=== 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: 6ad6eaa91393e2e3e5bb87bfd4fb06be6378d79a ################# 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 2022.11.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.2.2, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.2, # 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: 21.06.2023 # # \ 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=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.2 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta.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/6ad6eaa91393e2e3e5bb87bfd4fb06be6378d79a_performance_tests/01 job id: 47431042 --- 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/6ad6eaa91393e2e3e5bb87bfd4fb06be6378d79a_performance_tests/02 job id: 47431043 --- 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/6ad6eaa91393e2e3e5bb87bfd4fb06be6378d79a_performance_tests/03 job id: 47431044 --- 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/6ad6eaa91393e2e3e5bb87bfd4fb06be6378d79a_performance_tests/04 job id: 47431045 --- 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/6ad6eaa91393e2e3e5bb87bfd4fb06be6378d79a_performance_tests/05 job id: 47431046 --- 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/6ad6eaa91393e2e3e5bb87bfd4fb06be6378d79a_performance_tests/06 job id: 47431047 --- 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/6ad6eaa91393e2e3e5bb87bfd4fb06be6378d79a_performance_tests/07 job id: 47431048 --- 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/6ad6eaa91393e2e3e5bb87bfd4fb06be6378d79a_performance_tests/08 job id: 47431049 --- 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/6ad6eaa91393e2e3e5bb87bfd4fb06be6378d79a_performance_tests/09 job id: 47431050 --- 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/6ad6eaa91393e2e3e5bb87bfd4fb06be6378d79a_performance_tests/10 job id: 47431051 --- 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/6ad6eaa91393e2e3e5bb87bfd4fb06be6378d79a_performance_tests/11 job id: 47431052 --- 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/6ad6eaa91393e2e3e5bb87bfd4fb06be6378d79a_performance_tests/12 job id: 47431053 --- 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/6ad6eaa91393e2e3e5bb87bfd4fb06be6378d79a_performance_tests/13 job id: 47431054 --- 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/6ad6eaa91393e2e3e5bb87bfd4fb06be6378d79a_performance_tests/14 job id: 47431055 --- 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/6ad6eaa91393e2e3e5bb87bfd4fb06be6378d79a_performance_tests/15 job id: 47431056 --- 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/6ad6eaa91393e2e3e5bb87bfd4fb06be6378d79a_performance_tests/16 job id: 47431057 --- 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/6ad6eaa91393e2e3e5bb87bfd4fb06be6378d79a_performance_tests/17 job id: 47431058 --- 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/6ad6eaa91393e2e3e5bb87bfd4fb06be6378d79a_performance_tests/18 job id: 47431059 --- 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/6ad6eaa91393e2e3e5bb87bfd4fb06be6378d79a_performance_tests/19 job id: 47431060 --- 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/6ad6eaa91393e2e3e5bb87bfd4fb06be6378d79a_performance_tests/20 job id: 47431061 --- 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/6ad6eaa91393e2e3e5bb87bfd4fb06be6378d79a_performance_tests/21 job id: 47431063 --- 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/6ad6eaa91393e2e3e5bb87bfd4fb06be6378d79a_performance_tests/22 job id: 47431065 --- 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/6ad6eaa91393e2e3e5bb87bfd4fb06be6378d79a_performance_tests/23 job id: 47431066 --- 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/6ad6eaa91393e2e3e5bb87bfd4fb06be6378d79a_performance_tests/24 job id: 47431067 --- 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/6ad6eaa91393e2e3e5bb87bfd4fb06be6378d79a_performance_tests/25 job id: 47431068 --- 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/6ad6eaa91393e2e3e5bb87bfd4fb06be6378d79a_performance_tests/26 job id: 47431069 --- 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/6ad6eaa91393e2e3e5bb87bfd4fb06be6378d79a_performance_tests/27 job id: 47431070 --- 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/6ad6eaa91393e2e3e5bb87bfd4fb06be6378d79a_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 0.024 0.035 133.595 133.596 farming_run 1 2.0 132.999 133.000 133.563 133.566 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.464361E+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 1103589. MP_Allreduce 571 1938539. MP_Sync 25 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 10 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.006 0.027 114.627 114.627 qs_energies 1 2.0 0.000 0.000 114.425 114.426 mp2_main 1 3.0 0.000 0.001 112.348 112.350 mp2_gpw_main 1 4.0 0.019 0.026 111.346 111.347 mp2_ri_gpw_compute_in 1 5.0 0.171 0.172 92.466 92.963 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 54.845 55.343 mp2_eri_3c_integrate_gpw 272 7.0 0.152 0.166 41.293 46.606 get_2c_integrals 1 6.0 0.008 0.009 36.699 37.450 integrate_v_rspace 273 8.0 0.433 0.449 24.693 29.778 pw_transfer 6555 10.6 0.373 0.390 27.367 27.748 fft_wrap_pw1pw2 5465 11.4 0.045 0.047 26.018 26.429 grid_integrate_task_list 273 9.0 20.515 26.082 20.515 26.082 fft_wrap_pw1pw2_100 2178 12.4 1.182 1.331 23.542 23.957 compute_2c_integrals 1 7.0 0.002 0.002 19.371 19.372 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.857 19.096 mp2_eri_2c_integrate_gpw 1 9.0 2.395 2.440 18.855 19.093 rpa_ri_compute_en 1 5.0 0.019 0.020 18.772 18.981 cp_fm_cholesky_decompose 12 8.2 17.258 18.040 17.258 18.040 cholesky_decomp 1 7.0 0.000 0.000 16.171 16.947 fft3d_s 5443 13.4 16.163 16.571 16.185 16.593 ao_to_mo_and_store_B_mult_1 272 7.0 10.720 15.307 10.720 15.307 calculate_wavefunction 272 8.0 5.405 5.560 12.519 13.170 rpa_num_int 1 6.0 0.000 0.000 10.598 10.607 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.582 10.586 calc_mat_Q 8 8.0 0.000 0.000 9.451 9.538 contract_S_to_Q 8 9.0 0.000 0.000 8.874 8.960 calc_potential_gpw 544 9.5 0.005 0.005 8.220 8.616 parallel_gemm_fm 14 9.1 0.000 0.000 8.454 8.550 parallel_gemm_fm_cosma 14 10.1 8.454 8.550 8.454 8.550 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.195 8.438 potential_pw2rs 545 10.0 0.108 0.109 7.700 8.362 collocate_single_gaussian 272 10.0 0.039 0.042 7.460 7.690 create_integ_mat 1 6.0 0.014 0.026 7.583 7.592 array2fm 1 7.0 0.000 0.000 6.576 6.991 pw_scatter_s 2720 13.7 4.417 4.559 4.417 4.559 pw_gather_s 2722 13.2 3.848 4.208 3.848 4.208 array2fm_buffer_send 1 8.0 2.924 3.110 2.924 3.110 pw_poisson_solve 545 10.5 1.103 1.150 2.160 2.309 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.344357, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2797.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6ad6eaa91393e2e3e5bb87bfd4fb06be6378d79a_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 0.029 0.039 402.787 402.789 farming_run 1 2.0 402.019 402.025 402.741 402.743 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.229115E+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 743 386399. MP_Allreduce 2021 21391. MP_Sync 37 MP_Alltoall 77 27163929. MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.032 210.898 210.899 qs_energies 1 2.0 0.000 0.000 210.652 210.660 scf_env_do_scf 1 3.0 0.000 0.000 107.181 107.182 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.988 105.996 rebuild_ks_matrix 4 6.0 0.000 0.000 105.986 105.994 qs_ks_build_kohn_sham_matrix 4 7.0 0.062 0.070 105.986 105.994 hfx_ks_matrix 4 8.0 0.001 0.001 105.511 105.516 integrate_four_center 4 9.0 0.144 0.450 105.511 105.515 mp2_main 1 3.0 0.000 0.000 103.117 103.125 mp2_gpw_main 1 4.0 0.033 0.047 101.562 101.573 integrate_four_center_main 4 10.0 0.116 0.597 96.553 99.634 integrate_four_center_bin 263 11.0 96.437 99.591 96.437 99.591 init_scf_loop 1 4.0 0.000 0.000 92.805 92.805 mp2_ri_gpw_compute_in 1 5.0 0.064 0.064 74.731 75.754 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.098 55.121 mp2_eri_3c_integrate_gpw 91 7.0 0.142 0.158 41.711 46.853 integrate_v_rspace 95 8.0 0.394 0.563 28.171 33.152 pw_transfer 2240 10.6 0.144 0.165 30.052 30.421 fft_wrap_pw1pw2 1868 11.4 0.018 0.021 29.040 29.510 mp2_ri_gpw_compute_en 1 5.0 0.055 0.065 26.677 28.792 grid_integrate_task_list 95 9.0 23.366 28.558 23.366 28.558 fft_wrap_pw1pw2_100 730 12.4 1.305 1.434 26.703 27.273 ao_to_mo_and_store_B_mult_1 91 7.0 10.699 25.411 10.699 25.411 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.845 1.910 25.005 25.014 get_2c_integrals 1 6.0 0.000 0.000 20.426 20.566 compute_2c_integrals 1 7.0 0.002 0.003 19.379 19.398 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 18.975 19.233 mp2_eri_2c_integrate_gpw 1 9.0 1.730 1.857 18.974 19.233 fft3d_s 1823 13.4 18.461 18.931 18.475 18.944 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.373 14.373 calculate_wavefunction 91 8.0 2.012 2.037 9.737 9.964 mp2_ri_gpw_compute_en_expansio 172 7.0 0.557 0.600 8.781 9.393 potential_pw2rs 186 10.0 0.034 0.035 8.756 9.253 local_gemm 172 8.0 8.224 8.828 8.224 8.828 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.307 8.587 mp2_ri_gpw_compute_en_comm 22 7.0 0.498 0.528 7.979 8.338 calc_potential_gpw 182 9.5 0.002 0.002 7.957 8.205 collocate_single_gaussian 91 10.0 0.017 0.023 7.940 8.169 mp2_ri_gpw_compute_en_ener 172 7.0 6.355 6.441 6.355 6.441 mp_sendrecv_dm3 2068 8.0 6.013 6.397 6.013 6.397 mp_sync 37 10.5 3.709 5.723 3.709 5.723 pw_gather_s 912 13.2 4.906 5.462 4.906 5.462 pw_scatter_s 910 13.7 3.928 4.280 3.928 4.280 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.563517, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1516.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6ad6eaa91393e2e3e5bb87bfd4fb06be6378d79a_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 453.021696E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62385. MP_Allreduce 10329 270. MP_Sync 530 MP_Alltoall 2083 789670. 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 0.028 0.070 62.211 62.213 qs_mol_dyn_low 1 2.0 0.006 0.024 61.851 61.859 qs_forces 11 3.9 0.005 0.025 61.743 61.746 qs_energies 11 4.9 0.003 0.022 60.054 60.068 scf_env_do_scf 11 5.9 0.002 0.016 53.086 53.086 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 50.322 50.323 qs_scf_new_mos 108 7.5 0.000 0.001 37.620 37.895 qs_scf_loop_do_ot 108 8.5 0.001 0.001 37.619 37.895 dbcsr_multiply_generic 2286 12.5 0.095 0.099 35.862 36.381 ot_scf_mini 108 9.5 0.002 0.003 35.853 36.016 velocity_verlet 10 3.0 0.002 0.008 30.050 30.053 multiply_cannon 2286 13.5 0.190 0.197 26.976 28.715 multiply_cannon_loop 2286 14.5 1.485 1.573 25.797 27.489 ot_mini 108 10.5 0.001 0.001 20.926 21.179 qs_ot_get_derivative 108 11.5 0.001 0.001 17.862 18.037 mp_waitall_1 245248 16.5 10.053 16.088 10.053 16.088 multiply_cannon_metrocomm3 54864 15.5 0.067 0.073 6.354 13.336 multiply_cannon_multrec 54864 15.5 4.212 6.414 7.479 10.948 qs_ot_get_p 119 10.4 0.001 0.001 9.887 10.178 rebuild_ks_matrix 119 8.3 0.000 0.000 9.398 9.536 qs_ks_build_kohn_sham_matrix 119 9.3 0.011 0.014 9.397 9.536 qs_ks_update_qs_env 119 7.6 0.001 0.001 8.303 8.427 mp_sum_l 7287 12.8 5.847 7.816 5.847 7.816 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 6.276 6.705 qs_ot_p2m_diag 50 11.0 0.004 0.006 6.505 6.573 multiply_cannon_sync_h2d 54864 15.5 5.780 6.524 5.780 6.524 qs_ot_get_derivative_diag 49 12.0 0.001 0.002 6.012 6.102 qs_rho_update_rho_low 119 7.7 0.001 0.001 5.697 5.813 calculate_rho_elec 119 8.7 0.012 0.017 5.697 5.812 sum_up_and_integrate 119 10.3 0.012 0.014 5.680 5.688 integrate_v_rspace 119 11.3 0.002 0.003 5.668 5.677 cp_dbcsr_syevd 50 12.0 0.003 0.005 5.542 5.543 init_scf_run 11 5.9 0.000 0.001 5.249 5.250 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.249 5.250 cp_fm_diag_elpa 50 13.0 0.000 0.001 5.098 5.099 cp_fm_redistribute_end 50 14.0 2.582 5.031 2.606 5.039 cp_fm_diag_elpa_base 50 14.0 2.419 4.858 2.429 4.878 rs_pw_transfer 974 11.9 0.012 0.013 4.204 4.504 dbcsr_mm_accdrv_process 76910 16.1 1.138 1.795 3.187 4.455 multiply_cannon_metrocomm1 54864 15.5 0.053 0.057 2.220 3.935 density_rs2pw 119 9.7 0.004 0.005 3.462 3.568 calculate_dm_sparse 119 9.5 0.000 0.001 3.084 3.257 apply_preconditioner_dbcsr 119 12.6 0.000 0.001 2.984 3.166 apply_single 119 13.6 0.000 0.000 2.984 3.166 pw_transfer 1439 11.6 0.053 0.060 2.895 3.059 potential_pw2rs 119 12.3 0.004 0.004 2.912 2.999 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 2.818 2.985 make_m2s 4572 13.5 0.054 0.056 2.750 2.888 make_images 4572 14.5 0.133 0.138 2.666 2.804 ot_diis_step 108 11.5 0.006 0.008 2.781 2.782 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.705 2.772 init_scf_loop 11 6.9 0.001 0.006 2.714 2.714 fft3d_ps 1201 14.6 0.377 0.490 2.463 2.627 jit_kernel_multiply 13 15.8 1.989 2.626 1.989 2.626 calculate_first_density_matrix 1 7.0 0.000 0.003 2.603 2.607 wfi_extrapolate 11 7.9 0.001 0.001 2.552 2.552 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.358 2.361 fft_wrap_pw1pw2_140 487 13.2 0.182 0.197 2.164 2.331 acc_transpose_blocks 54864 15.5 0.228 0.250 1.740 2.186 mp_sum_d 4135 12.0 1.500 2.149 1.500 2.149 grid_integrate_task_list 119 12.3 1.992 2.125 1.992 2.125 mp_alltoall_d11v 2130 13.8 1.913 2.099 1.913 2.099 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 1.967 2.022 mp_waitany 12084 13.8 1.516 1.724 1.516 1.724 make_images_sizes 4572 15.5 0.004 0.004 1.347 1.652 mp_alltoall_i44 4572 16.5 1.343 1.648 1.343 1.648 mp_alltoall_z22v 1201 16.6 1.508 1.634 1.508 1.634 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.001 1.514 1.532 prepare_preconditioner 11 7.9 0.000 0.001 1.481 1.512 make_preconditioner 11 8.9 0.000 0.001 1.481 1.512 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.392 1.441 grid_collocate_task_list 119 9.7 1.358 1.426 1.358 1.426 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=62.213000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=431.818182, yerr=1.585054 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6ad6eaa91393e2e3e5bb87bfd4fb06be6378d79a_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 486.637568E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2194560 MPI messages size (bytes): total size 310.646604E+09 min size 0.000000E+00 max size 1.145520E+06 average size 141.553031E+03 MPI breakdown and total messages size (bytes): size <= 128 724648 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 281952 4619501568 32768 < size <= 131072 494448 39143342080 131072 < size <= 4194304 440000 264807943488 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62664. MP_Allreduce 10306 303. MP_Sync 54 MP_Alltoall 2060 1030466. 