=== 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: 1e516dface89e0648f7d6ff6212138fc2d72d23d ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, # Cray-FFTW 3.3.8.10, COSMA 2.6.6, ELPA 2023.05.001, # HDF5 1.14.2, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.2.2, LIBVORI 220621, LIBXSMM 1.17, # PLUMED 2.9.0, SIRIUS 7.4.3, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Last update: 04.10.2023 # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2023.05.001 USE_HDF5 := 1.14.2 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.9.0 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a endif ifneq ($(USE_HDF5),) USE_HDF5 := $(strip $(USE_HDF5)) HDF5_INC := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/include HDF5_LIB := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/lib CFLAGS += -I$(HDF5_INC) DFLAGS += -D__HDF5 LIBS += $(HDF5_LIB)/libhdf5_fortran.a LIBS += $(HDF5_LIB)/libhdf5_hl.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1e516dface89e0648f7d6ff6212138fc2d72d23d_performance_tests/01 job id: 49302813 --- 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/1e516dface89e0648f7d6ff6212138fc2d72d23d_performance_tests/02 job id: 49302814 --- 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/1e516dface89e0648f7d6ff6212138fc2d72d23d_performance_tests/03 job id: 49302815 --- 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/1e516dface89e0648f7d6ff6212138fc2d72d23d_performance_tests/04 job id: 49302816 --- 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/1e516dface89e0648f7d6ff6212138fc2d72d23d_performance_tests/05 job id: 49302817 --- 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/1e516dface89e0648f7d6ff6212138fc2d72d23d_performance_tests/06 job id: 49302818 --- 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/1e516dface89e0648f7d6ff6212138fc2d72d23d_performance_tests/07 job id: 49302819 --- 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/1e516dface89e0648f7d6ff6212138fc2d72d23d_performance_tests/08 job id: 49302820 --- 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/1e516dface89e0648f7d6ff6212138fc2d72d23d_performance_tests/09 job id: 49302821 --- 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/1e516dface89e0648f7d6ff6212138fc2d72d23d_performance_tests/10 job id: 49302822 --- 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/1e516dface89e0648f7d6ff6212138fc2d72d23d_performance_tests/11 job id: 49302823 --- 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/1e516dface89e0648f7d6ff6212138fc2d72d23d_performance_tests/12 job id: 49302824 --- 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/1e516dface89e0648f7d6ff6212138fc2d72d23d_performance_tests/13 job id: 49302825 --- 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/1e516dface89e0648f7d6ff6212138fc2d72d23d_performance_tests/14 job id: 49302826 --- 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/1e516dface89e0648f7d6ff6212138fc2d72d23d_performance_tests/15 job id: 49302827 --- 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/1e516dface89e0648f7d6ff6212138fc2d72d23d_performance_tests/16 job id: 49302828 --- 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/1e516dface89e0648f7d6ff6212138fc2d72d23d_performance_tests/17 job id: 49302829 --- 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/1e516dface89e0648f7d6ff6212138fc2d72d23d_performance_tests/18 job id: 49302830 --- 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/1e516dface89e0648f7d6ff6212138fc2d72d23d_performance_tests/19 job id: 49302831 --- 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/1e516dface89e0648f7d6ff6212138fc2d72d23d_performance_tests/20 job id: 49302832 --- 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/1e516dface89e0648f7d6ff6212138fc2d72d23d_performance_tests/21 job id: 49302833 --- 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/1e516dface89e0648f7d6ff6212138fc2d72d23d_performance_tests/22 job id: 49302834 --- 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/1e516dface89e0648f7d6ff6212138fc2d72d23d_performance_tests/23 job id: 49302835 --- 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/1e516dface89e0648f7d6ff6212138fc2d72d23d_performance_tests/24 job id: 49302836 --- 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/1e516dface89e0648f7d6ff6212138fc2d72d23d_performance_tests/25 job id: 49302837 --- 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/1e516dface89e0648f7d6ff6212138fc2d72d23d_performance_tests/26 job id: 49302838 --- 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/1e516dface89e0648f7d6ff6212138fc2d72d23d_performance_tests/27 job id: 49302839 --- 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/1e516dface89e0648f7d6ff6212138fc2d72d23d_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 0.020 0.032 134.333 134.334 farming_run 1 2.0 133.338 133.341 134.303 134.307 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.486954E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 230 1134128. MP_Allreduce 571 1938539. MP_Sync 25 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 10 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.005 0.024 114.354 114.355 qs_energies 1 2.0 0.000 0.000 114.153 114.155 mp2_main 1 3.0 0.000 0.000 112.086 112.088 mp2_gpw_main 1 4.0 0.027 0.034 111.080 111.082 mp2_ri_gpw_compute_in 1 5.0 0.172 0.174 92.219 92.611 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 54.563 54.956 mp2_eri_3c_integrate_gpw 272 7.0 0.152 0.165 40.964 46.078 get_2c_integrals 1 6.0 0.008 0.009 36.664 37.483 integrate_v_rspace 273 8.0 0.441 0.455 24.590 29.448 pw_transfer 6555 10.6 0.380 0.396 26.725 27.271 fft_wrap_pw1pw2 5465 11.4 0.045 0.046 25.359 25.934 grid_integrate_task_list 273 9.0 20.506 25.825 20.506 25.825 fft_wrap_pw1pw2_100 2178 12.4 1.091 1.277 22.915 23.479 rpa_ri_compute_en 1 5.0 0.019 0.021 18.746 18.946 compute_2c_integrals 1 7.0 0.002 0.002 18.924 18.925 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.530 18.637 mp2_eri_2c_integrate_gpw 1 9.0 2.383 2.426 18.527 18.633 cp_fm_cholesky_decompose 12 8.2 17.683 18.516 17.683 18.516 cholesky_decomp 1 7.0 0.000 0.000 16.601 17.432 fft3d_s 5443 13.4 16.084 16.351 16.107 16.373 ao_to_mo_and_store_B_mult_1 272 7.0 10.766 15.309 10.766 15.309 calculate_wavefunction 272 8.0 5.377 5.534 12.281 12.986 rpa_num_int 1 6.0 0.000 0.001 10.615 10.616 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.591 10.614 calc_mat_Q 8 8.0 0.000 0.000 9.465 9.563 contract_S_to_Q 8 9.0 0.000 0.000 8.887 8.987 calc_potential_gpw 544 9.5 0.005 0.006 8.236 8.668 parallel_gemm_fm 14 9.1 0.000 0.000 8.496 8.594 parallel_gemm_fm_cosma 14 10.1 8.496 8.594 8.496 8.594 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.039 8.248 potential_pw2rs 545 10.0 0.107 0.109 7.435 8.073 create_integ_mat 1 6.0 0.022 0.027 7.644 7.644 collocate_single_gaussian 272 10.0 0.040 0.042 7.301 7.498 array2fm 1 7.0 0.000 0.000 6.633 7.069 pw_scatter_s 2720 13.7 4.365 4.441 4.365 4.441 pw_gather_s 2722 13.2 3.421 3.748 3.421 3.748 array2fm_buffer_send 1 8.0 2.947 3.129 2.947 3.129 pw_poisson_solve 545 10.5 1.115 1.165 2.183 2.306 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.080241, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2793.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1e516dface89e0648f7d6ff6212138fc2d72d23d_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 205321. MP_Allreduce 424 9. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.023 0.034 420.222 420.223 farming_run 1 2.0 419.336 419.342 420.189 420.192 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.237299E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 757 478553. MP_Allreduce 2021 21391. MP_Sync 37 MP_Alltoall 77 16601612. MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.031 216.843 216.844 qs_energies 1 2.0 0.000 0.000 216.523 216.528 scf_env_do_scf 1 3.0 0.000 0.000 114.116 114.117 qs_ks_update_qs_env 5 5.0 0.000 0.000 113.216 113.222 rebuild_ks_matrix 4 6.0 0.000 0.000 113.215 113.221 qs_ks_build_kohn_sham_matrix 4 7.0 0.054 0.060 113.215 113.221 hfx_ks_matrix 4 8.0 0.001 0.001 112.869 112.874 integrate_four_center 4 9.0 0.155 0.480 112.868 112.873 integrate_four_center_main 4 10.0 0.101 0.590 101.972 104.361 integrate_four_center_bin 265 11.0 101.872 104.358 101.872 104.358 mp2_main 1 3.0 0.000 0.000 102.114 102.120 mp2_gpw_main 1 4.0 0.050 0.077 101.262 101.273 init_scf_loop 1 4.0 0.000 0.000 96.252 96.252 mp2_ri_gpw_compute_in 1 5.0 0.065 0.072 74.215 75.554 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.035 55.373 mp2_eri_3c_integrate_gpw 91 7.0 0.145 0.162 41.491 46.579 integrate_v_rspace 95 8.0 0.396 0.565 27.962 32.872 pw_transfer 2240 10.6 0.146 0.167 29.453 29.952 mp2_ri_gpw_compute_en 1 5.0 0.056 0.069 26.882 29.265 ao_to_mo_and_store_B_mult_1 91 7.0 10.863 29.050 10.863 29.050 fft_wrap_pw1pw2 1868 11.4 0.018 0.020 28.471 28.896 grid_integrate_task_list 95 9.0 23.384 28.486 23.384 28.486 fft_wrap_pw1pw2_100 730 12.4 1.271 1.463 26.228 26.561 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.830 1.896 24.876 24.886 get_2c_integrals 1 6.0 0.000 0.000 20.088 20.117 compute_2c_integrals 1 7.0 0.003 0.003 19.071 19.079 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 18.558 18.952 mp2_eri_2c_integrate_gpw 1 9.0 1.731 1.833 18.557 18.951 fft3d_s 1823 13.4 18.481 18.780 18.495 18.793 scf_env_do_scf_inner_loop 4 4.0 0.000 0.001 17.863 17.863 calculate_wavefunction 91 8.0 2.015 2.046 9.617 9.842 mp2_ri_gpw_compute_en_expansio 172 7.0 0.556 0.581 8.745 9.332 potential_pw2rs 186 10.0 0.033 0.034 8.389 8.923 local_gemm 172 8.0 8.189 8.751 8.189 8.751 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.112 8.537 mp2_ri_gpw_compute_en_comm 22 7.0 0.501 0.527 7.920 8.379 calc_potential_gpw 182 9.5 0.002 0.002 7.893 8.086 collocate_single_gaussian 91 10.0 0.017 0.024 7.732 8.012 integrate_four_center_load 4 10.0 0.000 0.000 6.742 6.747 hfx_load_balance 1 11.0 0.000 0.000 6.742 6.747 mp2_ri_gpw_compute_en_ener 172 7.0 6.336 6.402 6.336 6.402 mp_sendrecv_dm3 2068 8.0 5.953 6.395 5.953 6.395 mp_sync 37 10.5 2.886 5.972 2.886 5.972 pw_gather_s 912 13.2 4.465 4.918 4.465 4.918 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.253909, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1497.