0.001 11.980 12.194 sum_up_and_integrate 60 10.3 0.001 0.003 11.774 11.791 integrate_v_rspace 60 11.3 0.001 0.002 11.757 11.775 check_diag 80 13.5 8.593 8.888 10.985 11.148 fft_wrap_pw1pw2_240 915 15.0 0.847 0.935 10.859 11.117 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 11.001 11.058 qs_vxc_create 110 10.4 0.003 0.035 10.831 10.889 dbcsr_mm_accdrv_process 154766 15.8 6.272 6.436 10.066 10.839 fft3d_pb 915 16.0 2.405 2.752 9.011 9.367 multiply_cannon_metrocomm1 69360 15.2 0.095 0.102 4.974 8.749 make_m2s 2890 13.2 0.077 0.085 8.108 8.737 make_images 2890 14.2 0.237 0.258 8.001 8.626 acc_transpose_blocks 69360 15.2 0.360 0.382 7.857 8.331 make_full_single_inverse 7 9.6 0.001 0.003 8.081 8.113 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.001 7.849 7.865 xc_rho_set_and_dset_create 110 12.4 0.077 0.124 7.549 7.804 potential_pw2rs 60 12.3 0.002 0.003 7.711 7.773 calculate_first_density_matrix 1 7.0 0.001 0.010 7.750 7.761 transfer_rs2pw_30 110 11.6 1.256 1.330 7.151 7.742 mp_sendrecv_dv 168740 12.6 7.408 7.584 7.408 7.584 xc_vxc_pw_create 60 11.3 0.039 0.051 7.292 7.351 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 7.259 7.346 mp_alltoall_z22v 2340 17.7 6.075 6.639 6.075 6.639 xc_pw_derive 510 13.4 0.006 0.007 6.538 6.609 transfer_pw2rs 245 13.2 0.003 0.004 6.452 6.504 cp_fm_cholesky_decompose 14 10.2 5.743 5.752 5.743 5.752 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=285.271000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=563.600000, yerr=3.322650 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: bbd14a5fe6b002419d127c8ec63e3062d934e058 Summary: empty Status: OK