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 0.035 0.056 46.286 46.287 qs_mol_dyn_low 1 2.0 0.003 0.003 46.003 46.010 qs_forces 11 3.9 0.002 0.004 45.942 45.943 qs_energies 11 4.9 0.002 0.005 44.050 44.055 scf_env_do_scf 11 5.9 0.001 0.002 38.185 38.186 scf_env_do_scf_inner_loop 108 6.5 0.003 0.008 34.786 34.788 dbcsr_multiply_generic 2286 12.5 0.103 0.105 23.005 23.385 qs_scf_new_mos 108 7.5 0.001 0.001 22.809 23.068 qs_scf_loop_do_ot 108 8.5 0.001 0.001 22.809 23.067 velocity_verlet 10 3.0 0.001 0.001 22.312 22.313 ot_scf_mini 108 9.5 0.003 0.003 21.825 21.995 multiply_cannon 2286 13.5 0.208 0.221 16.776 18.461 multiply_cannon_loop 2286 14.5 0.911 0.974 15.404 16.912 ot_mini 108 10.5 0.001 0.001 12.927 13.169 mp_waitall_1 200699 16.5 7.195 12.049 7.195 12.049 qs_ot_get_derivative 108 11.5 0.001 0.001 10.361 10.539 multiply_cannon_metrocomm3 27432 15.5 0.067 0.070 4.237 9.544 rebuild_ks_matrix 119 8.3 0.000 0.000 8.842 8.980 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 8.842 8.980 multiply_cannon_multrec 27432 15.5 1.995 4.264 5.925 8.839 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.823 7.950 qs_ot_get_p 119 10.4 0.001 0.001 5.758 5.989 dbcsr_mm_accdrv_process 47894 16.0 2.964 4.938 3.859 5.720 sum_up_and_integrate 119 10.3 0.025 0.028 5.669 5.681 integrate_v_rspace 119 11.3 0.002 0.003 5.643 5.660 qs_rho_update_rho_low 119 7.7 0.001 0.001 5.482 5.521 calculate_rho_elec 119 8.7 0.021 0.024 5.482 5.520 rs_pw_transfer 974 11.9 0.010 0.011 4.908 5.389 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.971 4.822 init_scf_run 11 5.9 0.000 0.001 4.522 4.523 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.522 4.523 mp_sum_l 7287 12.8 2.296 4.234 2.296 4.234 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.126 4.204 apply_single 119 13.6 0.000 0.000 3.126 4.204 density_rs2pw 119 9.7 0.004 0.004 3.693 4.154 qs_ot_p2m_diag 50 11.0 0.009 0.013 3.928 3.950 make_m2s 4572 13.5 0.052 0.055 3.484 3.738 make_images 4572 14.5 0.201 0.237 3.394 3.646 cp_dbcsr_syevd 50 12.0 0.003 0.003 3.445 3.446 init_scf_loop 11 6.9 0.001 0.006 3.362 3.362 potential_pw2rs 119 12.3 0.006 0.006 3.293 3.333 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.919 2.920 cp_fm_redistribute_end 50 14.0 1.468 2.859 1.481 2.865 multiply_cannon_sync_h2d 27432 15.5 2.151 2.849 2.151 2.849 cp_fm_diag_elpa_base 50 14.0 1.339 2.712 1.373 2.767 pw_transfer 1439 11.6 0.066 0.070 2.646 2.715 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.554 2.624 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.476 2.567 calculate_first_density_matrix 1 7.0 0.001 0.003 2.512 2.513 ot_diis_step 108 11.5 0.011 0.011 2.507 2.508 calculate_dm_sparse 119 9.5 0.000 0.001 2.211 2.300 prepare_preconditioner 11 7.9 0.000 0.000 2.225 2.252 make_preconditioner 11 8.9 0.000 0.002 2.225 2.252 jit_kernel_multiply 11 16.1 0.843 2.194 0.843 2.194 fft3d_ps 1201 14.6 0.525 0.583 2.107 2.178 make_full_inverse_cholesky 11 9.9 0.000 0.001 2.078 2.146 make_images_data 4572 15.5 0.045 0.051 1.454 2.097 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.044 2.047 hybrid_alltoall_any 4725 16.4 0.051 0.113 1.248 1.948 grid_integrate_task_list 119 12.3 1.826 1.945 1.826 1.945 wfi_extrapolate 11 7.9 0.001 0.001 1.913 1.914 mp_alltoall_d11v 2130 13.8 1.688 1.862 1.688 1.862 fft_wrap_pw1pw2_140 487 13.2 0.201 0.211 1.718 1.788 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.705 1.746 mp_allgather_i34 2286 14.5 0.809 1.649 0.809 1.649 make_images_sizes 4572 15.5 0.005 0.005 1.214 1.642 mp_alltoall_i44 4572 16.5 1.210 1.637 1.210 1.637 mp_sendrecv_dv 16779 12.7 1.495 1.545 1.495 1.545 acc_transpose_blocks 27432 15.5 0.110 0.120 1.203 1.509 rs_pw_transfer_PW2RS_50 119 14.3 0.589 0.607 1.236 1.434 grid_collocate_task_list 119 9.7 1.277 1.424 1.277 1.424 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.371 1.380 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.306 1.348 mp_alltoall_z22v 1201 16.6 1.176 1.307 1.176 1.307 rs_pw_transfer_RS2PW_50 119 11.7 0.275 0.293 1.093 1.195 mp_sum_d 4135 12.0 0.661 1.130 0.661 1.130 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.082 1.095 dbcsr_complete_redistribute 329 12.2 0.082 0.110 0.961 1.056 cp_fm_cholesky_invert 11 10.9 1.048 1.053 1.048 1.053 rs_pw_transfer_RS2PW_140 130 11.5 0.139 0.150 0.617 1.035 mp_waitany 5720 13.7 0.611 1.014 0.611 1.014 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 0.994 0.995 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=46.287000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=462.545455, yerr=1.233151 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6ad6eaa91393e2e3e5bb87bfd4fb06be6378d79a_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 520.507392E+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 63497. MP_Allreduce 10155 305. MP_Sync 54 MP_Alltoall 1821 1607886. 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 0.012 0.030 39.403 39.404 qs_mol_dyn_low 1 2.0 0.003 0.003 39.174 39.182 qs_forces 11 3.9 0.002 0.002 39.087 39.089 qs_energies 11 4.9 0.001 0.001 37.298 37.300 scf_env_do_scf 11 5.9 0.000 0.001 31.987 31.988 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 28.586 28.587 velocity_verlet 10 3.0 0.001 0.001 19.213 19.215 dbcsr_multiply_generic 2286 12.5 0.096 0.098 18.336 18.506 qs_scf_new_mos 108 7.5 0.001 0.001 17.904 17.920 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.903 17.919 ot_scf_mini 108 9.5 0.002 0.002 17.067 17.076 multiply_cannon 2286 13.5 0.196 0.201 13.711 14.436 multiply_cannon_loop 2286 14.5 0.642 0.670 12.674 13.415 ot_mini 108 10.5 0.001 0.001 10.071 10.083 qs_ot_get_derivative 108 11.5 0.001 0.001 8.444 8.454 rebuild_ks_matrix 119 8.3 0.000 0.000 7.918 7.955 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.012 7.917 7.954 multiply_cannon_multrec 18288 15.5 1.929 2.843 6.782 7.146 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.970 7.004 dbcsr_mm_accdrv_process 38222 16.0 4.035 5.502 4.769 5.628 mp_waitall_1 158411 16.6 4.414 5.606 4.414 5.606 sum_up_and_integrate 119 10.3 0.031 0.031 5.347 5.352 integrate_v_rspace 119 11.3 0.002 0.003 5.316 5.324 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.874 4.892 calculate_rho_elec 119 8.7 0.031 0.032 4.874 4.892 qs_ot_get_p 119 10.4 0.001 0.001 4.587 4.609 rs_pw_transfer 974 11.9 0.009 0.010 4.282 4.566 init_scf_run 11 5.9 0.000 0.001 4.064 4.064 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.064 4.064 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.156 3.794 density_rs2pw 119 9.7 0.004 0.004 3.286 3.555 init_scf_loop 11 6.9 0.000 0.000 3.370 3.372 qs_ot_p2m_diag 50 11.0 0.012 0.013 3.096 3.103 potential_pw2rs 119 12.3 0.007 0.009 3.009 3.053 make_m2s 4572 13.5 0.044 0.045 2.761 3.015 make_images 4572 14.5 0.191 0.204 2.675 2.927 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.771 2.772 multiply_cannon_metrocomm3 18288 15.5 0.044 0.046 1.613 2.666 pw_transfer 1439 11.6 0.066 0.070 2.547 2.596 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.453 2.501 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.140 2.461 apply_single 119 13.6 0.000 0.000 2.139 2.461 prepare_preconditioner 11 7.9 0.000 0.000 2.409 2.412 make_preconditioner 11 8.9 0.000 0.000 2.409 2.412 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.374 2.384 cp_fm_diag_elpa_base 50 14.0 2.336 2.356 2.371 2.381 calculate_first_density_matrix 1 7.0 0.000 0.000 2.371 2.372 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.218 2.325 fft3d_ps 1201 14.6 0.536 0.554 1.944 1.999 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.989 1.998 jit_kernel_multiply 10 16.2 0.682 1.982 0.682 1.982 calculate_dm_sparse 119 9.5 0.000 0.000 1.966 1.982 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.921 1.923 grid_integrate_task_list 119 12.3 1.789 1.902 1.789 1.902 mp_sum_l 7287 12.8 1.476 1.778 1.476 1.778 fft_wrap_pw1pw2_140 487 13.2 0.250 0.258 1.675 1.724 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.679 1.688 wfi_extrapolate 11 7.9 0.001 0.001 1.600 1.600 make_images_data 4572 15.5 0.045 0.049 1.241 1.598 ot_diis_step 108 11.5 0.011 0.011 1.583 1.583 multiply_cannon_sync_h2d 18288 15.5 1.363 1.564 1.363 1.564 hybrid_alltoall_any 4725 16.4 0.056 0.115 1.019 1.403 grid_collocate_task_list 119 9.7 1.245 1.400 1.245 1.400 acc_transpose_blocks 18288 15.5 0.076 0.078 1.254 1.287 mp_alltoall_z22v 1201 16.6 1.191 1.268 1.191 1.268 mp_sendrecv_dv 11067 12.7 1.215 1.264 1.215 1.264 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.238 1.249 mp_alltoall_d11v 2130 13.8 1.075 1.245 1.075 1.245 make_images_sizes 4572 15.5 0.005 0.005 0.833 1.179 mp_alltoall_i44 4572 16.5 0.828 1.174 0.828 1.174 cp_fm_cholesky_invert 11 10.9 1.078 1.083 1.078 1.083 rs_pw_transfer_PW2RS_50 119 14.3 0.402 0.413 0.928 1.069 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.044 1.065 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.025 1.027 rs_pw_transfer_RS2PW_50 119 11.7 0.243 0.251 0.871 0.969 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.956 0.958 mp_allgather_i34 2286 14.5 0.496 0.941 0.496 0.941 mp_waitany 9880 13.7 0.603 0.873 0.603 0.873 dbcsr_complete_redistribute 329 12.2 0.093 0.103 0.731 0.861 acc_transpose_blocks_kernels 18288 16.5 0.212 0.220 0.802 0.835 rs_pw_transfer_RS2PW_140 130 11.5 0.119 0.122 0.569 0.823 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=39.404000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=495.545455, yerr=1.827250 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6ad6eaa91393e2e3e5bb87bfd4fb06be6378d79a_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 543.051776E+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 63496. 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 0.070 0.109 42.623 42.623 qs_mol_dyn_low 1 2.0 0.003 0.003 42.178 42.185 qs_forces 11 3.9 0.006 0.008 42.114 42.115 qs_energies 11 4.9 0.001 0.001 40.205 40.209 scf_env_do_scf 11 5.9 0.000 0.001 34.414 34.416 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 30.057 30.058 velocity_verlet 10 3.0 0.001 0.001 21.711 21.713 dbcsr_multiply_generic 2286 12.5 0.099 0.104 20.536 20.698 qs_scf_new_mos 108 7.5 0.001 0.001 19.489 19.542 qs_scf_loop_do_ot 108 8.5 0.001 0.001 19.489 19.541 ot_scf_mini 108 9.5 0.002 0.003 18.428 18.479 multiply_cannon 2286 13.5 0.227 0.261 15.530 15.938 multiply_cannon_loop 2286 14.5 0.951 0.992 14.247 14.683 ot_mini 108 10.5 0.001 0.001 10.806 10.863 multiply_cannon_multrec 27432 15.5 2.406 3.083 8.959 9.347 qs_ot_get_derivative 108 11.5 0.001 0.001 8.858 8.906 rebuild_ks_matrix 119 8.3 0.000 0.000 7.912 7.960 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.911 7.959 dbcsr_mm_accdrv_process 47916 15.9 5.299 7.255 6.461 7.673 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.021 7.062 sum_up_and_integrate 119 10.3 0.035 0.037 4.924 4.934 integrate_v_rspace 119 11.3 0.003 0.003 4.889 4.899 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.736 4.778 calculate_rho_elec 119 8.7 0.040 0.046 4.736 4.778 qs_ot_get_p 119 10.4 0.001 0.001 4.703 4.775 mp_waitall_1 137007 16.6 3.441 4.488 3.441 4.488 init_scf_run 11 5.9 0.000 0.001 4.366 4.367 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.366 4.366 init_scf_loop 11 6.9 0.000 0.000 4.322 4.323 rs_pw_transfer 974 11.9 0.009 0.009 3.623 3.819 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.332 3.744 make_m2s 4572 13.5 0.054 0.056 3.245 3.488 make_images 4572 14.5 0.271 0.332 3.137 3.376 prepare_preconditioner 11 7.9 0.000 0.000 3.326 3.334 make_preconditioner 11 8.9 0.000 0.000 3.326 3.334 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.903 3.242 density_rs2pw 119 9.7 0.004 0.004 3.046 3.233 qs_ot_p2m_diag 50 11.0 0.015 0.023 3.005 3.019 pw_transfer 1439 11.6 0.066 0.070 2.583 2.623 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.614 2.615 potential_pw2rs 119 12.3 0.009 0.010 2.571 2.587 calculate_first_density_matrix 1 7.0 0.000 0.000 2.578 2.581 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.490 2.531 calculate_dm_sparse 119 9.5 0.000 0.000 2.373 2.426 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.378 2.404 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.153 2.281 apply_single 119 13.6 0.000 0.000 2.152 2.280 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.219 2.233 cp_fm_diag_elpa_base 50 14.0 2.175 2.195 2.216 2.230 jit_kernel_multiply 10 15.8 1.102 2.094 1.102 2.094 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.051 2.052 fft3d_ps 1201 14.6 0.562 0.617 1.929 1.966 grid_integrate_task_list 119 12.3 1.818 1.922 1.818 1.922 fft_wrap_pw1pw2_140 487 13.2 0.288 0.303 1.877 1.921 multiply_cannon_metrocomm3 27432 15.5 0.039 0.040 0.951 1.911 ot_diis_step 108 11.5 0.012 0.012 1.889 1.890 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.797 1.809 make_images_data 4572 15.5 0.045 0.048 1.428 1.778 wfi_extrapolate 11 7.9 0.001 0.001 1.716 1.716 mp_sum_l 7287 12.8 1.342 1.637 1.342 1.637 acc_transpose_blocks 27432 15.5 0.111 0.113 1.538 1.578 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.447 1.459 hybrid_alltoall_any 4725 16.4 0.062 0.151 1.188 1.456 grid_collocate_task_list 119 9.7 1.248 1.360 1.248 1.360 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.309 1.343 mp_alltoall_d11v 2130 13.8 1.167 1.315 1.167 1.315 dbcsr_complete_redistribute 329 12.2 0.122 0.151 1.022 1.292 mp_alltoall_z22v 1201 16.6 1.251 1.284 1.251 1.284 cp_fm_cholesky_invert 11 10.9 1.222 1.227 1.222 1.227 cp_fm_upper_to_full 72 14.2 0.859 1.222 0.859 1.222 make_images_sizes 4572 15.5 0.005 0.005 0.791 1.113 mp_alltoall_i44 4572 16.5 0.786 1.108 0.786 1.108 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.099 1.103 multiply_cannon_sync_h2d 27432 15.5 0.969 1.045 0.969 1.045 mp_sendrecv_dv 8211 12.7 1.010 1.033 1.010 1.033 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.984 0.992 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.689 0.955 acc_transpose_blocks_kernels 27432 16.5 0.270 0.278 0.901 0.931 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.793 0.870 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=42.623000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=516.545455, yerr=2.641062 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6ad6eaa91393e2e3e5bb87bfd4fb06be6378d79a_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 611.856384E+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 63495. 