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1e516dface89e0648f7d6ff6212138fc2d72d23d_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 452.554752E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62379. MP_Allreduce 10329 270. MP_Sync 530 MP_Alltoall 2083 592243. MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.006 0.026 53.860 53.861 qs_mol_dyn_low 1 2.0 0.003 0.003 53.611 53.619 qs_forces 11 3.9 0.002 0.002 53.534 53.534 qs_energies 11 4.9 0.001 0.001 52.036 52.063 scf_env_do_scf 11 5.9 0.000 0.001 46.179 46.179 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 44.140 44.140 qs_scf_new_mos 108 7.5 0.000 0.001 34.234 34.557 qs_scf_loop_do_ot 108 8.5 0.000 0.001 34.233 34.557 dbcsr_multiply_generic 2286 12.5 0.097 0.101 33.869 34.365 ot_scf_mini 108 9.5 0.002 0.002 32.548 32.740 multiply_cannon 2286 13.5 0.192 0.203 26.378 28.066 multiply_cannon_loop 2286 14.5 1.846 1.970 25.733 27.440 velocity_verlet 10 3.0 0.001 0.001 26.585 26.585 ot_mini 108 10.5 0.001 0.001 19.735 19.984 qs_ot_get_derivative 108 11.5 0.001 0.001 16.642 16.830 mp_waitall_1 245248 16.5 8.597 14.955 8.597 14.955 multiply_cannon_metrocomm3 54864 15.5 0.074 0.083 6.196 12.932 multiply_cannon_multrec 54864 15.5 3.633 5.612 7.603 10.915 qs_ot_get_p 119 10.4 0.001 0.001 8.049 8.344 rebuild_ks_matrix 119 8.3 0.000 0.000 7.758 7.914 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 7.757 7.914 mp_sum_l 7287 12.8 5.610 7.304 5.610 7.304 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.837 6.976 multiply_cannon_sync_h2d 54864 15.5 5.116 6.408 5.116 6.408 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.875 6.289 dbcsr_mm_accdrv_process 76910 16.1 1.839 2.910 3.884 5.574 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.405 5.516 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.223 5.266 init_scf_run 11 5.9 0.000 0.001 4.576 4.576 scf_env_initial_rho_setup 11 6.9 0.000 0.001 4.576 4.576 sum_up_and_integrate 119 10.3 0.001 0.002 4.389 4.395 integrate_v_rspace 119 11.3 0.002 0.002 4.379 4.385 cp_dbcsr_syevd 50 12.0 0.002 0.003 4.377 4.377 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.189 4.190 cp_fm_redistribute_end 50 14.0 2.138 4.166 2.144 4.169 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.040 4.148 calculate_rho_elec 119 8.7 0.011 0.017 4.039 4.148 cp_fm_diag_elpa_base 50 14.0 2.020 4.070 2.024 4.078 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.923 3.184 apply_single 119 13.6 0.000 0.000 2.923 3.183 calculate_dm_sparse 119 9.5 0.000 0.000 2.904 3.079 multiply_cannon_metrocomm1 54864 15.5 0.057 0.062 1.799 2.914 ot_diis_step 108 11.5 0.006 0.006 2.803 2.803 acc_transpose_blocks 54864 15.5 0.229 0.260 2.242 2.781 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.547 2.649 jit_kernel_multiply 13 15.8 1.983 2.586 1.983 2.586 calculate_first_density_matrix 1 7.0 0.000 0.001 2.367 2.379 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.206 2.209 density_rs2pw 119 9.7 0.004 0.004 2.058 2.141 grid_integrate_task_list 119 12.3 2.019 2.134 2.019 2.134 wfi_extrapolate 11 7.9 0.001 0.001 2.133 2.133 mp_sum_d 4135 12.0 1.440 2.045 1.440 2.045 init_scf_loop 11 6.9 0.000 0.000 2.023 2.023 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.908 1.985 potential_pw2rs 119 12.3 0.004 0.004 1.750 1.761 pw_transfer 1439 11.6 0.052 0.055 1.580 1.648 make_m2s 4572 13.5 0.053 0.056 1.561 1.601 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.505 1.574 mp_alltoall_d11v 2130 13.8 1.303 1.540 1.303 1.540 make_images 4572 14.5 0.133 0.138 1.478 1.519 acc_transpose_blocks_sync 164592 16.5 1.222 1.470 1.222 1.470 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.001 1.391 1.436 mp_waitany 12084 13.8 1.230 1.417 1.230 1.417 transfer_rs2pw 487 10.6 0.005 0.006 1.303 1.411 grid_collocate_task_list 119 9.7 1.349 1.402 1.349 1.402 transfer_pw2rs 487 13.2 0.006 0.007 1.289 1.297 fft3d_ps 1201 14.6 0.368 0.474 1.219 1.284 fft_wrap_pw1pw2_140 487 13.2 0.141 0.155 1.167 1.238 dbcsr_dot_sd 1205 11.9 0.048 0.059 0.810 1.212 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=53.861000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=431.272727, yerr=1.052349 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1e516dface89e0648f7d6ff6212138fc2d72d23d_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 488.419328E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2194560 MPI messages size (bytes): total size 310.646604E+09 min size 0.000000E+00 max size 1.145520E+06 average size 141.553031E+03 MPI breakdown and total messages size (bytes): size <= 128 724648 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 281952 4619501568 32768 < size <= 131072 494448 39143342080 131072 < size <= 4194304 440000 264807943488 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62658. MP_Allreduce 10306 303. MP_Sync 54 MP_Alltoall 2060 960718. 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 0.013 0.033 39.717 39.718 qs_mol_dyn_low 1 2.0 0.003 0.003 39.422 39.429 qs_forces 11 3.9 0.002 0.003 39.312 39.312 qs_energies 11 4.9 0.001 0.001 37.655 37.659 scf_env_do_scf 11 5.9 0.000 0.001 32.088 32.088 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 29.697 29.697 dbcsr_multiply_generic 2286 12.5 0.101 0.104 23.184 23.523 qs_scf_new_mos 108 7.5 0.001 0.001 21.186 21.416 qs_scf_loop_do_ot 108 8.5 0.001 0.001 21.186 21.416 ot_scf_mini 108 9.5 0.002 0.003 20.267 20.434 multiply_cannon 2286 13.5 0.211 0.218 18.014 19.493 multiply_cannon_loop 2286 14.5 1.197 1.260 16.874 18.465 velocity_verlet 10 3.0 0.001 0.001 18.198 18.200 ot_mini 108 10.5 0.001 0.001 12.774 13.011 mp_waitall_1 200699 16.5 5.953 11.308 5.953 11.308 qs_ot_get_derivative 108 11.5 0.001 0.001 10.420 10.590 multiply_cannon_metrocomm3 27432 15.5 0.072 0.075 4.420 10.091 multiply_cannon_multrec 27432 15.5 1.869 4.242 6.766 9.669 dbcsr_mm_accdrv_process 47894 16.0 4.054 6.498 4.816 7.196 rebuild_ks_matrix 119 8.3 0.000 0.000 6.691 6.822 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.691 6.822 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.886 6.004 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 4.028 5.037 qs_ot_get_p 119 10.4 0.001 0.001 4.582 4.806 init_scf_run 11 5.9 0.000 0.001 4.358 4.359 scf_env_initial_rho_setup 11 6.9 0.000 0.000 4.358 4.359 mp_sum_l 7287 12.8 2.152 4.261 2.152 4.261 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.152 4.211 apply_single 119 13.6 0.000 0.000 3.152 4.211 sum_up_and_integrate 119 10.3 0.001 0.001 3.629 3.633 integrate_v_rspace 119 11.3 0.002 0.003 3.614 3.619 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.545 3.572 calculate_rho_elec 119 8.7 0.021 0.024 3.544 3.571 qs_ot_p2m_diag 50 11.0 0.009 0.013 2.965 2.984 make_m2s 4572 13.5 0.052 0.053 2.563 2.884 calculate_first_density_matrix 1 7.0 0.000 0.000 2.868 2.870 make_images 4572 14.5 0.205 0.243 2.474 2.798 calculate_dm_sparse 119 9.5 0.000 0.001 2.528 2.606 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.569 2.570 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.301 2.388 init_scf_loop 11 6.9 0.000 0.000 2.371 2.372 ot_diis_step 108 11.5 0.010 0.011 2.306 2.307 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.261 2.261 multiply_cannon_sync_h2d 27432 15.5 1.700 2.243 1.700 2.243 cp_fm_redistribute_end 50 14.0 1.148 2.238 1.151 2.240 cp_fm_diag_elpa_base 50 14.0 1.058 2.147 1.086 2.182 acc_transpose_blocks 27432 15.5 0.114 0.121 1.690 2.181 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.997 1.999 grid_integrate_task_list 119 12.3 1.835 1.947 1.835 1.947 density_rs2pw 119 9.7 0.004 0.004 1.846 1.907 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.791 1.832 pw_transfer 1439 11.6 0.065 0.069 1.794 1.828 jit_kernel_multiply 8 16.5 0.705 1.748 0.705 1.748 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.702 1.739 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.716 1.724 make_images_data 4572 15.5 0.047 0.053 1.222 1.571 prepare_preconditioner 11 7.9 0.000 0.000 1.503 1.530 make_preconditioner 11 8.9 0.000 0.000 1.503 1.530 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.403 1.459 wfi_extrapolate 11 7.9 0.001 0.001 1.432 1.432 fft_wrap_pw1pw2_140 487 13.2 0.160 0.169 1.363 1.400 hybrid_alltoall_any 4725 16.4 0.053 0.115 1.088 1.397 potential_pw2rs 119 12.3 0.006 0.006 1.386 1.392 fft3d_ps 1201 14.6 0.516 0.571 1.338 1.367 grid_collocate_task_list 119 9.7 1.287 1.348 1.287 1.348 mp_alltoall_d11v 2130 13.8 1.179 1.292 1.179 1.292 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.193 1.240 transfer_rs2pw 487 10.6 0.005 0.005 1.038 1.097 acc_transpose_blocks_kernels 27432 16.5 0.189 0.280 0.726 1.087 mp_sum_d 4135 12.0 0.524 1.015 0.524 1.015 qs_energies_init_hamiltonians 11 5.9 0.002 0.007 0.950 0.951 mp_allgather_i34 2286 14.5 0.562 0.951 0.562 0.951 acc_transpose_blocks_sync 82296 16.5 0.824 0.950 0.824 0.950 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.862 0.875 transfer_pw2rs 487 13.2 0.005 0.005 0.868 0.871 jit_kernel_transpose 5 15.5 0.537 0.809 0.537 0.809 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=39.718000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=464.818182, yerr=1.526623 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1e516dface89e0648f7d6ff6212138fc2d72d23d_performance_tests/05/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 59.051995E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3143552 0.0% 0.0% 100.0% average stack size 0.0 0.0 46.