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 0.016 0.032 33.816 33.817 qs_mol_dyn_low 1 2.0 0.003 0.003 33.631 33.638 qs_forces 11 3.9 0.002 0.002 33.448 33.449 qs_energies 11 4.9 0.001 0.001 31.593 31.597 scf_env_do_scf 11 5.9 0.000 0.001 26.417 26.418 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 23.148 23.148 velocity_verlet 10 3.0 0.001 0.001 17.617 17.628 dbcsr_multiply_generic 2286 12.5 0.092 0.097 13.875 13.992 qs_scf_new_mos 108 7.5 0.001 0.001 13.501 13.541 qs_scf_loop_do_ot 108 8.5 0.001 0.001 13.500 13.540 ot_scf_mini 108 9.5 0.002 0.002 12.704 12.737 multiply_cannon 2286 13.5 0.233 0.243 10.233 10.886 multiply_cannon_loop 2286 14.5 0.334 0.347 9.180 9.449 rebuild_ks_matrix 119 8.3 0.000 0.000 6.957 6.997 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.957 6.996 ot_mini 108 10.5 0.001 0.001 6.839 6.881 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.190 6.226 multiply_cannon_multrec 9144 15.5 1.597 1.816 5.765 6.003 qs_ot_get_derivative 108 11.5 0.001 0.001 5.409 5.445 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.475 4.492 calculate_rho_elec 119 8.7 0.060 0.061 4.474 4.492 sum_up_and_integrate 119 10.3 0.038 0.042 4.470 4.475 integrate_v_rspace 119 11.3 0.003 0.003 4.433 4.438 dbcsr_mm_accdrv_process 12550 15.8 3.300 3.880 4.067 4.140 qs_ot_get_p 119 10.4 0.001 0.001 3.840 3.887 init_scf_run 11 5.9 0.000 0.001 3.604 3.604 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.603 3.604 mp_waitall_1 115863 16.7 2.874 3.357 2.874 3.357 init_scf_loop 11 6.9 0.000 0.000 3.238 3.240 rs_pw_transfer 974 11.9 0.008 0.008 2.798 2.945 density_rs2pw 119 9.7 0.003 0.004 2.644 2.791 make_m2s 4572 13.5 0.034 0.036 2.547 2.736 make_images 4572 14.5 0.267 0.303 2.457 2.644 qs_ot_p2m_diag 50 11.0 0.022 0.023 2.530 2.536 pw_transfer 1439 11.6 0.066 0.069 2.454 2.465 prepare_preconditioner 11 7.9 0.000 0.000 2.389 2.394 make_preconditioner 11 8.9 0.000 0.000 2.389 2.394 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.360 2.373 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.242 2.292 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.274 2.276 calculate_first_density_matrix 1 7.0 0.000 0.000 2.131 2.134 potential_pw2rs 119 12.3 0.010 0.011 2.067 2.073 grid_integrate_task_list 119 12.3 1.904 1.978 1.904 1.978 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.896 1.922 calculate_dm_sparse 119 9.5 0.000 0.000 1.887 1.913 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.885 1.891 cp_fm_diag_elpa_base 50 14.0 1.848 1.872 1.882 1.889 fft_wrap_pw1pw2_140 487 13.2 0.364 0.373 1.801 1.817 fft3d_ps 1201 14.6 0.563 0.575 1.694 1.704 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.613 1.616 jit_kernel_multiply 9 16.1 0.728 1.577 0.728 1.577 make_images_data 4572 15.5 0.039 0.043 1.219 1.553 hybrid_alltoall_any 4725 16.4 0.063 0.175 1.146 1.537 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.512 1.529 grid_collocate_task_list 119 9.7 1.336 1.471 1.336 1.471 wfi_extrapolate 11 7.9 0.001 0.001 1.400 1.400 ot_diis_step 108 11.5 0.012 0.013 1.398 1.398 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.379 1.396 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.257 1.307 apply_single 119 13.6 0.000 0.000 1.257 1.307 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.239 1.239 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.218 1.227 mp_alltoall_d11v 2130 13.8 1.113 1.212 1.113 1.212 cp_fm_cholesky_invert 11 10.9 1.184 1.188 1.184 1.188 make_images_sizes 4572 15.5 0.005 0.005 0.666 1.046 mp_alltoall_i44 4572 16.5 0.661 1.041 0.661 1.041 mp_alltoall_z22v 1201 16.6 1.005 1.033 1.005 1.033 multiply_cannon_metrocomm1 9144 15.5 0.022 0.023 0.873 1.007 acc_transpose_blocks 9144 15.5 0.038 0.040 0.917 0.937 mp_allgather_i34 2286 14.5 0.384 0.931 0.384 0.931 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.866 0.915 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.892 0.896 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.864 0.871 mp_sum_l 7287 12.8 0.672 0.830 0.672 0.830 dbcsr_complete_redistribute 329 12.2 0.163 0.174 0.726 0.771 multiply_cannon_sync_h2d 9144 15.5 0.695 0.763 0.695 0.763 cp_fm_cholesky_decompose 22 10.9 0.745 0.749 0.745 0.749 mp_sendrecv_dv 5355 12.7 0.723 0.748 0.723 0.748 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.694 0.743 acc_transpose_blocks_kernels 9144 16.5 0.116 0.120 0.672 0.689 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=33.817000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=580.090909, yerr=3.553604 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6ad6eaa91393e2e3e5bb87bfd4fb06be6378d79a_performance_tests/08/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 235.585836E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 1388964 0.0% 0.0% 100.0% average stack size 0.0 0.0 105.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 752.705536E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63729. MP_Allreduce 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 0.018 0.037 42.941 42.941 qs_mol_dyn_low 1 2.0 0.003 0.003 42.722 42.729 qs_forces 11 3.9 0.002 0.002 42.660 42.662 qs_energies 11 4.9 0.001 0.001 40.640 40.644 scf_env_do_scf 11 5.9 0.001 0.001 34.508 34.508 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 26.441 26.442 velocity_verlet 10 3.0 0.001 0.001 24.085 24.091 dbcsr_multiply_generic 2286 12.5 0.100 0.103 17.950 18.094 qs_scf_new_mos 108 7.5 0.001 0.001 16.444 16.554 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.443 16.553 ot_scf_mini 108 9.5 0.002 0.002 15.335 15.448 multiply_cannon 2286 13.5 0.303 0.317 13.867 14.868 multiply_cannon_loop 2286 14.5 0.347 0.357 12.546 13.528 ot_mini 108 10.5 0.001 0.001 9.064 9.197 multiply_cannon_multrec 9144 15.5 3.525 5.192 8.707 8.854 init_scf_loop 11 6.9 0.000 0.000 8.038 8.039 rebuild_ks_matrix 119 8.3 0.000 0.000 7.448 7.607 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.447 7.607 qs_ot_get_derivative 108 11.5 0.001 0.001 6.969 7.083 prepare_preconditioner 11 7.9 0.000 0.000 7.048 7.063 make_preconditioner 11 8.9 0.000 0.000 7.048 7.063 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.609 6.941 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.723 6.868 dbcsr_mm_accdrv_process 12550 15.8 4.180 5.694 5.057 6.418 cp_fm_upper_to_full 72 14.2 3.242 4.625 3.242 4.625 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.412 4.463 calculate_rho_elec 119 8.7 0.119 0.121 4.412 4.463 sum_up_and_integrate 119 10.3 0.064 0.066 4.104 4.110 init_scf_run 11 5.9 0.000 0.001 4.084 4.084 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.083 4.083 integrate_v_rspace 119 11.3 0.003 0.004 4.039 4.046 mp_waitall_1 94719 16.7 2.565 3.634 2.565 3.634 qs_ot_get_p 119 10.4 0.001 0.001 3.458 3.611 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.526 2.966 dbcsr_complete_redistribute 329 12.2 0.288 0.292 2.009 2.849 pw_transfer 1439 11.6 0.068 0.069 2.777 2.783 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.678 2.685 make_m2s 4572 13.5 0.038 0.038 2.408 2.618 calculate_first_density_matrix 1 7.0 0.000 0.000 2.507 2.560 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.680 2.511 make_images 4572 14.5 0.350 0.380 2.286 2.495 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.263 2.476 apply_single 119 13.6 0.000 0.000 2.263 2.475 mp_alltoall_i22 627 13.8 1.506 2.374 1.506 2.374 calculate_dm_sparse 119 9.5 0.000 0.000 2.285 2.332 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 1.373 2.328 density_rs2pw 119 9.7 0.004 0.004 2.259 2.271 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.434 2.257 fft_wrap_pw1pw2_140 487 13.2 0.617 0.620 2.236 2.245 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.126 2.127 grid_integrate_task_list 119 12.3 2.060 2.077 2.060 2.077 ot_diis_step 108 11.5 0.014 0.014 2.067 2.067 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.983 1.984 mp_sum_l 7287 12.8 1.131 1.882 1.131 1.882 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.836 1.836 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.767 1.769 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.708 1.766 fft3d_ps 1201 14.6 0.596 0.605 1.696 1.701 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.632 1.667 rs_pw_transfer 974 11.9 0.009 0.009 1.562 1.586 cp_fm_cholesky_invert 11 10.9 1.573 1.576 1.573 1.576 hybrid_alltoall_any 4725 16.4 0.087 0.147 1.204 1.530 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.529 1.529 grid_collocate_task_list 119 9.7 1.508 1.528 1.508 1.528 cp_fm_diag_elpa_base 50 14.0 1.373 1.430 1.527 1.527 potential_pw2rs 119 12.3 0.014 0.014 1.485 1.489 make_images_data 4572 15.5 0.042 0.046 1.174 1.447 wfi_extrapolate 11 7.9 0.001 0.001 1.446 1.446 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.328 1.378 mp_alltoall_d11v 2130 13.8 1.178 1.238 1.178 1.238 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.175 1.196 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.098 1.118 multiply_cannon_sync_h2d 9144 15.5 1.043 1.051 1.043 1.051 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 0.989 1.032 jit_kernel_multiply 6 15.6 0.849 1.018 0.849 1.018 mp_alltoall_z22v 1201 16.6 0.964 0.976 0.964 0.976 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.937 0.952 qs_create_task_list 11 7.9 0.001 0.001 0.933 0.949 generate_qs_task_list 11 8.9 0.368 0.389 0.933 0.949 acc_transpose_blocks 9144 15.5 0.039 0.039 0.916 0.924 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=42.941000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=707.545455, yerr=13.950325 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6ad6eaa91393e2e3e5bb87bfd4fb06be6378d79a_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 501.620736E+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 66218. MP_Allreduce 9776 488. MP_Sync 52 MP_Alltoall 1938 1911174. 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 0.022 0.075 93.237 93.239 qs_mol_dyn_low 1 2.0 0.003 0.007 92.794 92.815 qs_forces 11 3.9 0.003 0.004 92.713 92.714 qs_energies 11 4.9 0.001 0.001 89.580 89.593 scf_env_do_scf 11 5.9 0.000 0.001 79.565 79.568 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 72.309 72.310 dbcsr_multiply_generic 2055 12.4 0.108 0.111 53.244 53.595 qs_scf_new_mos 99 7.5 0.001 0.001 51.883 52.014 qs_scf_loop_do_ot 99 8.5 0.001 0.001 51.883 52.013 ot_scf_mini 99 9.5 0.002 0.002 49.414 49.528 velocity_verlet 10 3.0 0.001 0.002 47.928 47.929 multiply_cannon 2055 13.4 0.180 0.185 42.955 44.448 multiply_cannon_loop 2055 14.4 1.525 1.568 41.608 43.046 ot_mini 99 10.5 0.001 0.001 27.926 28.039 qs_ot_get_derivative 99 11.5 0.001 0.001 21.099 21.224 multiply_cannon_multrec 49320 15.4 12.316 12.972 17.322 17.854 rebuild_ks_matrix 110 8.3 0.000 0.000 16.389 16.551 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.019 16.389 16.551 qs_ks_update_qs_env 110 7.6 0.001 0.001 14.363 14.504 mp_waitall_1 220248 16.4 12.352 13.651 12.352 13.651 qs_ot_get_p 110 10.4 0.001 0.001 12.583 12.743 multiply_cannon_sync_h2d 49320 15.4 10.077 10.807 10.077 10.807 qs_ot_p2m_diag 48 11.0 0.013 0.019 8.808 8.862 qs_rho_update_rho_low 110 7.6 0.001 0.001 8.424 8.582 calculate_rho_elec 110 8.6 0.021 0.026 8.424 8.581 sum_up_and_integrate 110 10.3 0.036 0.043 8.453 8.470 integrate_v_rspace 110 11.3 0.003 0.003 8.416 8.442 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.789 8.340 multiply_cannon_metrocomm3 49320 15.4 0.079 0.082 6.520 8.127 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.256 7.884 apply_single 110 13.6 0.000 0.000 7.256 7.884 init_scf_run 11 5.9 0.000 0.001 7.771 7.771 scf_env_initial_rho_setup 11 6.9 0.005 0.009 7.771 7.771 cp_dbcsr_syevd 48 12.0 0.003 0.003 7.713 7.713 init_scf_loop 11 6.9 0.000 0.000 7.211 7.213 cp_fm_diag_elpa 48 13.0 0.000 0.000 6.809 6.828 cp_fm_diag_elpa_base 48 14.0 6.785 6.808 6.806 6.825 ot_diis_step 99 11.5 0.005 0.005 6.640 6.641 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.264 6.342 rs_pw_transfer 902 11.9 0.011 0.013 5.274 6.067 density_rs2pw 110 9.6 0.004 0.005 5.092 5.891 mp_sum_l 6594 12.7 4.655 5.548 4.655 5.548 make_m2s 4110 13.4 0.062 0.066 5.221 5.346 make_images 4110 14.4 0.178 0.191 5.122 5.248 dbcsr_mm_accdrv_process 87628 16.1 1.924 1.987 4.882 5.165 pw_transfer 1331 11.6 0.056 0.065 4.892 5.022 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 4.803 4.935 prepare_preconditioner 11 7.9 0.000 0.000 4.829 4.850 make_preconditioner 11 8.9 0.000 0.000 4.829 4.850 wfi_extrapolate 11 7.9 0.001 0.001 4.675 4.675 make_full_inverse_cholesky 11 9.9 0.000 0.000 4.573 4.632 multiply_cannon_metrocomm1 49320 15.4 0.062 0.065 2.925 4.485 calculate_dm_sparse 110 9.5 0.001 0.001 4.036 4.135 fft3d_ps 1111 14.6 0.785 0.878 3.997 4.099 potential_pw2rs 110 12.3 0.006 0.007 3.900 3.945 fft_wrap_pw1pw2_140 451 13.1 0.451 0.496 3.801 3.936 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 3.763 3.768 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.554 3.630 grid_integrate_task_list 110 12.3 3.225 3.427 3.225 3.427 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.275 3.323 mp_alltoall_d11v 2046 13.8 2.559 3.067 2.559 3.067 calculate_first_density_matrix 1 7.0 0.000 0.001 2.951 2.963 mp_alltoall_z22v 1111 16.6 2.737 2.923 2.737 2.923 jit_kernel_multiply 13 15.9 2.682 2.725 2.682 2.725 mp_waitany 14300 13.8 1.939 2.650 1.939 2.650 cp_fm_cholesky_invert 11 10.9 2.540 2.545 2.540 2.545 make_images_data 4110 15.4 0.044 0.047 2.216 2.443 grid_collocate_task_list 110 9.6 2.159 2.395 2.159 2.395 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.161 2.182 acc_transpose_blocks 49320 15.4 0.210 0.216 2.108 2.168 hybrid_alltoall_any 4261 16.3 0.083 0.481 1.905 2.160 mp_sum_d 3889 11.9 1.456 1.972 1.456 1.972 make_images_sizes 4110 15.4 0.004 0.005 1.456 1.951 mp_alltoall_i44 4110 16.4 1.452 1.947 1.452 1.947 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.880 1.905 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=93.239000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.545455, yerr=2.775334 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6ad6eaa91393e2e3e5bb87bfd4fb06be6378d79a_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 587.997184E+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 66437. MP_Allreduce 9775 566. MP_Sync 52 MP_Alltoall 1717 2882115. 