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 519.405568E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63490. MP_Allreduce 10155 305. MP_Sync 54 MP_Alltoall 1821 1607811. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.028 32.982 32.983 qs_mol_dyn_low 1 2.0 0.003 0.003 32.576 32.596 qs_forces 11 3.9 0.002 0.002 32.491 32.492 qs_energies 11 4.9 0.001 0.002 30.949 30.951 scf_env_do_scf 11 5.9 0.000 0.001 25.776 25.776 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 23.379 23.380 dbcsr_multiply_generic 2286 12.5 0.096 0.098 17.925 17.985 qs_scf_new_mos 108 7.5 0.001 0.001 15.776 15.795 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.775 15.795 multiply_cannon 2286 13.5 0.195 0.203 14.685 15.394 velocity_verlet 10 3.0 0.008 0.016 15.056 15.058 ot_scf_mini 108 9.5 0.002 0.002 15.020 15.032 multiply_cannon_loop 2286 14.5 0.860 0.896 13.914 14.755 ot_mini 108 10.5 0.001 0.001 9.479 9.491 qs_ot_get_derivative 108 11.5 0.001 0.001 7.997 8.009 multiply_cannon_multrec 18288 15.5 2.052 3.102 7.345 7.542 dbcsr_mm_accdrv_process 38222 16.0 5.111 6.145 5.199 6.208 rebuild_ks_matrix 119 8.3 0.000 0.000 5.922 5.939 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.922 5.938 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.217 5.233 init_scf_run 11 5.9 0.000 0.001 4.015 4.015 scf_env_initial_rho_setup 11 6.9 0.000 0.001 4.015 4.015 mp_waitall_1 158411 16.6 2.582 3.719 2.582 3.719 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.962 3.601 sum_up_and_integrate 119 10.3 0.001 0.001 3.512 3.517 integrate_v_rspace 119 11.3 0.002 0.003 3.499 3.505 qs_ot_get_p 119 10.4 0.001 0.001 3.292 3.313 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.207 3.212 calculate_rho_elec 119 8.7 0.031 0.031 3.207 3.211 calculate_first_density_matrix 1 7.0 0.000 0.000 2.797 2.798 multiply_cannon_metrocomm3 18288 15.5 0.048 0.050 1.535 2.549 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.042 2.407 apply_single 119 13.6 0.000 0.000 2.042 2.406 init_scf_loop 11 6.9 0.000 0.000 2.377 2.379 acc_transpose_blocks 18288 15.5 0.080 0.082 2.285 2.360 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.202 2.211 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.195 2.197 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.036 2.043 calculate_dm_sparse 119 9.5 0.000 0.001 1.937 1.946 make_m2s 4572 13.5 0.044 0.045 1.777 1.928 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.905 1.906 grid_integrate_task_list 119 12.3 1.804 1.884 1.804 1.884 density_rs2pw 119 9.7 0.004 0.004 1.774 1.869 make_images 4572 14.5 0.192 0.204 1.692 1.842 pw_transfer 1439 11.6 0.065 0.068 1.803 1.811 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.711 1.720 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.713 1.719 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.659 1.672 cp_fm_diag_elpa_base 50 14.0 1.636 1.648 1.658 1.671 prepare_preconditioner 11 7.9 0.000 0.000 1.625 1.631 make_preconditioner 11 8.9 0.000 0.000 1.625 1.631 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.480 1.568 acc_transpose_blocks_kernels 18288 16.5 0.219 0.225 1.440 1.526 ot_diis_step 108 11.5 0.011 0.011 1.466 1.466 mp_sum_l 7287 12.8 1.080 1.461 1.080 1.461 fft_wrap_pw1pw2_140 487 13.2 0.212 0.217 1.384 1.395 potential_pw2rs 119 12.3 0.007 0.008 1.318 1.321 grid_collocate_task_list 119 9.7 1.233 1.306 1.233 1.306 jit_kernel_transpose 5 15.6 1.221 1.302 1.221 1.302 fft3d_ps 1201 14.6 0.526 0.542 1.279 1.287 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.276 1.281 multiply_cannon_sync_h2d 18288 15.5 1.039 1.167 1.039 1.167 wfi_extrapolate 11 7.9 0.001 0.001 1.166 1.166 transfer_rs2pw 487 10.6 0.005 0.005 0.947 1.030 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.955 0.971 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 0.941 0.941 make_images_data 4572 15.5 0.047 0.051 0.758 0.933 hybrid_alltoall_any 4725 16.4 0.058 0.117 0.658 0.866 acc_transpose_blocks_sync 54864 16.5 0.745 0.810 0.745 0.810 transfer_pw2rs 487 13.2 0.004 0.004 0.798 0.802 mp_alltoall_d11v 2130 13.8 0.657 0.782 0.657 0.782 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.775 0.777 jit_kernel_multiply 2 17.0 0.030 0.771 0.030 0.771 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.679 0.750 cp_fm_cholesky_invert 11 10.9 0.680 0.683 0.680 0.683 mp_alltoall_z22v 1201 16.6 0.607 0.675 0.607 0.675 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=32.983000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=495.181818, yerr=0.833196 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1e516dface89e0648f7d6ff6212138fc2d72d23d_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 556.302336E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63489. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.027 37.451 37.452 qs_mol_dyn_low 1 2.0 0.003 0.003 37.035 37.044 qs_forces 11 3.9 0.002 0.004 36.950 36.952 qs_energies 11 4.9 0.001 0.003 35.203 35.209 scf_env_do_scf 11 5.9 0.001 0.001 29.682 29.684 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 26.304 26.305 dbcsr_multiply_generic 2286 12.5 0.100 0.103 20.667 20.789 velocity_verlet 10 3.0 0.001 0.001 18.372 18.375 qs_scf_new_mos 108 7.5 0.001 0.001 18.288 18.343 qs_scf_loop_do_ot 108 8.5 0.001 0.001 18.288 18.342 multiply_cannon 2286 13.5 0.219 0.226 16.833 17.323 ot_scf_mini 108 9.5 0.002 0.003 17.257 17.314 multiply_cannon_loop 2286 14.5 1.518 1.574 15.903 16.291 ot_mini 108 10.5 0.001 0.001 10.684 10.751 multiply_cannon_multrec 27432 15.5 2.654 3.399 9.362 9.699 qs_ot_get_derivative 108 11.5 0.001 0.001 8.813 8.870 dbcsr_mm_accdrv_process 47916 15.9 6.131 7.802 6.604 8.253 rebuild_ks_matrix 119 8.3 0.000 0.000 6.280 6.326 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.280 6.326 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.563 5.604 init_scf_run 11 5.9 0.000 0.001 4.168 4.168 scf_env_initial_rho_setup 11 6.9 0.000 0.001 4.167 4.168 qs_ot_get_p 119 10.4 0.001 0.002 3.738 3.810 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.078 3.624 sum_up_and_integrate 119 10.3 0.001 0.001 3.474 3.483 integrate_v_rspace 119 11.3 0.003 0.004 3.463 3.472 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.322 3.359 calculate_rho_elec 119 8.7 0.040 0.046 3.322 3.359 init_scf_loop 11 6.9 0.000 0.002 3.357 3.358 acc_transpose_blocks 27432 15.5 0.121 0.124 2.809 3.056 calculate_first_density_matrix 1 7.0 0.000 0.002 2.773 2.774 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.252 2.733 apply_single 119 13.6 0.000 0.000 2.252 2.733 prepare_preconditioner 11 7.9 0.000 0.000 2.530 2.538 make_preconditioner 11 8.9 0.000 0.001 2.530 2.538 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.135 2.463 qs_ot_get_derivative_diag 49 12.0 0.001 0.002 2.347 2.376 make_m2s 4572 13.5 0.054 0.057 2.254 2.362 qs_ot_p2m_diag 50 11.0 0.016 0.023 2.314 2.325 calculate_dm_sparse 119 9.5 0.000 0.000 2.253 2.306 make_images 4572 14.5 0.273 0.334 2.147 2.253 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.246 2.247 mp_waitall_1 137007 16.6 1.730 2.223 1.730 2.223 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.071 2.083 pw_transfer 1439 11.6 0.065 0.069 1.918 1.955 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.912 1.912 grid_integrate_task_list 119 12.3 1.836 1.906 1.836 1.906 density_rs2pw 119 9.7 0.004 0.004 1.782 1.905 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.825 1.867 ot_diis_step 108 11.5 0.012 0.012 1.828 1.828 mp_sum_l 7287 12.8 1.156 1.747 1.156 1.747 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.658 1.667 cp_fm_diag_elpa_base 50 14.0 1.623 1.639 1.656 1.665 fft_wrap_pw1pw2_140 487 13.2 0.245 0.258 1.504 1.547 acc_transpose_blocks_sync 82296 16.5 1.406 1.521 1.406 1.521 acc_transpose_blocks_kernels 27432 16.5 0.271 0.278 1.254 1.386 fft3d_ps 1201 14.6 0.554 0.606 1.344 1.368 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.001 1.352 1.363 wfi_extrapolate 11 7.9 0.001 0.001 1.342 1.342 grid_collocate_task_list 119 9.7 1.249 1.335 1.249 1.335 multiply_cannon_metrocomm3 27432 15.5 0.041 0.042 0.741 1.316 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.287 1.308 potential_pw2rs 119 12.3 0.009 0.009 1.281 1.286 cp_fm_upper_to_full 72 14.2 0.814 1.160 0.814 1.160 jit_kernel_multiply 6 16.4 0.409 1.126 0.409 1.126 jit_kernel_transpose 5 15.6 0.983 1.117 0.983 1.117 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.087 1.089 dbcsr_complete_redistribute 329 12.2 0.121 0.146 0.780 1.064 transfer_rs2pw 487 10.6 0.004 0.005 0.912 1.043 make_images_data 4572 15.5 0.047 0.051 0.826 0.939 hybrid_alltoall_any 4725 16.4 0.065 0.156 0.721 0.880 build_core_hamiltonian_matrix_ 11 4.9 0.006 0.048 0.806 0.878 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.569 0.846 mp_alltoall_d11v 2130 13.8 0.724 0.816 0.724 0.816 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.807 0.811 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=37.452000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=528.727273, yerr=2.733327 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1e516dface89e0648f7d6ff6212138fc2d72d23d_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 627.810304E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63488. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.057 0.137 29.783 29.788 qs_mol_dyn_low 1 2.0 0.003 0.003 29.163 29.204 qs_forces 11 3.9 0.002 0.003 28.887 28.889 qs_energies 11 4.9 0.001 0.001 27.195 27.199 scf_env_do_scf 11 5.9 0.000 0.001 21.813 21.813 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 19.416 19.417 velocity_verlet 10 3.0 0.001 0.001 14.647 14.663 dbcsr_multiply_generic 2286 12.5 0.094 0.098 13.916 13.980 qs_scf_new_mos 108 7.5 0.001 0.001 11.907 11.936 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.907 11.935 multiply_cannon 2286 13.5 0.224 0.232 11.361 11.862 ot_scf_mini 108 9.5 0.002 0.002 11.206 11.231 multiply_cannon_loop 2286 14.5 0.642 0.664 10.446 10.649 multiply_cannon_multrec 9144 15.5 1.784 2.033 6.359 6.599 ot_mini 108 10.5 0.001 0.001 6.553 6.582 rebuild_ks_matrix 119 8.3 0.000 0.000 5.694 5.717 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.693 5.717 qs_ot_get_derivative 108 11.5 0.001 0.001 5.272 5.298 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.064 5.085 dbcsr_mm_accdrv_process 12550 15.8 3.751 4.480 4.464 4.544 init_scf_run 11 5.9 0.000 0.001 3.894 3.894 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.894 3.894 sum_up_and_integrate 119 10.3 0.001 0.001 3.386 3.389 integrate_v_rspace 119 11.3 0.003 0.003 3.375 3.379 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.255 3.263 calculate_rho_elec 119 8.7 0.060 0.061 3.255 3.262 qs_ot_get_p 119 10.4 0.001 0.001 2.782 2.820 calculate_first_density_matrix 1 7.0 0.000 0.000 2.772 2.773 init_scf_loop 11 6.9 0.000 0.000 2.378 2.379 acc_transpose_blocks 9144 15.5 0.042 0.043 2.003 2.066 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.019 2.020 calculate_dm_sparse 119 9.5 0.000 0.000 1.960 1.978 grid_integrate_task_list 119 12.3 1.870 1.941 1.870 1.941 make_m2s 4572 13.5 0.034 0.035 1.772 1.937 pw_transfer 1439 11.6 0.065 0.068 1.926 1.935 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.859 1.878 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.860 1.869 make_images 4572 14.5 0.270 0.300 1.683 1.845 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.833 1.844 mp_waitall_1 115863 16.7 1.316 1.812 1.316 1.812 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.802 1.806 density_rs2pw 119 9.7 0.003 0.004 1.656 1.692 prepare_preconditioner 11 7.9 0.000 0.000 1.664 1.667 make_preconditioner 11 8.9 0.000 0.000 1.664 1.667 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.558 1.584 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.561 1.562 fft_wrap_pw1pw2_140 487 13.2 0.322 0.332 1.515 1.526 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.356 1.361 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.343 1.354 jit_kernel_multiply 6 15.3 0.674 1.346 0.674 1.346 grid_collocate_task_list 119 9.7 1.297 1.339 1.297 1.339 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.293 1.301 cp_fm_diag_elpa_base 50 14.0 1.266 1.282 1.291 1.300 acc_transpose_blocks_kernels 9144 16.5 0.118 0.120 1.228 1.293 ot_diis_step 108 11.5 0.013 0.013 1.270 1.270 fft3d_ps 1201 14.6 0.556 0.566 1.247 1.256 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.211 1.247 apply_single 119 13.6 0.000 0.000 1.210 1.247 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.217 1.220 potential_pw2rs 119 12.3 0.010 0.011 1.203 1.205 jit_kernel_transpose 5 15.6 1.111 1.176 1.111 1.176 wfi_extrapolate 11 7.9 0.001 0.001 1.072 1.072 hybrid_alltoall_any 4725 16.4 0.065 0.175 0.746 0.980 make_images_data 4572 15.5 0.041 0.045 0.767 0.951 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.876 0.925 cp_fm_cholesky_invert 11 10.9 0.845 0.848 0.845 0.848 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.810 0.816 transfer_rs2pw 487 10.6 0.004 0.004 0.750 0.804 mp_alltoall_d11v 2130 13.8 0.729 0.787 0.729 0.787 acc_transpose_blocks_sync 27432 16.5 0.717 0.753 0.717 0.753 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.750 0.753 qs_env_update_s_mstruct 11 6.9 0.000 0.001 0.662 0.713 multiply_cannon_metrocomm3 9144 15.5 0.020 0.021 0.345 0.675 mp_allgather_i34 2286 14.5 0.232 0.653 0.232 0.653 transfer_pw2rs 487 13.2 0.003 0.004 0.649 0.650 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=29.788000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=594.000000, yerr=7.286351 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1e516dface89e0648f7d6ff6212138fc2d72d23d_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 774.418432E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63723. MP_Allreduce 10154 429. MP_Sync 54 MP_Alltoall 1582 7383731. MP_SendRecv 2499 189067. MP_ISendRecv 2499 189067. MP_Wait 6399 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.041 43.775 43.775 qs_mol_dyn_low 1 2.0 0.003 0.003 43.493 43.500 qs_forces 11 3.9 0.002 0.003 43.422 43.422 qs_energies 11 4.9 0.001 0.001 41.437 41.441 scf_env_do_scf 11 5.9 0.001 0.001 35.057 35.057 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 27.261 27.262 velocity_verlet 10 3.0 0.001 0.002 24.050 24.057 dbcsr_multiply_generic 2286 12.5 0.103 0.107 20.151 20.387 qs_scf_new_mos 108 7.5 0.001 0.001 17.847 17.949 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.846 17.949 multiply_cannon 2286 13.5 0.298 0.305 16.017 17.150 ot_scf_mini 108 9.5 0.002 0.002 16.684 16.790 multiply_cannon_loop 2286 14.5 0.869 0.890 14.742 15.915 ot_mini 108 10.5 0.001 0.001 10.322 10.445 multiply_cannon_multrec 9144 15.5 3.480 4.741 9.375 9.495 qs_ot_get_derivative 108 11.5 0.001 0.001 8.246 8.353 init_scf_loop 11 6.9 0.000 0.000 7.767 7.768 rebuild_ks_matrix 119 8.3 0.000 0.000 7.096 7.243 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.096 7.243 dbcsr_mm_accdrv_process 12550 15.8 4.778 6.999 5.755 7.031 prepare_preconditioner 11 7.9 0.000 0.000 6.801 6.815 make_preconditioner 11 8.9 0.000 0.000 6.801 6.814 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.357 6.694 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.402 6.535 cp_fm_upper_to_full 72 14.2 3.153 4.541 3.153 4.541 init_scf_run 11 5.9 0.000 0.001 4.282 4.282 scf_env_initial_rho_setup 11 6.9 0.000 0.001 4.281 4.282 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.050 4.071 calculate_rho_elec 119 8.7 0.118 0.121 4.050 4.071 sum_up_and_integrate 119 10.3 0.001 0.001 3.781 3.788 integrate_v_rspace 119 11.3 0.003 0.004 3.771 3.778 qs_ot_get_p 119 10.4 0.001 0.001 3.417 3.564 qs_ot_get_derivative_taylor 59 13.0 0.001 0.002 2.969 3.533 mp_waitall_1 94719 16.7 2.416 3.466 2.416 3.466 dbcsr_complete_redistribute 329 12.2 0.285 0.291 1.995 2.809 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.366 2.788 apply_single 119 13.6 0.000 0.000 2.365 2.788 calculate_first_density_matrix 1 7.0 0.000 0.000 2.767 2.773 pw_transfer 1439 11.6 0.068 0.069 2.647 2.653 calculate_dm_sparse 119 9.5 0.000 0.000 2.632 2.651 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.549 2.554 make_m2s 4572 13.5 0.037 0.038 2.374 2.531 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.675 2.494 make_images 4572 14.5 0.353 0.384 2.253 2.411 multiply_cannon_metrocomm3 9144 15.5 0.021 0.021 1.453 2.409 mp_alltoall_i22 627 13.8 1.453 2.295 1.453 2.295 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.438 2.251 acc_transpose_blocks 9144 15.5 0.045 0.046 2.069 2.226 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.120 2.175 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.163 2.165 fft_wrap_pw1pw2_140 487 13.2 0.576 0.582 2.120 2.129 grid_integrate_task_list 119 12.3 2.073 2.096 2.073 2.096 density_rs2pw 119 9.7 0.003 0.003 2.074 2.087 mp_sum_l 7287 12.8 1.211 2.085 1.211 2.085 ot_diis_step 108 11.5 0.014 0.014 2.048 2.048 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.015 2.017 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.860 1.879 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.813 1.814 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.657 1.657 fft3d_ps 1201 14.6 0.595 0.607 1.645 1.648 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.635 1.646 grid_collocate_task_list 119 9.7 1.517 1.533 1.517 1.533 wfi_extrapolate 11 7.9 0.001 0.001 1.448 1.448 cp_fm_cholesky_invert 11 10.9 1.431 1.434 1.431 1.434 jit_kernel_multiply 6 15.5 0.948 1.395 0.948 1.395 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.354 1.354 cp_fm_diag_elpa_base 50 14.0 1.209 1.261 1.352 1.352 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.319 1.342 potential_pw2rs 119 12.3 0.014 0.014 1.287 1.289 hybrid_alltoall_any 4725 16.4 0.090 0.151 1.083 1.276 mp_alltoall_d11v 2130 13.8 1.193 1.224 1.193 1.224 make_images_data 4572 15.5 0.046 0.049 1.014 1.211 acc_transpose_blocks_sync 27432 16.5 1.109 1.141 1.109 1.141 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.118 1.136 acc_transpose_blocks_kernels 9144 16.5 0.120 0.121 0.899 1.043 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.998 1.029 qs_create_task_list 11 7.9 0.000 0.000 0.947 0.959 generate_qs_task_list 11 8.9 0.369 0.388 0.947 0.959 mp_alltoall_z22v 1201 16.6 0.914 0.936 0.914 0.936 jit_kernel_transpose 5 15.6 0.779 0.923 0.779 0.923 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.905 0.920 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=43.775000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=728.636364, yerr=12.984415 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1e516dface89e0648f7d6ff6212138fc2d72d23d_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 500.695040E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66212. MP_Allreduce 9776 488. MP_Sync 52 MP_Alltoall 1938 454537. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.100 82.698 82.699 qs_mol_dyn_low 1 2.0 0.003 0.005 82.067 82.081 qs_forces 11 3.9 0.003 0.007 81.989 81.990 qs_energies 11 4.9 0.001 0.003 79.119 79.137 scf_env_do_scf 11 5.9 0.000 0.001 69.914 69.917 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 64.546 64.547 dbcsr_multiply_generic 2055 12.4 0.106 0.109 52.045 52.312 qs_scf_new_mos 99 7.5 0.000 0.001 47.633 47.794 qs_scf_loop_do_ot 99 8.5 0.000 0.001 47.632 47.793 ot_scf_mini 99 9.5 0.002 0.002 45.261 45.354 multiply_cannon 2055 13.4 0.186 0.190 43.530 44.382 multiply_cannon_loop 2055 14.4 1.808 1.853 42.594 43.495 velocity_verlet 10 3.0 0.001 0.001 42.875 42.876 ot_mini 99 10.5 0.001 0.001 26.960 27.065 qs_ot_get_derivative 99 11.5 0.001 0.001 20.199 20.321 multiply_cannon_multrec 49320 15.4 11.389 11.974 18.022 18.628 rebuild_ks_matrix 110 8.3 0.000 0.000 14.390 14.522 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.389 14.522 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.619 12.728 mp_waitall_1 220248 16.4 11.000 12.104 11.000 12.104 multiply_cannon_sync_h2d 49320 15.4 9.574 10.125 9.574 10.125 qs_ot_get_p 110 10.4 0.001 0.001 9.712 9.843 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.683 8.238 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.312 7.810 apply_single 110 13.6 0.000 0.000 7.312 7.810 multiply_cannon_metrocomm3 49320 15.4 0.085 0.088 6.688 7.769 init_scf_run 11 5.9 0.000 0.001 7.124 7.124 scf_env_initial_rho_setup 11 6.9 0.000 0.001 7.124 7.124 sum_up_and_integrate 110 10.3 0.002 0.003 6.927 6.940 integrate_v_rspace 110 11.3 0.003 0.004 6.902 6.921 dbcsr_mm_accdrv_process 87628 16.1 3.053 3.162 6.503 6.781 ot_diis_step 99 11.5 0.006 0.007 6.574 6.574 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.486 6.541 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.253 6.380 calculate_rho_elec 110 8.6 0.021 0.025 6.252 6.380 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.584 5.584 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.355 5.409 init_scf_loop 11 6.9 0.000 0.001 5.342 5.342 mp_sum_l 6594 12.7 4.191 5.147 4.191 5.147 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.027 5.053 cp_fm_diag_elpa_base 48 14.0 5.014 5.039 5.025 5.052 calculate_dm_sparse 110 9.5 0.001 0.001 4.011 4.155 wfi_extrapolate 11 7.9 0.001 0.001 3.999 3.999 make_m2s 4110 13.4 0.060 0.065 3.893 3.981 make_images 4110 14.4 0.177 0.189 3.798 3.886 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.696 3.701 multiply_cannon_metrocomm1 49320 15.4 0.065 0.067 2.525 3.528 density_rs2pw 110 9.6 0.004 0.005 3.288 3.499 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.355 3.403 grid_integrate_task_list 110 12.3 3.250 3.403 3.250 3.403 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.277 3.343 jit_kernel_multiply 13 15.9 3.174 3.316 3.174 3.316 prepare_preconditioner 11 7.9 0.000 0.000 3.203 3.228 make_preconditioner 11 8.9 0.000 0.000 3.203 3.227 pw_transfer 1331 11.6 0.055 0.069 3.033 3.108 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.995 3.034 calculate_first_density_matrix 1 7.0 0.000 0.004 3.027 3.031 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 2.944 3.023 acc_transpose_blocks 49320 15.4 0.215 0.223 2.662 2.873 fft_wrap_pw1pw2_140 451 13.1 0.390 0.427 2.519 2.602 potential_pw2rs 110 12.3 0.005 0.006 2.525 2.538 mp_alltoall_d11v 2046 13.8 2.036 2.500 2.036 2.500 fft3d_ps 1111 14.6 0.794 0.886 2.240 2.303 grid_collocate_task_list 110 9.6 2.156 2.296 2.156 2.296 transfer_rs2pw 451 10.6 0.005 0.006 1.872 2.008 mp_waitany 14300 13.8 1.834 1.977 1.834 1.977 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.917 1.944 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.900 1.924 make_images_data 4110 15.4 0.043 0.046 1.763 1.880 mp_sum_d 3889 11.9 1.351 1.837 1.351 1.837 hybrid_alltoall_any 4261 16.3 0.084 0.486 1.519 1.796 cp_fm_cholesky_invert 11 10.9 1.756 1.761 1.756 1.761 transfer_pw2rs 451 13.1 0.006 0.007 1.703 1.715 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=82.699000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.000000, yerr=2.044949 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1e516dface89e0648f7d6ff6212138fc2d72d23d_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 591.601664E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1972800 MPI messages size (bytes): total size 1.077520E+12 min size 0.000000E+00 max size 4.537280E+06 average size 546.188250E+03 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 222984 1826684928 8192 < size <= 32768 520356 13399818240 32768 < size <= 131072 372336 35386294272 131072 < size <= 4194304 787758 788321309808 4194304 < size <= 16777216 54450 238588003280 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66430. MP_Allreduce 9775 566. MP_Sync 52 MP_Alltoall 1717 3188570. 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 0.007 0.028 71.105 71.106 qs_mol_dyn_low 1 2.0 0.003 0.003 70.805 70.815 qs_forces 11 3.9 0.003 0.003 69.646 69.646 qs_energies 11 4.9 0.003 0.005 66.312 66.316 scf_env_do_scf 11 5.9 0.000 0.001 57.805 57.808 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 50.073 50.074 dbcsr_multiply_generic 2055 12.4 0.116 0.119 38.584 38.776 velocity_verlet 10 3.0 0.006 0.023 37.597 37.621 qs_scf_new_mos 99 7.5 0.001 0.001 34.055 34.203 qs_scf_loop_do_ot 99 8.5 0.001 0.001 34.055 34.203 multiply_cannon 2055 13.4 0.225 0.242 31.680 32.763 ot_scf_mini 99 9.5 0.003 0.003 32.372 32.497 multiply_cannon_loop 2055 14.4 1.168 1.197 30.352 31.273 ot_mini 99 10.5 0.001 0.001 18.983 19.129 multiply_cannon_multrec 24660 15.4 6.946 8.674 14.154 16.051 rebuild_ks_matrix 110 8.3 0.000 0.000 13.632 13.732 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 13.632 13.731 qs_ot_get_derivative 99 11.5 0.001 0.001 13.147 13.277 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.996 12.085 mp_waitall_1 176588 16.5 7.747 10.239 7.747 10.239 multiply_cannon_metrocomm3 24660 15.4 0.073 0.075 5.267 7.774 init_scf_loop 11 6.9 0.000 0.000 7.693 7.694 multiply_cannon_sync_h2d 24660 15.4 6.363 7.403 6.363 7.403 dbcsr_mm_accdrv_process 52282 16.1 5.544 6.116 7.038 7.362 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.557 7.195 apply_single 110 13.6 0.000 0.001 6.557 7.195 qs_ot_get_p 110 10.4 0.001 0.001 6.674 6.824 sum_up_and_integrate 110 10.3 0.001 0.003 6.415 6.429 integrate_v_rspace 110 11.3 0.003 0.003 6.388 6.401 init_scf_run 11 5.9 0.000 0.001 6.107 6.108 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.107 6.108 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.954 5.965 calculate_rho_elec 110 8.6 0.039 0.047 5.954 5.965 ot_diis_step 99 11.5 0.010 0.010 5.782 5.782 prepare_preconditioner 11 7.9 0.000 0.000 5.678 5.693 make_preconditioner 11 8.9 0.000 0.000 5.678 5.693 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.853 5.614 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.252 5.406 make_m2s 4110 13.4 0.057 0.060 4.356 4.820 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.727 4.748 make_images 4110 14.4 0.408 0.462 4.245 4.705 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.255 4.255 pw_transfer 1331 11.6 0.066 0.073 3.610 3.753 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.668 3.682 cp_fm_diag_elpa_base 48 14.0 3.619 3.637 3.665 3.679 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.503 3.650 wfi_extrapolate 11 7.9 0.001 0.001 3.515 3.515 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.367 3.449 density_rs2pw 110 9.6 0.004 0.005 3.183 3.362 grid_integrate_task_list 110 12.3 3.171 3.309 3.171 3.309 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.257 3.258 fft_wrap_pw1pw2_140 451 13.1 0.459 0.476 2.961 3.111 calculate_dm_sparse 110 9.5 0.001 0.001 3.016 3.047 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.917 2.957 make_images_data 4110 15.4 0.048 0.052 2.362 2.855 hybrid_alltoall_any 4261 16.3 0.106 0.453 2.080 2.831 fft3d_ps 1111 14.6 1.110 1.334 2.602 2.749 mp_sum_l 6594 12.7 1.914 2.709 1.914 2.709 cp_fm_cholesky_invert 11 10.9 2.657 2.664 2.657 2.664 