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 0.062 0.095 79.784 79.793 qs_mol_dyn_low 1 2.0 0.003 0.004 79.296 79.306 qs_forces 11 3.9 0.003 0.003 79.225 79.226 qs_energies 11 4.9 0.002 0.006 75.629 75.642 scf_env_do_scf 11 5.9 0.001 0.002 66.392 66.396 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 56.702 56.703 velocity_verlet 10 3.0 0.001 0.002 41.446 41.447 dbcsr_multiply_generic 2055 12.4 0.120 0.125 40.394 41.132 qs_scf_new_mos 99 7.5 0.001 0.001 38.273 38.415 qs_scf_loop_do_ot 99 8.5 0.001 0.001 38.272 38.414 ot_scf_mini 99 9.5 0.003 0.004 36.434 36.556 multiply_cannon 2055 13.4 0.222 0.243 31.904 34.621 multiply_cannon_loop 2055 14.4 0.930 0.950 30.288 32.869 ot_mini 99 10.5 0.001 0.001 20.420 20.584 multiply_cannon_multrec 24660 15.4 7.482 9.405 13.670 15.458 rebuild_ks_matrix 110 8.3 0.000 0.000 15.049 15.170 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.018 15.049 15.169 qs_ot_get_derivative 99 11.5 0.001 0.001 14.365 14.496 mp_waitall_1 176588 16.5 9.277 14.410 9.277 14.410 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.185 13.284 multiply_cannon_metrocomm3 24660 15.4 0.068 0.071 5.883 11.186 init_scf_loop 11 6.9 0.001 0.004 9.638 9.640 qs_ot_get_p 110 10.4 0.001 0.002 8.941 9.118 multiply_cannon_sync_h2d 24660 15.4 6.933 8.200 6.933 8.200 sum_up_and_integrate 110 10.3 0.052 0.059 7.714 7.726 integrate_v_rspace 110 11.3 0.002 0.003 7.661 7.676 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.425 7.443 calculate_rho_elec 110 8.6 0.040 0.048 7.424 7.442 prepare_preconditioner 11 7.9 0.000 0.000 7.418 7.438 make_preconditioner 11 8.9 0.000 0.002 7.417 7.438 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.623 7.383 apply_single 110 13.6 0.000 0.001 6.623 7.383 make_full_inverse_cholesky 11 9.9 0.000 0.001 6.974 7.152 init_scf_run 11 5.9 0.000 0.001 6.622 6.623 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.622 6.623 qs_ot_p2m_diag 48 11.0 0.029 0.044 6.507 6.554 dbcsr_mm_accdrv_process 52282 16.1 4.542 5.670 6.033 6.330 cp_dbcsr_syevd 48 12.0 0.003 0.005 5.959 5.960 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 5.179 5.910 ot_diis_step 99 11.5 0.010 0.010 5.889 5.890 make_m2s 4110 13.4 0.057 0.060 5.036 5.478 make_images 4110 14.4 0.401 0.445 4.926 5.363 rs_pw_transfer 902 11.9 0.012 0.013 4.407 5.265 density_rs2pw 110 9.6 0.005 0.005 4.434 5.134 cp_fm_diag_elpa 48 13.0 0.000 0.001 5.121 5.131 cp_fm_diag_elpa_base 48 14.0 5.060 5.081 5.116 5.126 pw_transfer 1331 11.6 0.068 0.076 4.383 4.605 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 4.275 4.501 qs_ot_get_derivative_diag 47 12.0 0.001 0.002 3.939 4.013 cp_fm_cholesky_invert 11 10.9 3.910 3.927 3.910 3.927 wfi_extrapolate 11 7.9 0.001 0.001 3.926 3.926 mp_sum_l 6594 12.7 2.830 3.733 2.830 3.733 fft_wrap_pw1pw2_140 451 13.1 0.516 0.531 3.282 3.507 fft3d_ps 1111 14.6 1.110 1.335 3.253 3.495 potential_pw2rs 110 12.3 0.008 0.009 3.389 3.463 grid_integrate_task_list 110 12.3 3.135 3.391 3.135 3.391 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.369 3.374 calculate_dm_sparse 110 9.5 0.001 0.001 3.173 3.207 make_images_data 4110 15.4 0.047 0.051 2.622 3.112 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.958 2.998 hybrid_alltoall_any 4261 16.3 0.103 0.445 2.275 2.891 mp_alltoall_d11v 2046 13.8 2.166 2.744 2.166 2.744 calculate_first_density_matrix 1 7.0 0.001 0.002 2.558 2.560 grid_collocate_task_list 110 9.6 2.095 2.541 2.095 2.541 qs_ot_get_orbitals 99 10.5 0.001 0.001 2.313 2.401 mp_allgather_i34 2055 14.4 0.942 2.064 0.942 2.064 multiply_cannon_metrocomm4 22605 15.4 0.078 0.084 0.808 2.052 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.027 2.044 mp_waitany 10164 13.8 1.419 2.039 1.419 2.039 mp_alltoall_z22v 1111 16.6 1.822 1.989 1.822 1.989 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.961 1.962 cp_fm_cholesky_decompose 22 10.9 1.855 1.861 1.855 1.861 mp_irecv_dv 57340 16.2 0.681 1.825 0.681 1.825 jit_kernel_multiply 9 16.3 1.141 1.788 1.141 1.788 rs_pw_transfer_RS2PW_140 121 11.5 0.208 0.220 1.031 1.751 dbcsr_complete_redistribute 325 12.2 0.226 0.281 1.380 1.653 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.593 1.618 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=79.793000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=554.818182, yerr=6.766471 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6ad6eaa91393e2e3e5bb87bfd4fb06be6378d79a_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 658.919424E+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 66428. 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 0.033 0.132 68.414 68.415 qs_mol_dyn_low 1 2.0 0.003 0.006 67.903 67.916 qs_forces 11 3.9 0.006 0.027 67.831 67.832 qs_energies 11 4.9 0.004 0.030 64.470 64.475 scf_env_do_scf 11 5.9 0.001 0.001 55.961 55.961 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 45.635 45.636 velocity_verlet 10 3.0 0.001 0.002 36.598 36.600 dbcsr_multiply_generic 2055 12.4 0.108 0.115 30.129 30.430 qs_scf_new_mos 99 7.5 0.001 0.001 28.804 28.910 qs_scf_loop_do_ot 99 8.5 0.001 0.001 28.803 28.909 ot_scf_mini 99 9.5 0.003 0.004 27.500 27.624 multiply_cannon 2055 13.4 0.211 0.220 22.565 23.792 multiply_cannon_loop 2055 14.4 0.621 0.634 21.138 22.419 ot_mini 99 10.5 0.001 0.001 14.759 14.886 rebuild_ks_matrix 110 8.3 0.000 0.000 13.493 13.657 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 13.493 13.657 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.883 12.031 mp_waitall_1 139946 16.5 8.400 11.503 8.400 11.503 multiply_cannon_multrec 16440 15.4 3.827 4.938 9.541 10.560 init_scf_loop 11 6.9 0.001 0.004 10.286 10.287 qs_ot_get_derivative 99 11.5 0.001 0.002 10.141 10.268 prepare_preconditioner 11 7.9 0.000 0.001 8.397 8.415 make_preconditioner 11 8.9 0.000 0.001 8.397 8.415 make_full_inverse_cholesky 11 9.9 0.000 0.001 7.618 8.033 qs_ot_get_p 110 10.4 0.001 0.001 7.681 7.827 multiply_cannon_metrocomm3 16440 15.4 0.042 0.046 4.544 7.653 sum_up_and_integrate 110 10.3 0.060 0.061 7.454 7.470 integrate_v_rspace 110 11.3 0.003 0.003 7.393 7.410 qs_rho_update_rho_low 110 7.6 0.001 0.002 6.962 6.975 calculate_rho_elec 110 8.6 0.059 0.059 6.962 6.975 init_scf_run 11 5.9 0.000 0.001 5.932 5.932 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.931 5.932 dbcsr_mm_accdrv_process 34862 16.1 4.598 5.276 5.570 5.763 make_m2s 4110 13.4 0.050 0.051 5.184 5.622 apply_preconditioner_dbcsr 110 12.6 0.000 0.001 5.090 5.601 apply_single 110 13.6 0.000 0.000 5.090 5.601 qs_ot_p2m_diag 48 11.0 0.042 0.044 5.568 5.578 make_images 4110 14.4 0.393 0.515 5.067 5.504 density_rs2pw 110 9.6 0.004 0.005 3.960 5.180 cp_dbcsr_syevd 48 12.0 0.003 0.003 5.137 5.139 rs_pw_transfer 902 11.9 0.010 0.011 3.882 5.022 ot_diis_step 99 11.5 0.011 0.013 4.578 4.579 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.362 4.373 cp_fm_diag_elpa_base 48 14.0 4.278 4.321 4.360 4.371 pw_transfer 1331 11.6 0.066 0.074 4.224 4.263 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 4.116 4.158 multiply_cannon_sync_h2d 16440 15.4 3.623 4.129 3.623 4.129 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.412 4.064 cp_fm_cholesky_invert 11 10.9 3.665 3.671 3.665 3.671 make_images_data 4110 15.4 0.043 0.047 2.937 3.578 grid_integrate_task_list 110 12.3 3.189 3.401 3.189 3.401 fft_wrap_pw1pw2_140 451 13.1 0.633 0.642 3.345 3.387 wfi_extrapolate 11 7.9 0.001 0.001 3.386 3.386 hybrid_alltoall_any 4261 16.3 0.106 0.375 2.541 3.384 potential_pw2rs 110 12.3 0.011 0.012 3.078 3.116 fft3d_ps 1111 14.6 1.094 1.103 2.943 2.983 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.869 2.871 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.770 2.833 mp_waitany 17072 13.8 1.382 2.645 1.382 2.645 calculate_dm_sparse 110 9.5 0.001 0.001 2.590 2.616 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.479 2.538 grid_collocate_task_list 110 9.6 2.129 2.520 2.129 2.520 mp_alltoall_d11v 2046 13.8 2.140 2.500 2.140 2.500 multiply_cannon_metrocomm4 14385 15.4 0.046 0.049 0.897 2.495 mp_sum_l 6594 12.7 1.839 2.490 1.839 2.490 calculate_first_density_matrix 1 7.0 0.000 0.003 2.414 2.415 mp_irecv_dv 48980 15.7 0.827 2.368 0.827 2.368 dbcsr_complete_redistribute 325 12.2 0.310 0.335 1.744 2.236 rs_pw_transfer_RS2PW_140 121 11.5 0.177 0.182 1.007 2.211 cp_fm_cholesky_decompose 22 10.9 2.076 2.100 2.076 2.100 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 2.046 2.048 mp_allgather_i34 2055 14.4 0.750 1.978 0.750 1.978 jit_kernel_multiply 10 16.5 0.584 1.891 0.584 1.891 cp_fm_upper_to_full 70 14.2 1.423 1.879 1.423 1.879 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.810 1.828 mp_alltoall_z22v 1111 16.6 1.638 1.681 1.638 1.681 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.174 1.653 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.480 1.489 rs_gather_matrices 110 12.3 0.233 0.261 1.040 1.485 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.363 1.455 make_images_sizes 4110 15.4 0.005 0.005 0.896 1.417 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.400 1.414 mp_alltoall_i44 4110 16.4 0.891 1.412 0.891 1.412 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=68.415000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=624.545455, yerr=8.998622 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6ad6eaa91393e2e3e5bb87bfd4fb06be6378d79a_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 722.731008E+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 66426. 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 0.014 0.033 73.654 73.655 qs_mol_dyn_low 1 2.0 0.003 0.003 73.385 73.396 qs_forces 11 3.9 0.003 0.003 73.315 73.316 qs_energies 11 4.9 0.001 0.001 69.730 69.735 scf_env_do_scf 11 5.9 0.000 0.001 60.893 60.895 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 47.219 47.219 velocity_verlet 10 3.0 0.001 0.001 41.204 41.206 dbcsr_multiply_generic 2055 12.4 0.114 0.119 31.211 31.463 qs_scf_new_mos 99 7.5 0.001 0.001 30.449 30.542 qs_scf_loop_do_ot 99 8.5 0.001 0.001 30.449 30.542 ot_scf_mini 99 9.5 0.003 0.003 28.733 28.825 multiply_cannon 2055 13.4 0.243 0.257 22.947 24.352 multiply_cannon_loop 2055 14.4 0.892 0.914 21.332 21.966 ot_mini 99 10.5 0.001 0.001 15.449 15.560 multiply_cannon_multrec 24660 15.4 4.240 6.800 12.693 13.907 init_scf_loop 11 6.9 0.000 0.000 13.619 13.620 rebuild_ks_matrix 110 8.3 0.000 0.000 13.232 13.344 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 13.232 13.343 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.678 11.771 prepare_preconditioner 11 7.9 0.000 0.000 11.754 11.768 make_preconditioner 11 8.9 0.000 0.000 11.754 11.768 make_full_inverse_cholesky 11 9.9 0.000 0.000 9.916 11.439 qs_ot_get_derivative 99 11.5 0.001 0.001 11.169 11.259 dbcsr_mm_accdrv_process 52304 16.0 6.810 8.038 8.306 9.222 qs_ot_get_p 110 10.4 0.001 0.001 8.058 8.188 mp_waitall_1 121746 16.5 5.581 7.606 5.581 7.606 sum_up_and_integrate 110 10.3 0.067 0.070 7.364 7.377 integrate_v_rspace 110 11.3 0.003 0.003 7.297 7.312 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.103 7.113 calculate_rho_elec 110 8.6 0.078 0.082 7.103 7.112 make_m2s 4110 13.4 0.059 0.061 6.440 6.822 make_images 4110 14.4 0.577 0.695 6.297 6.675 init_scf_run 11 5.9 0.000 0.001 5.971 5.972 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.971 5.971 qs_ot_p2m_diag 48 11.0 0.055 0.064 5.849 5.864 cp_dbcsr_syevd 48 12.0 0.003 0.003 5.305 5.306 cp_fm_upper_to_full 70 14.2 3.433 5.003 3.433 5.003 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.486 4.496 cp_fm_diag_elpa_base 48 14.0 4.310 4.375 4.482 4.493 density_rs2pw 110 9.6 0.004 0.004 3.892 4.410 pw_transfer 1331 11.6 0.066 0.077 4.322 4.383 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 4.214 4.279 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.133 4.238 apply_single 110 13.6 0.000 0.000 4.133 4.238 ot_diis_step 99 11.5 0.011 0.011 4.237 4.237 dbcsr_complete_redistribute 325 12.2 0.411 0.447 2.950 4.157 rs_pw_transfer 902 11.9 0.010 0.011 3.487 4.013 make_images_data 4110 15.4 0.046 0.050 3.274 3.859 cp_fm_cholesky_invert 11 10.9 3.836 3.847 3.836 3.847 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.614 3.658 hybrid_alltoall_any 4261 16.3 0.121 0.459 2.693 3.654 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.287 3.487 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.406 3.458 fft_wrap_pw1pw2_140 451 13.1 0.664 0.686 3.365 3.435 grid_integrate_task_list 110 12.3 3.231 3.414 3.231 3.414 multiply_cannon_metrocomm3 24660 15.4 0.036 0.037 1.604 3.353 wfi_extrapolate 11 7.9 0.001 0.001 3.350 3.350 multiply_cannon_sync_h2d 24660 15.4 3.168 3.312 3.168 3.312 calculate_dm_sparse 110 9.5 0.001 0.001 3.054 3.085 fft3d_ps 1111 14.6 1.086 1.117 2.981 3.018 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.828 3.012 mp_alltoall_i22 605 13.7 1.731 2.956 1.731 2.956 potential_pw2rs 110 12.3 0.013 0.013 2.922 2.951 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.916 2.918 mp_alltoall_d11v 2046 13.8 2.180 2.663 2.180 2.663 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.503 2.541 calculate_first_density_matrix 1 7.0 0.000 0.000 2.513 2.516 grid_collocate_task_list 110 9.6 2.224 2.511 2.224 2.511 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 2.280 2.281 cp_fm_cholesky_decompose 22 10.9 2.241 2.281 2.241 2.281 jit_kernel_multiply 10 15.9 1.163 1.910 1.163 1.910 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.855 1.884 mp_allgather_i34 2055 14.4 0.715 1.823 0.715 1.823 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.800 1.816 mp_alltoall_z22v 1111 16.6 1.679 1.756 1.679 1.756 mp_waitany 13376 13.8 1.224 1.732 1.224 1.732 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.597 1.698 mp_sum_l 6594 12.7 1.142 1.658 1.142 1.658 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.596 1.611 multiply_cannon_metrocomm4 20550 15.4 0.060 0.063 0.840 1.546 acc_transpose_blocks 24660 15.4 0.104 0.106 1.513 1.534 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=73.655000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=684.818182, yerr=3.737237 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6ad6eaa91393e2e3e5bb87bfd4fb06be6378d79a_performance_tests/13/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 807.299199E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 1438408 0.0% 0.0% 100.0% average stack size 0.0 0.0 684.2 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 834.449408E+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 66424. 