calculate_first_density_matrix 1 7.0 0.000 0.000 2.494 2.497 grid_collocate_task_list 110 9.6 2.170 2.325 2.170 2.325 potential_pw2rs 110 12.3 0.008 0.009 2.282 2.293 acc_transpose_blocks 24660 15.4 0.117 0.121 2.027 2.072 mp_alltoall_d11v 2046 13.8 1.819 2.029 1.819 2.029 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.996 2.011 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.862 1.863 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.794 1.807 jit_kernel_multiply 8 16.4 1.131 1.756 1.131 1.756 mp_allgather_i34 2055 14.4 0.637 1.717 0.637 1.717 multiply_cannon_metrocomm4 22605 15.4 0.077 0.083 0.786 1.653 transfer_rs2pw 451 10.6 0.007 0.007 1.445 1.595 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.573 1.583 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.388 1.491 mp_irecv_dv 57340 16.2 0.660 1.436 0.660 1.436 dbcsr_complete_redistribute 325 12.2 0.238 0.295 1.167 1.433 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=71.106000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=560.363636, yerr=6.168435 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1e516dface89e0648f7d6ff6212138fc2d72d23d_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 666.914816E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66421. MP_Allreduce 9774 562. MP_Sync 52 MP_Alltoall 1496 4511006. MP_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.008 0.028 60.135 60.135 qs_mol_dyn_low 1 2.0 0.003 0.003 59.853 59.896 qs_forces 11 3.9 0.003 0.003 59.774 59.775 qs_energies 11 4.9 0.001 0.001 56.565 56.569 scf_env_do_scf 11 5.9 0.000 0.001 48.535 48.535 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 40.156 40.156 velocity_verlet 10 3.0 0.001 0.002 32.193 32.195 dbcsr_multiply_generic 2055 12.4 0.107 0.112 29.490 29.792 qs_scf_new_mos 99 7.5 0.001 0.001 25.717 25.850 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.717 25.849 ot_scf_mini 99 9.5 0.002 0.002 24.475 24.631 multiply_cannon 2055 13.4 0.213 0.225 23.281 24.394 multiply_cannon_loop 2055 14.4 0.814 0.832 22.253 23.293 ot_mini 99 10.5 0.001 0.001 14.386 14.548 rebuild_ks_matrix 110 8.3 0.000 0.000 11.992 12.164 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.992 12.163 multiply_cannon_multrec 16440 15.4 3.637 4.792 10.422 11.441 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.535 10.695 qs_ot_get_derivative 99 11.5 0.001 0.001 9.959 10.116 mp_waitall_1 139946 16.5 6.866 9.875 6.866 9.875 init_scf_loop 11 6.9 0.000 0.000 8.345 8.347 multiply_cannon_metrocomm3 16440 15.4 0.046 0.048 4.417 7.275 dbcsr_mm_accdrv_process 34862 16.1 5.788 6.317 6.629 6.747 prepare_preconditioner 11 7.9 0.000 0.000 6.629 6.649 make_preconditioner 11 8.9 0.000 0.000 6.629 6.649 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.965 6.325 sum_up_and_integrate 110 10.3 0.001 0.002 6.209 6.225 integrate_v_rspace 110 11.3 0.003 0.003 6.183 6.199 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.660 5.671 calculate_rho_elec 110 8.6 0.059 0.059 5.659 5.671 init_scf_run 11 5.9 0.000 0.001 5.644 5.644 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.643 5.644 qs_ot_get_p 110 10.4 0.001 0.001 5.309 5.488 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.996 5.454 apply_single 110 13.6 0.000 0.000 4.996 5.454 ot_diis_step 99 11.5 0.011 0.011 4.397 4.397 make_m2s 4110 13.4 0.049 0.051 4.057 4.318 make_images 4110 14.4 0.399 0.519 3.942 4.203 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.233 3.945 multiply_cannon_sync_h2d 16440 15.4 3.236 3.729 3.236 3.729 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.694 3.697 pw_transfer 1331 11.6 0.065 0.073 3.527 3.535 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.420 3.430 grid_integrate_task_list 110 12.3 3.199 3.428 3.199 3.428 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.329 3.330 density_rs2pw 110 9.6 0.004 0.004 2.910 3.160 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.023 3.025 fft_wrap_pw1pw2_140 451 13.1 0.580 0.589 2.962 2.973 wfi_extrapolate 11 7.9 0.001 0.001 2.937 2.937 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.800 2.810 cp_fm_diag_elpa_base 48 14.0 2.736 2.767 2.798 2.808 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.699 2.749 calculate_first_density_matrix 1 7.0 0.000 0.000 2.616 2.618 make_images_data 4110 15.4 0.045 0.050 2.259 2.610 hybrid_alltoall_any 4261 16.3 0.109 0.383 2.005 2.605 cp_fm_cholesky_invert 11 10.9 2.561 2.567 2.561 2.567 calculate_dm_sparse 110 9.5 0.001 0.001 2.511 2.538 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.409 2.493 multiply_cannon_metrocomm4 14385 15.4 0.048 0.052 0.891 2.483 grid_collocate_task_list 110 9.6 2.222 2.393 2.222 2.393 fft3d_ps 1111 14.6 1.089 1.099 2.362 2.378 mp_irecv_dv 48980 15.7 0.815 2.344 0.815 2.344 mp_sum_l 6594 12.7 1.529 2.156 1.529 2.156 potential_pw2rs 110 12.3 0.011 0.012 2.074 2.078 mp_alltoall_d11v 2046 13.8 1.709 2.047 1.709 2.047 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.954 1.955 acc_transpose_blocks 16440 15.4 0.078 0.079 1.879 1.923 dbcsr_complete_redistribute 325 12.2 0.323 0.371 1.413 1.876 cp_fm_upper_to_full 70 14.2 1.403 1.775 1.403 1.775 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.650 1.665 cp_fm_cholesky_decompose 22 10.9 1.542 1.565 1.542 1.565 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.368 1.515 transfer_rs2pw 451 10.6 0.005 0.006 1.241 1.498 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.454 1.465 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.948 1.403 mp_waitany 17072 13.8 1.108 1.347 1.108 1.347 mp_allgather_i34 2055 14.4 0.342 1.332 0.342 1.332 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.274 1.279 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=60.135000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=631.545455, yerr=8.585015 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1e516dface89e0648f7d6ff6212138fc2d72d23d_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 738.992128E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66419. MP_Allreduce 9774 603. MP_Sync 52 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.008 0.028 67.374 67.375 qs_mol_dyn_low 1 2.0 0.003 0.003 67.165 67.175 qs_forces 11 3.9 0.003 0.003 65.796 65.797 qs_energies 11 4.9 0.001 0.001 62.416 62.419 scf_env_do_scf 11 5.9 0.000 0.001 53.973 53.975 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 42.417 42.418 velocity_verlet 10 3.0 0.016 0.021 38.383 38.440 dbcsr_multiply_generic 2055 12.4 0.114 0.120 31.230 31.498 qs_scf_new_mos 99 7.5 0.001 0.001 27.933 28.041 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.933 28.040 ot_scf_mini 99 9.5 0.003 0.003 26.246 26.352 multiply_cannon 2055 13.4 0.238 0.253 24.272 25.286 multiply_cannon_loop 2055 14.4 1.407 1.456 22.875 23.545 ot_mini 99 10.5 0.001 0.001 15.423 15.556 multiply_cannon_multrec 24660 15.4 4.184 7.149 13.392 14.597 rebuild_ks_matrix 110 8.3 0.000 0.000 11.862 11.967 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.862 11.967 init_scf_loop 11 6.9 0.000 0.000 11.515 11.515 qs_ot_get_derivative 99 11.5 0.001 0.001 11.227 11.339 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.475 10.569 dbcsr_mm_accdrv_process 52304 16.0 7.884 9.129 9.052 10.147 prepare_preconditioner 11 7.9 0.000 0.000 9.795 9.808 make_preconditioner 11 8.9 0.000 0.000 9.795 9.807 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.022 9.439 mp_waitall_1 121746 16.5 4.339 6.401 4.339 6.401 sum_up_and_integrate 110 10.3 0.001 0.002 6.177 6.190 integrate_v_rspace 110 11.3 0.003 0.003 6.151 6.164 qs_ot_get_p 110 10.4 0.001 0.001 5.725 5.894 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.772 5.781 calculate_rho_elec 110 8.6 0.078 0.081 5.772 5.780 make_m2s 4110 13.4 0.060 0.062 5.344 5.681 init_scf_run 11 5.9 0.000 0.001 5.580 5.580 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.580 5.580 make_images 4110 14.4 0.577 0.703 5.203 5.536 cp_fm_upper_to_full 70 14.2 3.313 4.774 3.313 4.774 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.063 4.174 apply_single 110 13.6 0.000 0.000 4.063 4.174 ot_diis_step 99 11.5 0.011 0.011 4.157 4.158 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.891 3.907 dbcsr_complete_redistribute 325 12.2 0.422 0.461 2.638 3.744 pw_transfer 1331 11.6 0.065 0.075 3.610 3.645 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.589 3.644 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.504 3.543 grid_integrate_task_list 110 12.3 3.298 3.415 3.298 3.415 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.301 3.360 multiply_cannon_metrocomm3 24660 15.4 0.038 0.039 1.579 3.348 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.332 3.332 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.117 3.213 density_rs2pw 110 9.6 0.004 0.004 2.911 3.150 fft_wrap_pw1pw2_140 451 13.1 0.610 0.632 3.017 3.060 calculate_dm_sparse 110 9.5 0.001 0.001 2.974 3.008 wfi_extrapolate 11 7.9 0.001 0.001 2.978 2.978 hybrid_alltoall_any 4261 16.3 0.122 0.463 2.280 2.954 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.938 2.941 make_images_data 4110 15.4 0.048 0.052 2.626 2.939 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.764 2.846 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.803 2.812 cp_fm_diag_elpa_base 48 14.0 2.658 2.712 2.801 2.811 mp_alltoall_i22 605 13.7 1.605 2.742 1.605 2.742 