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 0.042 0.056 64.655 64.656 qs_mol_dyn_low 1 2.0 0.003 0.003 64.263 64.273 qs_forces 11 3.9 0.021 0.023 64.190 64.191 qs_energies 11 4.9 0.001 0.001 60.325 60.332 scf_env_do_scf 11 5.9 0.000 0.001 51.213 51.213 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 41.434 41.435 velocity_verlet 10 3.0 0.001 0.001 35.802 35.805 dbcsr_multiply_generic 2055 12.4 0.106 0.112 25.703 25.978 qs_scf_new_mos 99 7.5 0.001 0.001 24.823 24.913 qs_scf_loop_do_ot 99 8.5 0.001 0.001 24.823 24.913 ot_scf_mini 99 9.5 0.002 0.002 23.446 23.516 multiply_cannon 2055 13.4 0.247 0.261 18.772 20.297 multiply_cannon_loop 2055 14.4 0.325 0.338 17.089 17.648 rebuild_ks_matrix 110 8.3 0.000 0.000 12.900 12.987 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 12.900 12.986 ot_mini 99 10.5 0.001 0.001 12.303 12.359 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.467 11.544 mp_waitall_1 103326 16.6 7.683 10.019 7.683 10.019 init_scf_loop 11 6.9 0.000 0.000 9.718 9.719 multiply_cannon_multrec 8220 15.4 3.267 4.724 7.763 8.975 qs_ot_get_derivative 99 11.5 0.001 0.001 8.231 8.294 prepare_preconditioner 11 7.9 0.000 0.000 7.922 7.931 make_preconditioner 11 8.9 0.000 0.000 7.922 7.931 make_full_inverse_cholesky 11 9.9 0.000 0.000 7.422 7.543 sum_up_and_integrate 110 10.3 0.080 0.082 7.224 7.238 integrate_v_rspace 110 11.3 0.003 0.003 7.144 7.157 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.124 7.141 calculate_rho_elec 110 8.6 0.115 0.116 7.124 7.141 qs_ot_get_p 110 10.4 0.001 0.001 6.867 6.953 init_scf_run 11 5.9 0.000 0.001 5.851 5.852 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.851 5.851 make_m2s 4110 13.4 0.039 0.040 5.192 5.683 make_images 4110 14.4 0.639 0.690 5.060 5.551 dbcsr_mm_accdrv_process 17442 15.9 2.997 4.171 4.367 5.369 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 3.236 5.100 qs_ot_p2m_diag 48 11.0 0.081 0.084 5.044 5.052 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.634 4.634 pw_transfer 1331 11.6 0.066 0.071 4.537 4.555 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 4.430 4.452 cp_fm_cholesky_invert 11 10.9 4.318 4.324 4.318 4.324 density_rs2pw 110 9.6 0.004 0.004 3.689 4.068 ot_diis_step 99 11.5 0.012 0.012 4.014 4.014 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.888 3.983 apply_single 110 13.6 0.000 0.000 3.887 3.982 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.890 3.897 cp_fm_diag_elpa_base 48 14.0 3.821 3.858 3.887 3.895 fft_wrap_pw1pw2_140 451 13.1 0.831 0.846 3.686 3.710 make_images_data 4110 15.4 0.038 0.044 2.946 3.674 grid_integrate_task_list 110 12.3 3.355 3.492 3.355 3.492 rs_pw_transfer 902 11.9 0.010 0.010 2.912 3.301 hybrid_alltoall_any 4261 16.3 0.201 0.865 2.741 3.268 wfi_extrapolate 11 7.9 0.001 0.001 3.163 3.163 multiply_cannon_sync_h2d 8220 15.4 2.857 3.051 2.857 3.051 fft3d_ps 1111 14.6 1.143 1.155 2.978 2.989 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.811 2.813 calculate_dm_sparse 110 9.5 0.001 0.001 2.719 2.753 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.703 2.703 potential_pw2rs 110 12.3 0.015 0.016 2.609 2.620 grid_collocate_task_list 110 9.6 2.323 2.585 2.323 2.585 calculate_first_density_matrix 1 7.0 0.000 0.000 2.568 2.570 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.422 2.453 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.401 2.433 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.379 2.416 mp_alltoall_d11v 2046 13.8 2.110 2.374 2.110 2.374 mp_allgather_i34 2055 14.4 0.868 2.207 0.868 2.207 cp_fm_cholesky_decompose 22 10.9 2.072 2.093 2.072 2.093 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.766 1.995 multiply_cannon_metrocomm1 8220 15.4 0.021 0.022 1.387 1.947 dbcsr_complete_redistribute 325 12.2 0.552 0.594 1.720 1.826 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.767 1.777 mp_waitany 9240 13.8 1.238 1.678 1.238 1.678 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.540 1.668 mp_alltoall_z22v 1111 16.6 1.597 1.615 1.597 1.615 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.572 1.587 make_images_sizes 4110 15.4 0.005 0.005 0.808 1.436 mp_alltoall_i44 4110 16.4 0.804 1.432 0.804 1.432 multiply_cannon_metrocomm4 6165 15.4 0.018 0.020 0.479 1.367 jit_kernel_multiply 7 15.7 1.058 1.364 1.058 1.364 mp_sum_l 6594 12.7 1.050 1.341 1.050 1.341 qs_create_task_list 11 7.9 0.002 0.003 1.230 1.336 generate_qs_task_list 11 8.9 0.373 0.440 1.228 1.334 mp_irecv_dv 24056 15.7 0.458 1.325 0.458 1.325 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.272 1.313 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=64.656000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=786.363636, yerr=9.819024 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6ad6eaa91393e2e3e5bb87bfd4fb06be6378d79a_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 1.340219E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67104. MP_Allreduce 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 0.090 0.108 95.495 95.495 qs_mol_dyn_low 1 2.0 0.003 0.003 94.973 94.982 qs_forces 11 3.9 0.013 0.014 94.411 94.411 qs_energies 11 4.9 0.001 0.001 90.012 90.015 scf_env_do_scf 11 5.9 0.000 0.001 79.380 79.380 velocity_verlet 10 3.0 0.001 0.001 60.535 60.602 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 48.765 48.765 init_scf_loop 11 6.9 0.000 0.000 30.529 30.531 dbcsr_multiply_generic 2055 12.4 0.119 0.125 29.985 30.055 qs_scf_new_mos 99 7.5 0.001 0.001 29.022 29.073 qs_scf_loop_do_ot 99 8.5 0.001 0.001 29.022 29.072 prepare_preconditioner 11 7.9 0.000 0.000 28.358 28.365 make_preconditioner 11 8.9 0.000 0.000 28.358 28.365 make_full_inverse_cholesky 11 9.9 0.000 0.000 22.238 27.785 ot_scf_mini 99 9.5 0.002 0.002 27.208 27.239 multiply_cannon 2055 13.4 0.351 0.380 21.889 22.803 multiply_cannon_loop 2055 14.4 0.344 0.347 19.958 20.259 cp_fm_upper_to_full 70 14.2 12.737 18.565 12.737 18.565 rebuild_ks_matrix 110 8.3 0.000 0.000 15.121 15.156 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 15.120 15.155 ot_mini 99 10.5 0.001 0.001 14.546 14.579 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.584 13.616 dbcsr_complete_redistribute 325 12.2 1.018 1.040 7.920 11.325 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.769 10.172 qs_ot_get_derivative 99 11.5 0.001 0.001 9.889 9.925 multiply_cannon_multrec 8220 15.4 4.383 4.558 9.662 9.764 mp_waitall_1 84994 16.7 8.665 9.629 8.665 9.629 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.104 9.473 mp_alltoall_i22 605 13.7 5.710 9.123 5.710 9.123 qs_rho_update_rho_low 110 7.6 0.001 0.001 8.640 8.677 calculate_rho_elec 110 8.6 0.227 0.227 8.640 8.676 sum_up_and_integrate 110 10.3 0.150 0.151 8.089 8.104 integrate_v_rspace 110 11.3 0.003 0.003 7.938 7.955 qs_ot_get_p 110 10.4 0.001 0.001 7.274 7.312 make_m2s 4110 13.4 0.043 0.043 6.233 6.767 make_images 4110 14.4 0.878 0.928 6.042 6.574 cp_fm_cholesky_invert 11 10.9 6.313 6.318 6.313 6.318 init_scf_run 11 5.9 0.000 0.001 6.229 6.229 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.229 6.229 pw_transfer 1331 11.6 0.075 0.076 5.808 5.816 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 5.691 5.698 qs_ot_p2m_diag 48 11.0 0.151 0.156 5.459 5.467 dbcsr_mm_accdrv_process 11614 15.7 3.286 3.722 5.136 5.367 multiply_cannon_metrocomm3 8220 15.4 0.018 0.019 4.883 5.227 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.729 5.211 apply_single 110 13.6 0.000 0.000 4.729 5.211 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.966 4.967 fft_wrap_pw1pw2_140 451 13.1 1.334 1.339 4.823 4.833 ot_diis_step 99 11.5 0.015 0.016 4.624 4.624 density_rs2pw 110 9.6 0.004 0.004 4.510 4.559 make_images_data 4110 15.4 0.041 0.044 3.496 4.401 hybrid_alltoall_any 4261 16.3 0.256 0.552 3.438 4.321 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.223 4.224 cp_fm_diag_elpa_base 48 14.0 3.643 3.880 4.221 4.221 multiply_cannon_sync_h2d 8220 15.4 3.947 3.952 3.947 3.952 grid_integrate_task_list 110 12.3 3.690 3.763 3.690 3.763 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.713 3.714 wfi_extrapolate 11 7.9 0.001 0.001 3.699 3.699 fft3d_ps 1111 14.6 1.305 1.316 3.625 3.629 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.123 3.564 calculate_dm_sparse 110 9.5 0.001 0.001 3.278 3.308 rs_pw_transfer 902 11.9 0.011 0.011 3.072 3.101 potential_pw2rs 110 12.3 0.021 0.022 3.026 3.031 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.003 3.005 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.752 2.769 grid_collocate_task_list 110 9.6 2.681 2.690 2.681 2.690 cp_fm_cholesky_decompose 22 10.9 2.655 2.678 2.655 2.678 mp_alltoall_d11v 2046 13.8 2.462 2.573 2.462 2.573 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.424 2.431 calculate_first_density_matrix 1 7.0 0.000 0.000 2.387 2.388 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.241 2.294 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.138 2.143 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.087 2.141 mp_alltoall_z22v 1111 16.6 2.068 2.074 2.068 2.074 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.933 1.961 qs_create_task_list 11 7.9 0.016 0.017 1.908 1.954 generate_qs_task_list 11 8.9 0.731 0.784 1.893 1.939 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=95.495000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1213.000000, yerr=50.190546 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6ad6eaa91393e2e3e5bb87bfd4fb06be6378d79a_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 630.423552E+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 57910. MP_Allreduce 11059 797. MP_Sync 87 MP_Alltoall 2226 2406411. 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 0.016 0.034 210.175 210.176 qs_mol_dyn_low 1 2.0 0.003 0.004 209.712 209.724 qs_forces 11 3.9 0.006 0.006 209.626 209.627 qs_energies 11 4.9 0.001 0.001 203.999 204.015 scf_env_do_scf 11 5.9 0.001 0.001 187.301 187.305 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 166.154 166.157 qs_scf_new_mos 117 7.6 0.001 0.001 125.717 125.969 qs_scf_loop_do_ot 117 8.6 0.001 0.001 125.716 125.969 dbcsr_multiply_generic 2507 12.6 0.177 0.182 124.884 125.302 velocity_verlet 10 3.0 0.001 0.002 125.017 125.019 ot_scf_mini 117 9.6 0.003 0.003 119.124 119.355 multiply_cannon 2507 13.6 0.237 0.251 101.216 102.480 multiply_cannon_loop 2507 14.6 2.095 2.141 98.979 100.299 ot_mini 117 10.6 0.001 0.001 66.501 66.751 multiply_cannon_multrec 60168 15.6 33.255 35.759 41.461 43.311 qs_ot_get_derivative 117 11.6 0.001 0.001 41.733 41.977 rebuild_ks_matrix 128 8.3 0.001 0.001 33.932 34.181 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 33.931 34.181 mp_waitall_1 267128 16.5 28.896 32.101 28.896 32.101 qs_ot_get_p 128 10.4 0.001 0.001 30.648 30.890 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.446 30.677 multiply_cannon_sync_h2d 60168 15.6 27.311 28.774 27.311 28.774 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.289 24.972 apply_single 128 13.6 0.001 0.001 24.288 24.971 ot_diis_step 117 11.6 0.007 0.008 24.489 24.491 qs_ot_p2m_diag 83 11.4 0.079 0.091 24.001 24.065 cp_dbcsr_syevd 83 12.4 0.005 0.005 21.527 21.528 init_scf_loop 11 6.9 0.000 0.001 21.071 21.073 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 20.118 20.361 cp_fm_diag_elpa 83 13.4 0.000 0.001 18.470 18.502 cp_fm_diag_elpa_base 83 14.4 18.387 18.427 18.465 18.499 multiply_cannon_metrocomm3 60168 15.6 0.112 0.117 15.821 18.039 prepare_preconditioner 11 7.9 0.000 0.000 16.453 16.499 make_preconditioner 11 8.9 0.000 0.000 16.452 16.499 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.629 15.837 sum_up_and_integrate 128 10.3 0.089 0.108 14.442 14.460 make_m2s 5014 13.6 0.105 0.114 14.011 14.436 integrate_v_rspace 128 11.3 0.004 0.005 14.352 14.374 make_images 5014 14.6 0.403 0.423 13.832 14.266 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.972 14.154 calculate_rho_elec 128 8.7 0.046 0.065 13.971 14.153 init_scf_run 11 5.9 0.000 0.001 12.564 12.565 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.564 12.565 density_rs2pw 128 9.7 0.006 0.007 7.252 10.427 mp_sum_l 7950 12.9 8.386 9.531 8.386 9.531 cp_fm_cholesky_invert 11 10.9 9.437 9.445 9.437 9.445 wfi_extrapolate 11 7.9 0.001 0.001 9.250 9.250 rs_pw_transfer 1046 11.9 0.016 0.018 6.024 9.225 calculate_dm_sparse 128 9.5 0.001 0.001 8.446 8.535 dbcsr_mm_accdrv_process 124484 16.2 3.142 3.259 7.769 8.263 multiply_cannon_metrocomm1 60168 15.6 0.090 0.094 6.342 8.172 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.026 8.155 pw_transfer 1547 11.6 0.075 0.089 7.855 8.033 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.815 7.903 fft_wrap_pw1pw2 1291 12.7 0.011 0.012 7.650 7.827 make_images_data 5014 15.6 0.067 0.073 6.875 7.697 grid_integrate_task_list 128 12.3 7.053 7.657 7.053 7.657 hybrid_alltoall_any 5200 16.5 0.292 2.266 5.990 7.348 fft_wrap_pw1pw2_140 523 13.2 1.279 1.327 6.606 6.783 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.697 6.708 mp_alltoall_d11v 2415 14.1 4.323 6.006 4.323 6.006 grid_collocate_task_list 128 9.7 4.672 5.969 4.672 5.969 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.828 5.945 mp_waitany 16020 13.9 2.698 5.912 2.698 5.912 fft3d_ps 1291 14.7 2.134 2.770 5.406 5.717 rs_pw_transfer_RS2PW_140 139 11.5 0.282 0.301 2.150 5.348 potential_pw2rs 128 12.3 0.009 0.011 4.791 4.816 cp_fm_cholesky_decompose 22 10.9 4.779 4.793 4.779 4.793 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=210.176000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=597.909091, yerr=6.960336 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6ad6eaa91393e2e3e5bb87bfd4fb06be6378d79a_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 839.880704E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100942E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834332048 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58363. MP_Allreduce 11058 960. MP_Sync 87 MP_Alltoall 1969 5804166. MP_SendRecv 12032 47072. MP_ISendRecv 12032 47072. MP_Wait 25916 MP_ISend 11748 212467. MP_IRecv 11748 212467. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.034 0.056 196.534 196.536 qs_mol_dyn_low 1 2.0 0.003 0.005 196.105 196.118 qs_forces 11 3.9 0.004 0.004 196.013 196.014 qs_energies 11 4.9 0.002 0.006 189.068 189.080 scf_env_do_scf 11 5.9 0.001 0.001 172.426 172.436 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 138.259 138.261 velocity_verlet 10 3.0 0.001 0.002 123.389 123.391 dbcsr_multiply_generic 2507 12.6 0.190 0.195 98.215 99.330 qs_scf_new_mos 117 7.6 0.001 0.001 97.317 97.766 qs_scf_loop_do_ot 117 8.6 0.001 0.002 97.317 97.765 ot_scf_mini 117 9.6 0.004 0.005 92.513 93.024 multiply_cannon 2507 13.6 0.479 0.526 77.850 82.097 multiply_cannon_loop 2507 14.6 1.253 1.294 74.380 76.763 ot_mini 117 10.6 0.001 0.001 50.233 50.675 mp_waitall_1 214728 16.6 24.640 37.852 24.640 37.852 multiply_cannon_multrec 30084 15.6 22.200 25.962 31.832 35.829 rebuild_ks_matrix 128 8.3 0.001 0.001 33.969 34.419 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.019 33.969 34.419 init_scf_loop 11 6.9 0.001 0.006 34.075 34.076 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.511 30.917 prepare_preconditioner 11 7.9 0.000 0.000 29.610 29.679 make_preconditioner 11 8.9 0.000 0.001 29.609 29.679 make_full_inverse_cholesky 11 9.9 0.000 0.001 28.282 28.823 qs_ot_get_derivative 117 11.6 0.001 0.001 28.270 28.768 multiply_cannon_metrocomm3 30084 15.6 0.090 0.096 15.520 28.291 qs_ot_get_p 128 10.4 0.001 0.001 23.795 24.341 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.029 23.162 apply_single 128 13.6 0.001 0.001 22.029 23.162 multiply_cannon_sync_h2d 30084 15.6 19.518 22.214 19.518 22.214 ot_diis_step 117 11.6 0.014 0.015 21.794 21.796 qs_ot_p2m_diag 83 11.4 0.188 0.216 18.842 18.880 cp_dbcsr_syevd 83 12.4 0.006 0.006 17.657 17.658 cp_fm_cholesky_invert 11 10.9 17.347 17.360 17.347 17.360 make_m2s 5014 13.6 0.093 0.099 14.530 16.165 make_images 5014 14.6 1.163 1.349 14.317 15.952 sum_up_and_integrate 128 10.3 0.116 0.133 15.089 15.118 integrate_v_rspace 128 11.3 0.004 0.004 14.972 15.009 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.489 14.520 calculate_rho_elec 128 8.7 0.089 0.106 14.489 14.519 cp_fm_diag_elpa 83 13.4 0.000 0.001 14.252 14.281 cp_fm_diag_elpa_base 83 14.4 13.979 14.091 14.245 14.270 init_scf_run 11 5.9 0.000 0.001 11.797 11.799 scf_env_initial_rho_setup 11 6.9 0.009 0.023 11.797 11.799 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 11.149 11.505 make_images_data 5014 15.6 0.065 0.074 8.708 10.554 density_rs2pw 128 9.7 0.007 0.007 7.796 10.252 multiply_cannon_metrocomm4 27577 15.6 0.105 0.119 3.679 10.233 hybrid_alltoall_any 5200 16.5 0.344 1.527 7.465 10.031 mp_irecv_dv 69486 16.3 3.478 9.843 3.478 9.843 dbcsr_mm_accdrv_process 62242 16.2 4.334 5.348 9.085 9.699 pw_transfer 1547 11.6 0.087 0.095 8.908 8.960 rs_pw_transfer 1046 11.9 0.014 0.016 6.325 8.814 fft_wrap_pw1pw2 1291 12.7 0.011 0.012 8.681 8.735 wfi_extrapolate 11 7.9 0.001 0.001 8.587 8.588 fft_wrap_pw1pw2_140 523 13.2 1.336 1.355 7.529 7.594 grid_integrate_task_list 128 12.3 7.186 7.576 7.186 7.576 cp_fm_cholesky_decompose 22 10.9 7.355 7.432 7.355 7.432 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.254 6.994 calculate_dm_sparse 128 9.5 0.001 0.001 6.475 6.634 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.235 6.244 fft3d_ps 1291 14.7 2.813 2.976 5.986 6.031 grid_collocate_task_list 128 9.7 4.786 5.971 4.786 5.971 mp_sum_l 7950 12.9 4.063 5.857 4.063 5.857 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.352 5.464 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.378 5.441 potential_pw2rs 128 12.3 0.016 0.019 5.322 5.342 mp_alltoall_d11v 2415 14.1 4.222 5.026 4.222 5.026 mp_allgather_i34 2507 14.6 2.041 5.012 2.041 5.012 mp_waitany 11748 13.9 2.555 5.006 2.555 5.006 rs_pw_transfer_RS2PW_140 139 11.5 0.351 0.377 2.132 4.594 mp_sum_d 4465 12.1 2.714 4.131 2.714 4.131 dbcsr_complete_redistribute 395 12.7 0.776 0.853 3.248 4.125 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=196.536000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=799.090909, yerr=4.481366 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6ad6eaa91393e2e3e5bb87bfd4fb06be6378d79a_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 2.928533E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 3984192 0.0% 0.0% 100.0% average stack size 0.0 0.0 1695.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 951.173120E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1042912 MPI messages size (bytes): total size 2.716210E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.604448E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 281856 36943429632 131072 < size <= 4194304 660064 996105256960 4194304 < size <= 16777216 65632 931531265168 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4020 57958. MP_Allreduce 11127 998. MP_Sync 87 MP_Alltoall 1712 9388896. MP_SendRecv 7936 75008. MP_ISendRecv 7936 75008. MP_Wait 21820 MP_ISend 11748 275205. MP_IRecv 11748 275205. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.032 177.868 177.869 qs_mol_dyn_low 1 2.0 0.003 0.003 177.467 177.479 qs_forces 11 3.9 0.004 0.004 177.356 177.360 qs_energies 11 4.9 0.001 0.001 170.711 170.721 scf_env_do_scf 11 5.9 0.001 0.001 155.190 155.191 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 119.605 119.606 velocity_verlet 10 3.0 0.003 0.010 114.031 114.034 dbcsr_multiply_generic 2507 12.6 0.187 0.192 81.328 82.394 qs_scf_new_mos 117 7.6 0.001 0.001 81.215 81.610 qs_scf_loop_do_ot 117 8.6 0.001 0.001 81.214 81.609 ot_scf_mini 117 9.6 0.003 0.004 77.093 77.475 multiply_cannon 2507 13.6 0.501 0.524 61.617 65.676 multiply_cannon_loop 2507 14.6 0.864 0.886 58.429 61.080 ot_mini 117 10.6 0.001 0.001 42.461 42.851 init_scf_loop 11 6.9 0.000 0.000 35.482 35.483 mp_waitall_1 170520 16.6 24.631 33.438 24.631 33.438 prepare_preconditioner 11 7.9 0.000 0.000 31.422 31.472 make_preconditioner 11 8.9 0.000 0.000 31.422 31.472 rebuild_ks_matrix 128 8.3 0.001 0.001 30.832 31.276 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.022 30.832 31.275 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.040 30.489 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.737 28.136 multiply_cannon_multrec 20056 15.6 13.586 16.485 22.235 25.226 multiply_cannon_metrocomm3 20056 15.6 0.058 0.062 14.919 23.841 qs_ot_get_derivative 117 11.6 0.001 0.002 22.756 23.128 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.815 20.790 apply_single 128 13.6 0.001 0.001 19.815 20.790 qs_ot_get_p 128 10.4 0.001 0.001 20.051 20.446 ot_diis_step 117 11.6 0.018 0.018 19.602 19.602 multiply_cannon_sync_h2d 20056 15.6 14.307 15.789 14.307 15.789 qs_ot_p2m_diag 83 11.4 0.266 0.272 15.752 15.758 make_m2s 5014 13.6 0.082 0.087 14.685 15.685 make_images 5014 14.6 1.178 1.273 14.450 15.448 cp_dbcsr_syevd 83 12.4 0.005 0.006 14.721 14.722 cp_fm_cholesky_invert 11 10.9 14.707 14.717 14.707 14.717 sum_up_and_integrate 128 10.3 0.131 0.145 14.686 14.715 integrate_v_rspace 128 11.3 0.004 0.005 14.554 14.586 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.473 14.495 calculate_rho_elec 128 8.7 0.132 0.148 14.472 14.494 cp_fm_diag_elpa 83 13.4 0.000 0.000 11.533 11.544 cp_fm_diag_elpa_base 83 14.4 11.110 11.270 11.529 11.540 init_scf_run 11 5.9 0.000 0.001 10.524 10.524 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.524 10.524 make_images_data 5014 15.6 0.059 0.066 8.973 10.338 hybrid_alltoall_any 5200 16.5 0.435 1.985 7.794 9.467 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 8.962 9.237 density_rs2pw 128 9.7 0.006 0.007 7.265 9.222 multiply_cannon_metrocomm4 17549 15.6 0.062 0.072 3.434 9.035 pw_transfer 1547 11.6 0.086 0.103 8.738 8.836 mp_irecv_dv 50230 16.2 3.312 8.792 3.312 8.792 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.513 8.615 dbcsr_mm_accdrv_process 41502 16.2 4.390 5.151 8.095 8.182 grid_integrate_task_list 128 12.3 7.274 7.696 7.274 7.696 fft_wrap_pw1pw2_140 523 13.2 1.412 1.442 7.479 7.588 cp_fm_cholesky_decompose 22 10.9 7.536 7.549 7.536 7.549 wfi_extrapolate 11 7.9 0.001 0.001 7.532 7.532 cp_fm_upper_to_full 105 14.8 5.797 7.376 5.797 7.376 rs_pw_transfer 1046 11.9 0.014 0.014 5.382 7.355 dbcsr_complete_redistribute 395 12.7 1.168 1.203 4.498 6.363 calculate_dm_sparse 128 9.5 0.001 0.001 5.719 5.842 grid_collocate_task_list 128 9.7 4.992 5.841 4.992 5.841 fft3d_ps 1291 14.7 2.740 2.964 5.630 5.716 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.411 5.417 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.552 5.237 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.327 5.186 mp_alltoall_d11v 2415 14.1 4.200 4.977 4.200 4.977 potential_pw2rs 128 12.3 0.021 0.024 4.779 4.790 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.532 4.683 mp_allgather_i34 2507 14.6 1.710 4.433 1.710 4.433 mp_sum_l 7950 12.9 3.187 4.429 3.187 4.429 mp_waitany 11748 13.9 2.310 4.322 2.310 4.322 transfer_fm_to_dbcsr 11 9.9 0.020 0.028 2.362 4.155 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.011 4.040 mp_alltoall_i22 716 14.1 2.001 3.964 2.001 3.964 rs_pw_transfer_RS2PW_140 139 11.5 0.332 0.351 1.899 3.851 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 3.724 3.726 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.596 3.644 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=177.869000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=896.545455, yerr=11.388700 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6ad6eaa91393e2e3e5bb87bfd4fb06be6378d79a_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019187724288 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019187724288 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796577E+12 0.0% 0.0% 100.0% flops max/rank 4.320341E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500544 0.0% 0.0% 100.0% number of processed stacks 5927808 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.2 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.154802E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1133160 MPI messages size (bytes): total size 2.008142E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.772161E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 315952 35695099904 131072 < size <= 4194304 709496 778939400192 4194304 < size <= 16777216 69840 660837747920 16777216 < size 30480 532676608000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3979 58540. MP_Allreduce 11025 1085. MP_Sync 86 MP_Alltoall 1700 12496361. MP_SendRecv 5842 75008. MP_ISendRecv 5842 75008. MP_Wait 22272 MP_ISend 14840 244848. MP_IRecv 14840 244848. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.027 0.066 203.921 203.923 qs_mol_dyn_low 1 2.0 0.003 0.003 203.406 203.419 qs_forces 11 3.9 0.011 0.064 203.284 203.294 qs_energies 11 4.9 0.006 0.043 195.630 195.640 scf_env_do_scf 11 5.9 0.001 0.001 177.745 177.751 velocity_verlet 10 3.0 0.003 0.014 133.626 133.629 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 126.443 126.446 qs_scf_new_mos 116 7.6 0.001 0.001 87.155 87.447 qs_scf_loop_do_ot 116 8.6 0.001 0.001 87.154 87.447 ot_scf_mini 116 9.6 0.003 0.004 82.623 82.945 dbcsr_multiply_generic 2485 12.5 0.185 0.191 80.664 81.330 multiply_cannon 2485 13.5 0.547 0.584 54.681 57.386 multiply_cannon_loop 2485 14.5 1.181 1.211 50.695 52.294 init_scf_loop 11 6.9 0.002 0.017 51.164 51.167 prepare_preconditioner 11 7.9 0.000 0.001 46.916 46.941 make_preconditioner 11 8.9 0.000 0.001 46.916 46.941 make_full_inverse_cholesky 11 9.9 0.012 0.023 40.119 45.304 ot_mini 116 10.6 0.001 0.001 43.458 43.758 rebuild_ks_matrix 127 8.3 0.001 0.001 31.065 31.316 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.021 31.065 31.316 multiply_cannon_multrec 29820 15.5 14.063 18.712 25.965 30.629 mp_waitall_1 146592 16.7 18.439 28.431 18.439 28.431 qs_ks_update_qs_env 127 7.6 0.001 0.001 27.742 27.962 qs_ot_get_p 127 10.4 0.001 0.001 25.269 25.564 qs_ot_get_derivative 116 11.6 0.001 0.002 23.907 24.233 make_m2s 4970 13.5 0.097 0.101 21.472 22.566 make_images 4970 14.5 1.940 2.272 21.167 22.261 qs_ot_p2m_diag 82 11.4 0.339 0.389 20.761 20.813 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 18.967 19.541 apply_single 127 13.6 0.001 0.001 18.967 19.541 ot_diis_step 116 11.6 0.017 0.019 19.419 19.421 cp_dbcsr_syevd 82 12.4 0.005 0.006 19.401 19.403 cp_fm_cholesky_invert 11 10.9 18.722 18.732 18.722 18.732 cp_fm_upper_to_full 104 14.8 11.754 17.396 11.754 17.396 sum_up_and_integrate 127 10.3 0.139 0.151 15.785 15.840 cp_fm_diag_elpa 82 13.4 0.000 0.000 15.807 15.821 qs_rho_update_rho_low 127 7.7 0.001 0.001 15.781 15.813 calculate_rho_elec 127 8.7 0.175 0.191 15.781 15.812 cp_fm_diag_elpa_base 82 14.4 14.766 15.109 15.799 15.812 integrate_v_rspace 127 11.3 0.004 0.004 15.646 15.706 multiply_cannon_metrocomm3 29820 15.5 0.047 0.050 6.136 14.974 dbcsr_complete_redistribute 393 12.7 1.511 1.661 9.673 13.579 make_images_data 4970 15.5 0.061 0.067 11.555 13.455 multiply_cannon_sync_h2d 29820 15.5 11.645 12.723 11.645 12.723 copy_fm_to_dbcsr 208 11.6 0.002 0.002 8.142 12.067 hybrid_alltoall_any 5155 16.4 0.525 2.194 10.091 12.014 dbcsr_mm_accdrv_process 61748 16.2 7.347 8.302 11.478 11.984 init_scf_run 11 5.9 0.000 0.001 11.601 11.604 scf_env_initial_rho_setup 11 6.9 0.019 0.021 11.600 11.604 transfer_fm_to_dbcsr 11 9.9 0.001 0.006 6.774 10.604 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 10.112 10.354 pw_transfer 1535 11.6 0.086 0.101 9.786 9.934 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 9.560 9.713 mp_alltoall_i22 712 14.1 5.767 9.667 5.767 9.667 density_rs2pw 127 9.7 0.006 0.006 8.029 9.275 cp_fm_cholesky_decompose 22 10.9 9.037 9.125 9.037 9.125 fft_wrap_pw1pw2_140 519 13.2 1.545 1.579 8.265 8.434 wfi_extrapolate 11 7.9 0.001 0.001 8.308 8.308 grid_integrate_task_list 127 12.3 7.431 7.947 7.431 7.947 rs_pw_transfer 1038 11.9 0.013 0.014 6.305 7.547 multiply_cannon_metrocomm4 24850 15.5 0.079 0.090 2.768 6.970 mp_irecv_dv 75445 16.2 2.624 6.707 2.624 6.707 mp_alltoall_d11v 2401 14.1 5.697 6.706 5.697 6.706 fft3d_ps 1281 14.7 2.816 2.894 6.509 6.622 calculate_dm_sparse 127 9.5 0.001 0.001 6.290 6.373 grid_collocate_task_list 127 9.7 5.116 5.898 5.116 5.898 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.689 5.781 potential_pw2rs 127 12.3 0.023 0.023 5.588 5.668 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.569 4.660 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 4.604 4.607 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.501 4.589 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.328 4.390 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=203.923000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1088.454545, yerr=14.779369 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6ad6eaa91393e2e3e5bb87bfd4fb06be6378d79a_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420241154048 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528896499712 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514757E+12 0.0% 0.0% 100.0% flops max/rank 5.865088E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755939872 0.0% 0.0% 100.0% number of processed stacks 1960712 0.0% 0.0% 100.0% average stack size 0.0 0.0 3445.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.507348E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 240672 MPI messages size (bytes): total size 1.331455E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.532238E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 113904 59718500352 4194304 < size <= 16777216 104976 550376570880 16777216 < size 20208 721350154512 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8931 51. MP_Alltoall 9654 799394. MP_ISend 40068 2102573. MP_IRecv 40068 2101675. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58352. MP_Allreduce 11057 1167. MP_Sync 87 MP_Alltoall 1712 18838210. MP_SendRecv 3840 122880. MP_ISendRecv 3840 122880. MP_Wait 16122 MP_ISend 10680 423556. MP_IRecv 10680 423556. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.022 0.060 195.803 195.804 qs_mol_dyn_low 1 2.0 0.003 0.003 195.323 195.337 qs_forces 11 3.9 0.004 0.004 195.211 195.215 qs_energies 11 4.9 0.001 0.001 187.339 187.345 scf_env_do_scf 11 5.9 0.001 0.001 168.939 168.946 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 127.724 127.726 velocity_verlet 10 3.0 0.001 0.001 127.064 127.068 qs_scf_new_mos 117 7.6 0.001 0.001 86.560 86.650 qs_scf_loop_do_ot 117 8.6 0.001 0.001 86.559 86.649 ot_scf_mini 117 9.6 0.003 0.004 81.966 82.032 dbcsr_multiply_generic 2507 12.6 0.181 0.193 76.963 77.415 multiply_cannon 2507 13.6 0.578 0.603 55.326 60.364 multiply_cannon_loop 2507 14.6 0.446 0.459 50.483 51.350 ot_mini 117 10.6 0.001 0.001 41.827 41.889 init_scf_loop 11 6.9 0.000 0.000 41.046 41.047 prepare_preconditioner 11 7.9 0.000 0.000 36.940 36.961 make_preconditioner 11 8.9 0.000 0.000 36.940 36.961 mp_waitall_1 125778 16.7 28.029 35.212 28.029 35.212 make_full_inverse_cholesky 11 9.9 0.015 0.027 34.700 35.021 rebuild_ks_matrix 128 8.3 0.001 0.001 31.386 31.462 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.018 31.385 31.461 qs_ks_update_qs_env 128 7.6 0.001 0.001 28.414 28.484 qs_ot_get_p 128 10.4 0.001 0.001 26.985 27.076 qs_ot_p2m_diag 83 11.4 0.496 0.501 22.568 22.594 qs_ot_get_derivative 117 11.6 0.001 0.002 21.745 21.804 cp_dbcsr_syevd 83 12.4 0.005 0.006 21.245 21.249 cp_fm_cholesky_invert 11 10.9 21.166 21.173 21.166 21.173 multiply_cannon_multrec 10028 15.6 10.369 14.230 17.707 20.524 ot_diis_step 117 11.6 0.020 0.020 20.015 20.016 make_m2s 5014 13.6 0.067 0.073 17.618 19.960 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.570 19.903 apply_single 128 13.6 0.001 0.001 19.570 19.902 make_images 5014 14.6 2.298 2.796 17.310 19.646 multiply_cannon_metrocomm3 10028 15.6 0.023 0.024 11.752 19.033 cp_fm_diag_elpa 83 13.4 0.000 0.000 17.386 17.395 cp_fm_diag_elpa_base 83 14.4 17.083 17.213 17.380 17.390 qs_rho_update_rho_low 128 7.7 0.001 0.001 16.873 16.897 calculate_rho_elec 128 8.7 0.259 0.269 16.873 16.896 sum_up_and_integrate 128 10.3 0.181 0.192 16.239 16.290 integrate_v_rspace 128 11.3 0.004 0.004 16.057 16.111 make_images_data 5014 15.6 0.051 0.061 10.703 13.587 hybrid_alltoall_any 5200 16.5 0.820 3.586 10.287 12.648 multiply_cannon_sync_h2d 10028 15.6 11.599 11.967 11.599 11.967 init_scf_run 11 5.9 0.000 0.001 11.343 11.343 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.342 11.343 pw_transfer 1547 11.6 0.087 0.095 10.533 10.573 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 10.308 10.356 cp_fm_cholesky_decompose 22 10.9 9.915 10.028 9.915 10.028 density_rs2pw 128 9.7 0.006 0.006 8.101 9.000 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 8.923 8.959 fft_wrap_pw1pw2_140 523 13.2 1.908 1.942 8.827 8.880 multiply_cannon_metrocomm1 10028 15.6 0.029 0.029 5.766 8.705 grid_integrate_task_list 128 12.3 7.714 8.298 7.714 8.298 wfi_extrapolate 11 7.9 0.001 0.001 8.260 8.260 mp_allgather_i34 2507 14.6 3.052 7.726 3.052 7.726 dbcsr_mm_accdrv_process 20762 16.1 2.552 3.433 6.974 7.605 mp_alltoall_d11v 2415 14.1 6.086 7.350 6.086 7.350 fft3d_ps 1291 14.7 2.770 2.841 6.789 6.826 calculate_dm_sparse 128 9.5 0.001 0.001 6.300 6.383 rs_pw_transfer 1046 11.9 0.013 0.013 5.432 6.379 dbcsr_complete_redistribute 395 12.7 2.110 2.161 5.749 6.194 grid_collocate_task_list 128 9.7 5.467 6.162 5.467 6.162 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.385 5.429 potential_pw2rs 128 12.3 0.027 0.028 5.388 5.407 qs_energies_init_hamiltonians 11 5.9 0.003 0.007 5.343 5.344 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.176 4.197 copy_fm_to_dbcsr 209 11.7 0.001 0.002 3.711 4.084 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.868 3.941 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=195.804000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1398.454545, yerr=37.266518 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6ad6eaa91393e2e3e5bb87bfd4fb06be6378d79a_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430454546432 0.0% 0.0% 100.0% flops 32 x 32 x 32 1973537472512 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 11613065416704 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 94.243750E+12 0.0% 0.0% 100.0% flops max/rank 11.787404E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806546176 0.0% 0.0% 100.0% number of processed stacks 1982128 0.0% 0.0% 100.0% average stack size 0.0 0.0 3434.0 marketing flops 145.661668E+12 ------------------------------------------------------------------------------- # multiplications 2534 max memory usage/rank 3.127087E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101360 MPI messages size (bytes): total size 1.145137E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.297725E+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 45848 35601252352 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592388592 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 58564. MP_Allreduce 11184 1500. MP_Sync 88 MP_Alltoall 1724 36993632. MP_SendRecv 1806 218624. MP_ISendRecv 1806 218624. MP_Wait 9876 MP_ISend 6456 1080169. MP_IRecv 6456 1080169. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.025 0.042 326.416 326.417 qs_mol_dyn_low 1 2.0 0.003 0.003 325.831 325.844 qs_forces 11 3.9 0.004 0.004 325.729 325.734 qs_energies 11 4.9 0.001 0.001 316.005 316.025 scf_env_do_scf 11 5.9 0.001 0.001 292.272 292.288 velocity_verlet 10 3.0 0.001 0.001 234.310 234.317 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 157.061 157.064 init_scf_loop 11 6.9 0.000 0.000 134.933 134.938 prepare_preconditioner 11 7.9 0.000 0.000 129.637 129.658 make_preconditioner 11 8.9 0.000 0.000 129.637 129.658 make_full_inverse_cholesky 11 9.9 0.036 0.039 104.945 126.637 qs_scf_new_mos 118 7.6 0.001 0.001 102.591 102.694 qs_scf_loop_do_ot 118 8.6 0.001 0.001 102.590 102.693 ot_scf_mini 118 9.6 0.004 0.004 97.628 97.779 dbcsr_multiply_generic 2534 12.6 0.215 0.226 88.576 89.021 cp_fm_upper_to_full 106 14.8 54.058 77.644 54.058 77.644 multiply_cannon 2534 13.6 0.718 0.777 61.730 63.022 multiply_cannon_loop 2534 14.6 0.480 0.490 57.677 59.049 ot_mini 118 10.6 0.001 0.001 48.268 48.474 dbcsr_complete_redistribute 397 12.7 3.971 4.026 30.908 44.098 rebuild_ks_matrix 129 8.3 0.001 0.001 40.586 40.708 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.018 40.586 40.707 copy_fm_to_dbcsr 210 11.7 0.001 0.002 27.219 40.374 cp_fm_cholesky_invert 11 10.9 39.507 39.515 39.507 39.515 transfer_fm_to_dbcsr 11 9.9 0.030 0.033 24.646 37.639 qs_ks_update_qs_env 129 7.6 0.001 0.001 37.336 37.451 mp_waitall_1 104780 16.8 31.947 36.101 31.947 36.101 mp_alltoall_i22 720 14.1 22.433 35.744 22.433 35.744 qs_ot_get_p 129 10.4 0.001 0.001 33.014 33.249 qs_ot_p2m_diag 84 11.4 0.889 0.895 28.306 28.362 qs_ot_get_derivative 118 11.6 0.002 0.002 27.290 27.446 cp_dbcsr_syevd 84 12.4 0.006 0.006 26.343 26.347 make_m2s 5068 13.6 0.077 0.079 21.388 22.538 qs_rho_update_rho_low 129 7.7 0.001 0.001 22.441 22.485 calculate_rho_elec 129 8.7 0.485 0.485 22.440 22.484 cp_fm_diag_elpa 84 13.4 0.000 0.000 22.473 22.474 cp_fm_diag_elpa_base 84 14.4 17.876 19.555 22.468 22.469 multiply_cannon_metrocomm3 10136 15.6 0.023 0.025 21.061 22.250 make_images 5068 14.6 3.785 4.005 20.903 22.054 sum_up_and_integrate 129 10.3 0.324 0.326 21.180 21.262 integrate_v_rspace 129 11.3 0.004 0.006 20.855 20.937 ot_diis_step 118 11.6 0.023 0.023 20.908 20.908 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 20.119 20.297 apply_single 129 13.6 0.001 0.001 20.119 20.297 multiply_cannon_multrec 10136 15.6 10.619 12.409 18.229 18.293 multiply_cannon_sync_h2d 10136 15.6 15.791 15.810 15.791 15.810 make_images_data 5068 15.6 0.058 0.063 11.724 13.631 hybrid_alltoall_any 5255 16.5 1.310 3.053 11.537 13.518 init_scf_run 11 5.9 0.000 0.001 13.427 13.427 scf_env_initial_rho_setup 11 6.9 0.001 0.001 13.427 13.427 pw_transfer 1559 11.6 0.093 0.094 13.138 13.161 fft_wrap_pw1pw2 1301 12.7 0.011 0.011 12.901 12.924 cp_fm_cholesky_decompose 22 10.9 11.303 11.359 11.303 11.359 fft_wrap_pw1pw2_140 527 13.2 3.094 3.125 11.232 11.256 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 10.837 10.945 mp_alltoall_d11v 2429 14.1 9.725 10.849 9.725 10.849 wfi_extrapolate 11 7.9 0.001 0.001 9.958 9.958 density_rs2pw 129 9.7 0.005 0.005 9.459 9.608 dbcsr_mm_accdrv_process 20954 16.1 3.770 5.452 7.366 9.253 grid_integrate_task_list 129 12.3 8.665 8.834 8.665 8.834 qs_energies_init_hamiltonians 11 5.9 0.002 0.005 8.073 8.076 fft3d_ps 1301 14.7 2.871 2.884 7.944 7.978 calculate_dm_sparse 129 9.5 0.001 0.001 6.888 7.024 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.711 6.832 rs_scatter_matrices 140 9.7 3.747 4.644 6.487 6.705 grid_collocate_task_list 129 9.7 6.493 6.553 6.493 6.553 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=326.417000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2772.545455, yerr=180.689771 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6ad6eaa91393e2e3e5bb87bfd4fb06be6378d79a_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 1.259500E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255669. MP_Allreduce 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 0.016 0.035 85.389 85.390 qs_energies 1 2.0 0.000 0.000 84.890 84.904 ls_scf 1 3.0 0.000 0.000 84.012 84.026 dbcsr_multiply_generic 111 6.7 0.016 0.016 72.681 72.866 multiply_cannon 111 7.7 0.019 0.027 55.956 56.877 multiply_cannon_loop 111 8.7 0.212 0.227 52.483 53.539 ls_scf_main 1 4.0 0.000 0.000 52.599 52.599 density_matrix_trs4 2 5.0 0.002 0.003 46.994 47.111 ls_scf_init_scf 1 4.0 0.000 0.000 28.351 28.352 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.203 27.255 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.094 25.114 mp_waitall_1 11031 10.9 22.022 24.840 22.022 24.840 multiply_cannon_multrec 2664 9.7 8.233 9.106 15.628 17.499 multiply_cannon_sync_h2d 2664 9.7 13.862 15.589 13.862 15.589 make_m2s 222 7.7 0.009 0.011 13.006 13.639 make_images 222 8.7 0.101 0.111 12.984 13.617 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.351 12.757 multiply_cannon_metrocomm3 2664 9.7 0.010 0.011 5.387 8.832 make_images_data 222 9.7 0.005 0.007 7.606 8.180 dbcsr_mm_accdrv_process 4760 10.4 0.503 0.631 7.013 7.974 hybrid_alltoall_any 227 10.6 0.218 1.849 6.523 7.830 dbcsr_mm_accdrv_process_sort 4760 11.4 6.311 7.221 6.311 7.221 calculate_norms 4752 9.8 5.579 6.293 5.579 6.293 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.988 5.194 mp_sum_l 887 5.1 3.180 4.865 3.180 4.865 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.067 3.717 mp_irecv_dv 6231 10.9 2.050 3.684 2.050 3.684 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.334 3.490 arnoldi_extremal 4 6.8 0.000 0.001 3.438 3.470 arnoldi_normal_ev 4 7.8 0.001 0.003 3.438 3.470 build_subspace 16 8.4 0.009 0.012 3.337 3.339 make_images_sizes 222 9.7 0.000 0.000 0.735 3.170 mp_alltoall_i44 222 10.7 0.735 3.170 0.735 3.170 ls_scf_post 1 4.0 0.000 0.000 3.063 3.076 ls_scf_store_result 1 5.0 0.000 0.000 2.877 2.924 dbcsr_special_finalize 555 9.7 0.005 0.006 2.306 2.723 dbcsr_merge_single_wm 555 10.7 0.457 0.591 2.298 2.715 make_images_pack 222 9.7 2.209 2.638 2.211 2.640 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.387 2.619 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.403 2.520 dbcsr_matrix_vector_mult_local 304 10.0 2.052 2.445 2.054 2.447 dbcsr_sort_data 658 11.4 2.098 2.445 2.098 2.445 buffer_matrices_ensure_size 222 8.7 1.755 2.067 1.755 2.067 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.798 1.800 rebuild_ks_matrix 3 7.3 0.000 0.000 1.789 1.790 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.789 1.790 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.390000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1131.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6ad6eaa91393e2e3e5bb87bfd4fb06be6378d79a_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 2.098913E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 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 0.029 0.045 90.651 90.652 qs_energies 1 2.0 0.000 0.000 90.177 90.181 ls_scf 1 3.0 0.000 0.000 88.861 88.865 dbcsr_multiply_generic 111 6.7 0.015 0.016 74.487 74.846 multiply_cannon 111 7.7 0.028 0.038 52.995 56.788 ls_scf_main 1 4.0 0.000 0.000 54.390 54.399 multiply_cannon_loop 111 8.7 0.118 0.126 49.720 52.988 density_matrix_trs4 2 5.0 0.002 0.003 48.602 48.828 ls_scf_init_scf 1 4.0 0.000 0.000 30.963 30.964 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.731 29.828 mp_waitall_1 9105 10.9 19.973 28.850 19.973 28.850 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.331 27.348 multiply_cannon_multrec 1332 9.7 13.249 16.099 22.548 26.414 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.196 20.180 make_m2s 222 7.7 0.008 0.009 14.735 15.335 make_images 222 8.7 1.580 1.973 14.704 15.305 dbcsr_mm_accdrv_process 4041 10.4 0.279 0.449 8.907 10.470 dbcsr_mm_accdrv_process_sort 4041 11.4 8.501 10.020 8.501 10.020 make_images_data 222 9.7 0.004 0.005 8.158 9.091 hybrid_alltoall_any 227 10.6 0.526 2.485 7.584 8.530 mp_sum_l 887 5.1 5.447 8.527 5.447 8.527 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.219 7.770 mp_irecv_dv 3311 11.0 3.200 7.699 3.200 7.699 calculate_norms 2376 9.8 6.047 6.749 6.047 6.749 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.120 6.680 multiply_cannon_sync_h2d 1332 9.7 4.870 6.335 4.870 6.335 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.055 5.270 arnoldi_extremal 4 6.8 0.000 0.002 5.002 5.021 arnoldi_normal_ev 4 7.8 0.001 0.004 5.002 5.021 build_subspace 16 8.4 0.014 0.021 4.743 4.746 ls_scf_post 1 4.0 0.000 0.000 3.508 3.512 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.274 3.478 ls_scf_store_result 1 5.0 0.000 0.000 3.218 3.327 dbcsr_matrix_vector_mult_local 304 10.0 2.736 3.220 2.738 3.221 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.250 2.812 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.636 2.745 make_images_pack 222 9.7 2.031 2.435 2.034 2.437 mp_allgather_i34 111 8.7 0.961 2.108 0.961 2.108 dbcsr_sort_data 436 11.2 1.842 2.038 1.842 2.038 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.892 1.894 rebuild_ks_matrix 3 7.3 0.000 0.000 1.880 1.881 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.880 1.881 buffer_matrices_ensure_size 222 8.7 1.013 1.860 1.013 1.860 dbcsr_data_new 4174 10.1 1.622 1.859 1.622 1.859 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=90.652000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1729.