cp_fm_cholesky_invert 11 10.9 2.686 2.694 2.686 2.694 acc_transpose_blocks 24660 15.4 0.114 0.117 2.570 2.648 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.585 2.616 multiply_cannon_sync_h2d 24660 15.4 2.365 2.560 2.365 2.560 calculate_first_density_matrix 1 7.0 0.000 0.000 2.507 2.510 grid_collocate_task_list 110 9.6 2.263 2.439 2.263 2.439 fft3d_ps 1111 14.6 1.084 1.115 2.386 2.403 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.234 2.235 potential_pw2rs 110 12.3 0.012 0.013 2.008 2.012 mp_alltoall_d11v 2046 13.8 1.752 1.910 1.752 1.910 jit_kernel_multiply 8 15.9 0.826 1.881 0.826 1.881 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.796 1.834 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.613 1.706 cp_fm_cholesky_decompose 22 10.9 1.615 1.664 1.615 1.664 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.628 1.639 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.532 1.547 mp_sum_l 6594 12.7 0.924 1.530 0.924 1.530 multiply_cannon_metrocomm4 20550 15.4 0.061 0.064 0.876 1.451 transfer_rs2pw 451 10.6 0.005 0.006 1.182 1.442 acc_transpose_blocks_sync 73980 16.4 1.368 1.430 1.368 1.430 mp_sum_dm 438 4.9 1.321 1.405 1.321 1.405 mp_irecv_dv 62702 16.1 0.772 1.370 0.772 1.370 md_write_output 11 3.9 0.055 1.308 0.058 1.356 mp_waitany 13376 13.8 1.028 1.354 1.028 1.354 update_particle_set 20 4.0 0.000 0.000 1.281 1.352 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=67.375000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=698.727273, yerr=9.535493 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1e516dface89e0648f7d6ff6212138fc2d72d23d_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 868.032512E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66417. MP_Allreduce 9774 644. MP_Sync 52 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.028 54.391 54.392 qs_mol_dyn_low 1 2.0 0.003 0.003 54.064 54.075 qs_forces 11 3.9 0.003 0.003 53.985 53.986 qs_energies 11 4.9 0.001 0.001 50.374 50.378 scf_env_do_scf 11 5.9 0.000 0.001 42.184 42.184 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 34.643 34.644 velocity_verlet 10 3.0 0.001 0.001 30.663 30.667 dbcsr_multiply_generic 2055 12.4 0.105 0.109 23.176 23.296 qs_scf_new_mos 99 7.5 0.001 0.001 20.346 20.396 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.345 20.395 ot_scf_mini 99 9.5 0.002 0.002 19.103 19.132 multiply_cannon 2055 13.4 0.237 0.249 17.661 19.099 multiply_cannon_loop 2055 14.4 0.603 0.628 16.405 16.752 rebuild_ks_matrix 110 8.3 0.000 0.000 11.471 11.506 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 11.470 11.505 ot_mini 99 10.5 0.001 0.001 10.647 10.668 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.209 10.240 multiply_cannon_multrec 8220 15.4 3.225 4.352 7.695 8.614 init_scf_loop 11 6.9 0.000 0.000 7.496 7.497 mp_waitall_1 103326 16.6 5.709 7.468 5.709 7.468 qs_ot_get_derivative 99 11.5 0.001 0.001 6.979 7.007 sum_up_and_integrate 110 10.3 0.001 0.002 6.106 6.123 integrate_v_rspace 110 11.3 0.003 0.004 6.079 6.096 prepare_preconditioner 11 7.9 0.000 0.000 5.869 5.872 make_preconditioner 11 8.9 0.000 0.000 5.869 5.872 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.831 5.846 calculate_rho_elec 110 8.6 0.114 0.114 5.830 5.846 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.456 5.526 dbcsr_mm_accdrv_process 17442 15.9 3.086 3.814 4.331 5.250 init_scf_run 11 5.9 0.000 0.001 5.077 5.077 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.077 5.077 qs_ot_get_p 110 10.4 0.001 0.001 4.528 4.541 make_m2s 4110 13.4 0.038 0.039 4.198 4.486 make_images 4110 14.4 0.649 0.711 4.069 4.356 multiply_cannon_metrocomm3 8220 15.4 0.019 0.020 2.922 4.263 pw_transfer 1331 11.6 0.065 0.070 3.759 3.769 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.728 3.759 apply_single 110 13.6 0.000 0.000 3.727 3.759 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.651 3.665 ot_diis_step 99 11.5 0.012 0.012 3.647 3.647 grid_integrate_task_list 110 12.3 3.383 3.509 3.383 3.509 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.202 3.206 fft_wrap_pw1pw2_140 451 13.1 0.776 0.787 3.176 3.192 density_rs2pw 110 9.6 0.004 0.004 2.859 3.046 cp_dbcsr_syevd 48 12.0 0.003 0.003 2.867 2.868 cp_fm_cholesky_invert 11 10.9 2.813 2.817 2.813 2.817 hybrid_alltoall_any 4261 16.3 0.200 0.851 2.209 2.722 wfi_extrapolate 11 7.9 0.001 0.001 2.689 2.689 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.659 2.660 make_images_data 4110 15.4 0.040 0.046 2.241 2.623 grid_collocate_task_list 110 9.6 2.363 2.543 2.363 2.543 calculate_dm_sparse 110 9.5 0.001 0.001 2.500 2.538 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.526 2.527 multiply_cannon_sync_h2d 8220 15.4 2.377 2.516 2.377 2.516 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.368 2.375 cp_fm_diag_elpa_base 48 14.0 2.312 2.333 2.366 2.374 fft3d_ps 1111 14.6 1.128 1.144 2.319 2.330 calculate_first_density_matrix 1 7.0 0.000 0.000 2.293 2.294 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.171 2.189 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.053 2.088 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.784 1.991 potential_pw2rs 110 12.3 0.015 0.015 1.917 1.919 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.884 1.900 mp_alltoall_d11v 2046 13.8 1.577 1.735 1.577 1.735 cp_fm_cholesky_decompose 22 10.9 1.637 1.669 1.637 1.669 qs_env_update_s_mstruct 11 6.9 0.000 0.001 1.487 1.605 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.590 1.596 mp_allgather_i34 2055 14.4 0.429 1.528 0.429 1.528 dbcsr_complete_redistribute 325 12.2 0.560 0.580 1.397 1.495 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.460 1.472 jit_kernel_multiply 8 15.6 0.932 1.459 0.932 1.459 acc_transpose_blocks 8220 15.4 0.039 0.041 1.386 1.423 qs_create_task_list 11 7.9 0.000 0.000 1.209 1.310 generate_qs_task_list 11 8.9 0.374 0.443 1.209 1.310 transfer_rs2pw 451 10.6 0.005 0.005 1.091 1.290 mp_waitany 9240 13.8 1.040 1.240 1.040 1.240 multiply_cannon_metrocomm1 8220 15.4 0.022 0.023 0.737 1.145 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.107 1.136 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=54.392000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=817.727273, yerr=12.671306 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1e516dface89e0648f7d6ff6212138fc2d72d23d_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 1.384071E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67098. MP_Allreduce 9752 812. MP_Sync 52 MP_Alltoall 1474 16505187. MP_SendRecv 2310 360267. MP_ISendRecv 2310 360267. MP_Wait 5214 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.038 0.051 87.276 87.277 qs_mol_dyn_low 1 2.0 0.003 0.003 86.963 86.972 qs_forces 11 3.9 0.003 0.004 86.892 86.893 qs_energies 11 4.9 0.001 0.001 82.801 82.803 scf_env_do_scf 11 5.9 0.001 0.001 72.420 72.420 velocity_verlet 10 3.0 0.001 0.002 54.983 54.989 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 44.510 44.511 dbcsr_multiply_generic 2055 12.4 0.119 0.125 30.659 30.761 init_scf_loop 11 6.9 0.000 0.000 27.838 27.841 qs_scf_new_mos 99 7.5 0.001 0.001 27.429 27.461 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.428 27.460 prepare_preconditioner 11 7.9 0.000 0.000 25.866 25.873 make_preconditioner 11 8.9 0.000 0.000 25.866 25.873 ot_scf_mini 99 9.5 0.002 0.002 25.639 25.664 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.181 25.359 multiply_cannon 2055 13.4 0.333 0.355 23.253 24.007 multiply_cannon_loop 2055 14.4 0.828 0.854 21.480 22.160 cp_fm_upper_to_full 70 14.2 12.598 18.023 12.598 18.023 ot_mini 99 10.5 0.001 0.001 14.859 14.887 rebuild_ks_matrix 110 8.3 0.000 0.001 13.443 13.477 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 13.442 13.476 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.198 12.229 multiply_cannon_multrec 8220 15.4 4.404 4.583 10.589 10.827 dbcsr_complete_redistribute 325 12.2 1.015 1.048 7.238 10.434 qs_ot_get_derivative 99 11.5 0.001 0.001 10.306 10.333 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.230 9.420 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.669 8.834 mp_alltoall_i22 605 13.7 5.299 8.473 5.299 8.473 mp_waitall_1 84994 16.7 7.398 8.217 7.398 8.217 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.006 7.043 calculate_rho_elec 110 8.6 0.224 0.225 7.006 7.043 sum_up_and_integrate 110 10.3 0.002 0.002 6.639 6.655 integrate_v_rspace 110 11.3 0.004 0.004 6.610 6.627 dbcsr_mm_accdrv_process 11614 15.7 4.082 4.554 6.037 6.218 init_scf_run 11 5.9 0.000 0.001 6.125 6.125 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.125 6.125 make_m2s 4110 13.4 0.043 0.044 5.279 5.772 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.092 5.614 apply_single 110 13.6 0.000 0.000 5.092 5.614 make_images 4110 14.4 0.885 0.940 5.091 5.582 qs_ot_get_p 110 10.4 0.001 0.001 5.517 5.547 multiply_cannon_metrocomm3 8220 15.4 0.019 0.020 4.812 5.209 cp_fm_cholesky_invert 11 10.9 5.122 5.127 5.122 5.127 pw_transfer 1331 11.6 0.074 0.075 4.868 4.874 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 4.752 4.757 ot_diis_step 99 11.5 0.015 0.016 4.536 4.536 fft_wrap_pw1pw2_140 451 13.1 1.280 1.286 4.154 4.161 qs_ot_p2m_diag 48 11.0 0.151 0.156 3.985 3.993 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.040 3.885 grid_integrate_task_list 110 12.3 3.670 3.737 3.670 3.737 