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6ad6eaa91393e2e3e5bb87bfd4fb06be6378d79a_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 2.786951E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265470. MP_Allreduce 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 0.033 0.054 93.831 93.832 qs_energies 1 2.0 0.000 0.000 93.108 93.111 ls_scf 1 3.0 0.000 0.000 91.706 91.711 dbcsr_multiply_generic 111 6.7 0.016 0.016 75.862 76.159 ls_scf_main 1 4.0 0.000 0.000 57.290 57.295 multiply_cannon 111 7.7 0.037 0.071 52.605 56.583 multiply_cannon_loop 111 8.7 0.101 0.109 48.985 52.871 density_matrix_trs4 2 5.0 0.002 0.003 51.233 51.383 mp_waitall_1 7281 11.0 22.790 32.104 22.790 32.104 ls_scf_init_scf 1 4.0 0.000 0.000 30.872 30.876 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.668 29.736 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.266 27.275 multiply_cannon_multrec 888 9.7 12.665 15.445 21.275 24.687 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.693 21.896 make_m2s 222 7.7 0.007 0.008 16.620 17.892 make_images 222 8.7 1.985 2.315 16.581 17.851 hybrid_alltoall_any 227 10.6 0.624 2.891 8.854 10.243 make_images_data 222 9.7 0.004 0.005 9.193 10.201 dbcsr_mm_accdrv_process 3754 10.4 0.246 0.418 8.140 9.392 dbcsr_mm_accdrv_process_sort 3754 11.4 7.770 8.974 7.770 8.974 mp_sum_l 887 5.1 5.108 8.175 5.108 8.175 multiply_cannon_sync_h2d 888 9.7 6.056 7.214 6.056 7.214 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.475 6.921 mp_irecv_dv 2335 11.1 2.460 6.870 2.460 6.870 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.897 6.389 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.518 6.367 arnoldi_extremal 4 6.8 0.000 0.001 5.507 5.521 arnoldi_normal_ev 4 7.8 0.001 0.005 5.507 5.521 build_subspace 16 8.4 0.014 0.020 5.201 5.209 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.892 5.079 calculate_norms 1584 9.8 4.335 4.705 4.335 4.705 dbcsr_matrix_vector_mult 304 9.0 0.009 0.021 3.596 3.964 mp_allgather_i34 111 8.7 1.448 3.643 1.448 3.643 dbcsr_matrix_vector_mult_local 304 10.0 3.015 3.606 3.017 3.608 ls_scf_post 1 4.0 0.000 0.000 3.544 3.547 ls_scf_store_result 1 5.0 0.000 0.000 3.292 3.371 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.943 3.050 dbcsr_sort_data 325 11.1 1.843 2.152 1.843 2.152 make_images_pack 222 9.7 1.822 2.117 1.825 2.120 make_images_sizes 222 9.7 0.000 0.000 1.115 2.075 mp_alltoall_i44 222 10.7 1.114 2.074 1.114 2.074 dbcsr_data_release 9322 10.9 1.306 2.042 1.306 2.042 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.934 1.936 rebuild_ks_matrix 3 7.3 0.000 0.000 1.916 1.918 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.916 1.918 dbcsr_finalize 304 7.8 0.027 0.032 1.612 1.885 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=93.832000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2160.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6ad6eaa91393e2e3e5bb87bfd4fb06be6378d79a_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 3.324686E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 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 0.034 0.046 96.645 96.646 qs_energies 1 2.0 0.000 0.000 96.023 96.029 ls_scf 1 3.0 0.000 0.000 94.369 94.375 dbcsr_multiply_generic 111 6.7 0.017 0.018 77.683 77.971 ls_scf_main 1 4.0 0.000 0.000 58.510 58.512 multiply_cannon 111 7.7 0.048 0.101 51.132 55.900 density_matrix_trs4 2 5.0 0.002 0.003 52.371 52.500 multiply_cannon_loop 111 8.7 0.115 0.127 45.982 49.110 ls_scf_init_scf 1 4.0 0.000 0.000 32.611 32.613 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.384 31.441 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.906 28.922 mp_waitall_1 6369 11.0 21.711 28.102 21.711 28.102 multiply_cannon_multrec 1332 9.7 14.286 17.370 22.256 25.163 make_m2s 222 7.7 0.008 0.009 20.928 22.475 make_images 222 8.7 3.143 3.605 20.877 22.426 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 8.614 16.392 make_images_data 222 9.7 0.004 0.005 11.548 13.212 hybrid_alltoall_any 227 10.6 0.801 3.775 10.805 12.591 dbcsr_mm_accdrv_process 3641 10.4 0.220 0.403 7.596 9.137 dbcsr_mm_accdrv_process_sort 3641 11.4 7.223 8.721 7.223 8.721 mp_sum_l 887 5.1 4.038 7.475 4.038 7.475 multiply_cannon_sync_h2d 1332 9.7 5.492 6.011 5.492 6.011 multiply_cannon_metrocomm4 1110 9.7 0.004 0.007 2.086 5.942 mp_irecv_dv 3229 10.9 2.063 5.903 2.063 5.903 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.076 5.842 arnoldi_extremal 4 6.8 0.000 0.001 5.616 5.631 arnoldi_normal_ev 4 7.8 0.001 0.005 5.616 5.631 build_subspace 16 8.4 0.014 0.021 5.267 5.275 mp_allgather_i34 111 8.7 2.264 4.750 2.264 4.750 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.523 4.700 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.387 4.532 calculate_norms 2376 9.8 4.196 4.521 4.196 4.521 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.738 4.087 dbcsr_matrix_vector_mult_local 304 10.0 3.155 3.681 3.157 3.683 dbcsr_sort_data 658 11.4 3.034 3.398 3.034 3.398 ls_scf_post 1 4.0 0.000 0.000 3.248 3.255 dbcsr_special_finalize 555 9.7 0.006 0.007 2.774 3.242 dbcsr_merge_single_wm 555 10.7 0.532 0.663 2.765 3.234 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.042 3.100 ls_scf_store_result 1 5.0 0.000 0.000 2.986 3.047 dbcsr_data_release 10477 10.7 1.600 2.398 1.600 2.398 dbcsr_finalize 304 7.8 0.050 0.062 1.810 1.991 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.951 1.952 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=96.646000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2723.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6ad6eaa91393e2e3e5bb87bfd4fb06be6378d79a_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 4.696691E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265558. MP_Allreduce 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 0.049 0.064 99.737 99.738 qs_energies 1 2.0 0.000 0.000 98.939 98.941 ls_scf 1 3.0 0.000 0.000 96.993 96.998 dbcsr_multiply_generic 111 6.7 0.018 0.018 77.965 78.211 ls_scf_main 1 4.0 0.000 0.000 62.425 62.438 multiply_cannon 111 7.7 0.117 0.190 55.557 60.794 density_matrix_trs4 2 5.0 0.002 0.003 55.124 55.206 multiply_cannon_loop 111 8.7 0.069 0.076 50.941 52.397 mp_waitall_1 5436 11.0 26.310 31.842 26.310 31.842 ls_scf_init_scf 1 4.0 0.000 0.000 30.957 30.964 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.607 29.669 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.453 27.465 multiply_cannon_multrec 444 9.7 14.040 16.638 21.167 23.140 make_m2s 222 7.7 0.006 0.006 17.694 20.280 make_images 222 8.7 3.730 4.436 17.630 20.218 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.916 16.182 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.371 14.404 make_images_data 222 9.7 0.003 0.004 9.819 12.390 hybrid_alltoall_any 227 10.6 0.788 3.756 9.578 12.258 dbcsr_mm_accdrv_process 3003 10.4 0.168 0.370 6.817 7.937 dbcsr_mm_accdrv_process_sort 3003 11.4 6.496 7.600 6.496 7.600 multiply_cannon_sync_h2d 444 9.7 6.580 7.487 6.580 7.487 mp_allgather_i34 111 8.7 2.829 7.078 2.829 7.078 arnoldi_extremal 4 6.8 0.002 0.005 6.092 6.103 arnoldi_normal_ev 4 7.8 0.002 0.005 6.090 6.098 build_subspace 16 8.4 0.015 0.019 5.695 5.703 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.588 4.716 mp_sum_l 887 5.1 2.757 4.565 2.757 4.565 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.693 4.454 mp_irecv_dv 1241 11.2 1.676 4.424 1.676 4.424 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.223 4.419 dbcsr_matrix_vector_mult_local 304 10.0 3.604 4.061 3.606 4.063 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.650 3.760 calculate_norms 792 9.8 3.570 3.684 3.570 3.684 ls_scf_post 1 4.0 0.000 0.000 3.611 3.613 make_images_sizes 222 9.7 0.000 0.000 1.206 3.521 mp_alltoall_i44 222 10.7 1.206 3.521 1.206 3.521 ls_scf_store_result 1 5.0 0.000 0.000 3.364 3.427 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.866 3.371 dbcsr_finalize 304 7.8 0.062 0.078 2.208 2.302 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.145 2.148 dbcsr_merge_all 275 8.9 0.474 0.529 2.054 2.138 rebuild_ks_matrix 3 7.3 0.000 0.000 2.112 2.115 qs_ks_build_kohn_sham_matrix 3 8.3 0.013 0.015 2.112 2.115 dbcsr_data_release 10123 10.8 1.356 2.027 1.356 2.027 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=99.738000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3651.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6ad6eaa91393e2e3e5bb87bfd4fb06be6378d79a_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 8.845849E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284111. MP_Allreduce 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 0.127 0.187 108.236 108.249 qs_energies 1 2.0 0.000 0.000 106.647 106.658 ls_scf 1 3.0 0.000 0.000 103.728 103.738 dbcsr_multiply_generic 111 6.7 0.023 0.026 77.393 77.491 ls_scf_main 1 4.0 0.000 0.000 65.872 65.883 density_matrix_trs4 2 5.0 0.002 0.003 56.643 56.691 multiply_cannon 111 7.7 0.202 0.345 50.125 52.069 multiply_cannon_loop 111 8.7 0.069 0.074 46.695 47.474 ls_scf_init_scf 1 4.0 0.000 0.000 34.165 34.167 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.564 32.584 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.728 29.738 mp_waitall_1 4527 11.1 22.175 25.722 22.175 25.722 make_m2s 222 7.7 0.007 0.007 23.748 24.673 make_images 222 8.7 4.603 4.967 23.638 24.562 multiply_cannon_multrec 444 9.7 17.933 18.621 22.604 23.205 hybrid_alltoall_any 227 10.6 1.666 3.646 12.770 15.318 make_images_data 222 9.7 0.003 0.004 12.959 15.283 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.629 11.131 multiply_cannon_sync_h2d 444 9.7 8.847 8.893 8.847 8.893 arnoldi_extremal 4 6.8 0.001 0.001 7.150 7.160 arnoldi_normal_ev 4 7.8 0.003 0.009 7.150 7.159 build_subspace 16 8.4 0.026 0.036 6.598 6.608 dbcsr_matrix_vector_mult 304 9.0 0.017 0.035 5.299 5.401 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.225 5.316 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.974 5.238 dbcsr_matrix_vector_mult_local 304 10.0 4.872 5.123 4.875 5.125 dbcsr_mm_accdrv_process 1814 10.4 0.297 0.360 4.505 4.633 dbcsr_mm_accdrv_process_sort 1814 11.4 4.184 4.314 4.184 4.314 ls_scf_post 1 4.0 0.000 0.000 3.691 3.702 make_images_sizes 222 9.7 0.000 0.000 1.424 3.466 mp_alltoall_i44 222 10.7 1.423 3.466 1.423 3.466 ls_scf_store_result 1 5.0 0.000 0.000 3.434 3.444 mp_allgather_i34 111 8.7 1.053 3.340 1.053 3.340 calculate_norms 792 9.8 3.240 3.277 3.240 3.277 dbcsr_finalize 304 7.8 0.082 0.090 3.090 3.132 dbcsr_merge_all 275 8.9 0.894 0.917 2.873 2.922 qs_energies_init_hamiltonians 1 3.0 0.001 0.004 2.889 2.889 dbcsr_complete_redistribute 5 7.6 1.427 1.460 2.768 2.886 dbcsr_data_release 12724 10.6 2.332 2.847 2.332 2.847 matrix_ls_to_qs 2 6.0 0.000 0.000 2.402 2.536 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.514 2.517 dbcsr_sort_data 325 11.1 2.446 2.494 2.446 2.494 rebuild_ks_matrix 3 7.3 0.000 0.000 2.446 2.449 qs_ks_build_kohn_sham_matrix 3 8.3 0.010 0.011 2.446 2.449 dbcsr_new_transposed 4 7.5 0.255 0.275 2.305 2.323 dbcsr_frobenius_norm 74 6.6 2.057 2.136 2.190 2.229 dbcsr_add_d 103 6.2 0.000 0.000 2.136 2.205 dbcsr_add_anytype 103 7.2 0.860 0.891 2.136 2.205 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=108.249000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6846.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6ad6eaa91393e2e3e5bb87bfd4fb06be6378d79a_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 585.609216E+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 76559. MP_Allreduce 13232 2081. MP_Sync 1064 MP_Alltoall 2588 813144493. 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 0.027 0.095 298.231 298.232 qs_mol_dyn_low 1 2.0 0.004 0.067 297.087 297.108 qs_forces 5 3.8 0.005 0.024 296.086 296.089 qs_energies 5 4.8 0.001 0.035 292.528 292.553 scf_env_do_scf 5 5.8 0.000 0.001 274.430 274.433 scf_env_do_scf_inner_loop 105 6.6 0.002 0.006 238.848 238.852 qs_scf_new_mos 105 7.6 0.001 0.001 187.901 188.139 qs_scf_loop_do_ot 105 8.6 0.001 0.001 187.901 188.139 ot_scf_mini 105 9.6 0.003 0.004 177.821 177.989 dbcsr_multiply_generic 1445 12.2 0.126 0.134 139.493 140.343 velocity_verlet 4 3.0 0.001 0.006 120.235 120.237 multiply_cannon 1445 13.2 0.278 0.293 117.743 120.203 multiply_cannon_loop 1445 14.2 2.399 2.510 115.029 117.520 qs_ot_get_p 112 10.4 0.001 0.001 84.597 85.028 qs_ot_p2m_diag 40 11.0 0.020 0.030 72.315 72.404 cp_dbcsr_syevd 40 12.0 0.002 0.003 68.569 68.572 ot_mini 105 10.6 0.001 0.003 62.376 62.508 cp_fm_syevd 40 13.0 0.000 0.000 61.430 61.593 cp_fm_redistribute_end 40 14.0 27.905 55.664 27.915 55.671 cp_fm_syevd_base 40 14.0 27.736 55.509 27.736 55.509 mp_waitall_1 488190 16.1 38.591 47.678 38.591 47.678 multiply_cannon_multrec 69360 15.2 31.239 36.938 39.032 44.492 qs_ot_get_derivative 55 11.6 0.001 0.001 40.591 40.815 multiply_cannon_metrocomm3 69360 15.2 0.195 0.204 26.418 36.848 init_scf_loop 7 6.6 0.000 0.000 35.512 35.514 multiply_cannon_sync_h2d 69360 15.2 30.073 34.317 30.073 34.317 rebuild_ks_matrix 110 8.4 0.000 0.000 33.426 33.599 qs_ks_build_kohn_sham_matrix 110 9.4 0.012 0.016 33.425 33.599 qs_ks_update_qs_env 112 7.6 0.001 0.001 30.614 30.773 prepare_preconditioner 7 7.6 0.000 0.000 30.004 30.035 make_preconditioner 7 8.6 0.000 0.000 30.004 30.035 qs_rho_update_rho_low 110 7.6 0.001 0.002 24.519 24.905 calculate_rho_elec 110 8.6 0.030 0.034 24.518 24.905 apply_preconditioner_dbcsr 62 12.6 0.000 0.001 23.133 23.430 apply_single 62 13.6 0.000 0.000 23.132 23.430 rs_pw_transfer 690 11.5 0.011 0.012 20.255 21.424 make_full_inverse_cholesky 7 9.6 0.000 0.000 21.328 21.392 ot_new_cg_direction 55 11.6 0.001 0.002 21.096 21.098 density_rs2pw 110 9.6 0.006 0.008 17.969 19.109 qs_ot_get_orbitals 105 10.6 0.001 0.001 15.509 15.770 init_scf_run 5 5.8 0.000 0.001 15.247 15.248 scf_env_initial_rho_setup 5 6.8 0.003 0.011 15.247 15.248 cp_fm_cholesky_invert 7 10.6 14.616 14.639 14.616 14.639 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 14.484 14.591 mp_sum_l 4764 12.2 12.893 13.988 12.893 13.988 pw_transfer 1645 12.4 0.086 0.115 13.528 13.808 fft_wrap_pw1pw2 1425 13.5 0.014 0.017 13.383 13.666 sum_up_and_integrate 60 10.3 0.028 0.032 12.111 12.123 integrate_v_rspace 60 11.3 0.001 0.002 12.083 12.096 calculate_dm_sparse 110 9.5 0.000 0.001 11.692 11.867 check_diag 80 13.5 8.597 8.873 11.595 11.755 fft_wrap_pw1pw2_240 915 15.0 1.221 1.317 11.113 11.382 qs_vxc_create 110 10.4 0.002 0.005 11.163 11.226 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 11.126 11.188 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.010 10.111 10.127 acc_transpose_blocks 69360 15.2 0.363 0.387 9.511 9.799 calculate_first_density_matrix 1 7.0 0.000 0.004 9.643 9.655 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 9.550 9.625 multiply_cannon_metrocomm1 69360 15.2 0.092 0.098 4.718 9.498 fft3d_pb 915 16.0 2.393 2.697 9.135 9.486 make_m2s 2890 13.2 0.081 0.090 8.459 9.049 make_images 2890 14.2 0.244 0.264 8.350 8.939 dbcsr_mm_accdrv_process 154766 15.8 4.037 4.215 7.664 8.436 make_full_single_inverse 7 9.6 0.001 0.001 8.245 8.277 potential_pw2rs 60 12.3 0.003 0.003 7.880 7.915 xc_rho_set_and_dset_create 110 12.4 0.077 0.096 7.638 7.888 rs_pw_transfer_RS2PW_30 110 11.6 1.596 1.672 7.581 7.866 acc_transpose_blocks_kernels 69360 16.2 0.862 0.900 7.566 7.754 xc_vxc_pw_create 60 11.3 0.039 0.049 7.505 7.568 mp_sendrecv_dv 168740 12.6 7.407 7.561 7.407 7.561 jit_kernel_transpose 5 15.0 6.705 6.865 6.705 6.865 xc_pw_derive 510 13.4 0.006 0.007 6.773 6.839 mp_alltoall_z22v 2340 17.7 6.278 6.693 6.278 6.693 cp_fm_cholesky_decompose 14 10.2 6.174 6.184 6.174 6.184 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=298.232000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=557.200000, yerr=2.227106 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 6ad6eaa91393e2e3e5bb87bfd4fb06be6378d79a Summary: empty Status: OK