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.650 3.651 calculate_dm_sparse 110 9.5 0.001 0.001 3.622 3.649 hybrid_alltoall_any 4261 16.3 0.262 0.561 2.784 3.514 density_rs2pw 110 9.6 0.004 0.004 3.481 3.490 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.463 3.464 make_images_data 4110 15.4 0.045 0.048 2.759 3.366 wfi_extrapolate 11 7.9 0.001 0.001 3.280 3.280 multiply_cannon_sync_h2d 8220 15.4 3.118 3.154 3.118 3.154 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.902 2.905 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.877 2.877 cp_fm_diag_elpa_base 48 14.0 2.341 2.539 2.876 2.876 fft3d_ps 1111 14.6 1.300 1.319 2.804 2.809 calculate_first_density_matrix 1 7.0 0.000 0.000 2.733 2.733 grid_collocate_task_list 110 9.6 2.663 2.672 2.663 2.672 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.638 2.654 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.410 2.422 qs_env_update_s_mstruct 11 6.9 0.001 0.001 2.177 2.225 potential_pw2rs 110 12.3 0.021 0.021 2.184 2.186 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.102 2.184 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 2.130 2.148 mp_alltoall_d11v 2046 13.8 1.986 2.045 1.986 2.045 cp_fm_cholesky_decompose 22 10.9 1.940 1.955 1.940 1.955 qs_create_task_list 11 7.9 0.000 0.000 1.888 1.932 generate_qs_task_list 11 8.9 0.735 0.795 1.887 1.932 jit_kernel_multiply 8 15.1 1.760 1.870 1.760 1.870 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.842 1.846 acc_transpose_blocks 8220 15.4 0.041 0.042 1.774 1.815 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.730 1.766 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=87.277000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1252.636364, yerr=53.087017 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1e516dface89e0648f7d6ff6212138fc2d72d23d_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 1.103326E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 12044928 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.1 marketing flops 145.647559E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 630.874112E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10431456 MPI messages size (bytes): total size 4.526012E+12 min size 0.000000E+00 max size 4.537280E+06 average size 433.881156E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3605352 96403587072 32768 < size <= 131072 1305088 74666999808 131072 < size <= 4194304 5190114 3200148350408 4194304 < size <= 16777216 263934 1154745839544 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4024 57450. MP_Allreduce 11139 794. MP_Sync 87 MP_Alltoall 2242 2314331. MP_SendRecv 24510 18752. MP_ISendRecv 24510 18752. MP_Wait 42802 MP_ISend 16140 108019. MP_IRecv 16140 108019. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.077 211.904 211.905 qs_mol_dyn_low 1 2.0 0.004 0.009 211.404 211.430 qs_forces 11 3.9 0.014 0.075 211.271 211.274 qs_energies 11 4.9 0.002 0.010 205.571 205.588 scf_env_do_scf 11 5.9 0.001 0.001 188.664 188.668 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 167.645 167.647 dbcsr_multiply_generic 2527 12.6 0.182 0.186 127.428 127.897 qs_scf_new_mos 118 7.6 0.001 0.001 127.220 127.570 qs_scf_loop_do_ot 118 8.6 0.001 0.001 127.220 127.569 velocity_verlet 10 3.0 0.001 0.001 126.572 126.574 ot_scf_mini 118 9.6 0.003 0.004 120.424 120.772 multiply_cannon 2527 13.6 0.244 0.252 102.932 104.349 multiply_cannon_loop 2527 14.6 2.417 2.487 100.684 102.123 ot_mini 118 10.6 0.001 0.001 67.420 67.769 multiply_cannon_multrec 60648 15.6 31.985 33.879 41.895 44.061 qs_ot_get_derivative 118 11.6 0.001 0.002 42.356 42.680 rebuild_ks_matrix 129 8.3 0.001 0.001 34.169 34.477 qs_ks_build_kohn_sham_matrix 129 9.3 0.015 0.018 34.169 34.477 mp_waitall_1 269254 16.5 29.838 33.565 29.838 33.565 qs_ks_update_qs_env 129 7.6 0.001 0.001 30.727 31.019 qs_ot_get_p 129 10.4 0.001 0.001 30.580 30.995 multiply_cannon_sync_h2d 60648 15.6 26.520 28.043 26.520 28.043 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 24.594 25.097 apply_single 129 13.6 0.001 0.001 24.593 25.097 ot_diis_step 118 11.6 0.008 0.010 24.805 24.807 qs_ot_p2m_diag 83 11.4 0.079 0.092 23.756 23.831 cp_dbcsr_syevd 83 12.4 0.005 0.005 20.984 20.985 init_scf_loop 11 6.9 0.000 0.000 20.944 20.946 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.898 20.167 multiply_cannon_metrocomm3 60648 15.6 0.120 0.127 16.294 18.443 cp_fm_diag_elpa 83 13.4 0.000 0.000 17.900 17.931 cp_fm_diag_elpa_base 83 14.4 17.826 17.864 17.896 17.928 prepare_preconditioner 11 7.9 0.000 0.000 16.307 16.355 make_preconditioner 11 8.9 0.000 0.000 16.307 16.355 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.556 15.756 make_m2s 5054 13.6 0.105 0.113 14.167 14.554 sum_up_and_integrate 129 10.3 0.002 0.006 14.413 14.432 make_images 5054 14.6 0.400 0.419 13.982 14.378 integrate_v_rspace 129 11.3 0.004 0.012 14.355 14.373 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.603 13.755 calculate_rho_elec 129 8.7 0.045 0.064 13.603 13.754 init_scf_run 11 5.9 0.000 0.001 12.651 12.652 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.651 12.651 mp_sum_l 8010 12.9 9.063 10.456 9.063 10.456 dbcsr_mm_accdrv_process 125468 16.2 4.712 4.935 9.483 10.199 cp_fm_cholesky_invert 11 10.9 9.371 9.379 9.371 9.379 wfi_extrapolate 11 7.9 0.001 0.001 9.192 9.192 calculate_dm_sparse 129 9.5 0.001 0.001 8.866 8.961 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 8.716 8.935 qs_ot_get_orbitals 118 10.6 0.001 0.001 8.178 8.273 multiply_cannon_metrocomm1 60648 15.6 0.095 0.099 6.606 8.168 pw_transfer 1559 11.6 0.075 0.094 7.765 7.990 make_images_data 5054 15.6 0.067 0.072 6.980 7.965 density_rs2pw 129 9.7 0.006 0.007 7.129 7.859 fft_wrap_pw1pw2 1301 12.7 0.010 0.013 7.561 7.790 grid_integrate_task_list 129 12.3 7.130 7.563 7.130 7.563 hybrid_alltoall_any 5240 16.5 0.298 2.298 6.084 7.508 fft_wrap_pw1pw2_140 527 13.2 1.131 1.184 6.602 6.847 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.693 6.708 fft3d_ps 1301 14.7 2.208 2.871 5.557 5.928 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.772 5.908 mp_alltoall_d11v 2423 14.1 4.494 5.787 4.494 5.787 grid_collocate_task_list 129 9.7 4.869 5.288 4.869 5.288 potential_pw2rs 129 12.3 0.009 0.010 4.806 4.854 cp_fm_cholesky_decompose 22 10.9 4.776 4.790 4.776 4.790 mp_sum_d 4499 12.1 3.594 4.338 3.594 4.338 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=211.905000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=596.818182, yerr=6.132153 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1e516dface89e0648f7d6ff6212138fc2d72d23d_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 846.577664E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100942E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834332048 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58357. MP_Allreduce 11058 960. MP_Sync 87 MP_Alltoall 1969 6536660. 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 0.009 0.029 188.992 188.994 qs_mol_dyn_low 1 2.0 0.003 0.003 188.563 188.602 qs_forces 11 3.9 0.004 0.005 188.448 188.449 qs_energies 11 4.9 0.001 0.002 181.711 181.722 scf_env_do_scf 11 5.9 0.001 0.001 164.943 164.953 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 132.444 132.446 velocity_verlet 10 3.0 0.001 0.001 118.739 118.741 dbcsr_multiply_generic 2507 12.6 0.190 0.196 97.739 98.946 qs_scf_new_mos 117 7.6 0.001 0.001 93.742 94.321 qs_scf_loop_do_ot 117 8.6 0.001 0.001 93.741 94.320 ot_scf_mini 117 9.6 0.004 0.004 88.916 89.552 multiply_cannon 2507 13.6 0.503 0.562 77.689 82.338 multiply_cannon_loop 2507 14.6 1.578 1.643 74.338 77.142 ot_mini 117 10.6 0.001 0.001 50.199 50.758 mp_waitall_1 214728 16.6 23.871 38.002 23.871 38.002 multiply_cannon_multrec 30084 15.6 21.042 25.587 31.913 36.967 rebuild_ks_matrix 128 8.3 0.001 0.001 32.063 32.785 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.019 32.062 32.784 init_scf_loop 11 6.9 0.000 0.000 32.411 32.412 qs_ks_update_qs_env 128 7.6 0.001 0.001 28.834 29.467 qs_ot_get_derivative 117 11.6 0.001 0.002 28.173 28.801 prepare_preconditioner 11 7.9 0.000 0.000 28.139 28.214 make_preconditioner 11 8.9 0.000 0.000 28.139 28.214 multiply_cannon_metrocomm3 30084 15.6 0.097 0.102 15.565 28.028 make_full_inverse_cholesky 11 9.9 0.000 0.000 26.842 27.361 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.219 23.339 apply_single 128 13.6 0.001 0.001 22.219 23.339 ot_diis_step 117 11.6 0.014 0.015 21.849 21.851 qs_ot_get_p 128 10.4 0.001 0.001 20.760 21.426 multiply_cannon_sync_h2d 30084 15.6 17.946 20.836 17.946 20.836 cp_fm_cholesky_invert 11 10.9 16.483 16.495 16.483 16.495 qs_ot_p2m_diag 83 11.4 0.188 0.217 15.912 15.946 make_m2s 5014 13.6 0.089 0.094 14.094 15.767 make_images 5014 14.6 1.173 1.369 13.880 15.548 cp_dbcsr_syevd 83 12.4 0.006 0.006 14.717 14.718 sum_up_and_integrate 128 10.3 0.002 0.004 13.840 13.869 integrate_v_rspace 128 11.3 0.003 0.004 13.780 13.811 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.251 13.281 calculate_rho_elec 128 8.7 0.087 0.104 13.250 13.281 init_scf_run 11 5.9 0.000 0.001 11.976 11.977 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.976 11.977 cp_fm_diag_elpa 83 13.4 0.000 0.000 11.579 11.614 cp_fm_diag_elpa_base 83 14.4 11.327 11.428 11.573 11.604 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 11.069 11.526 multiply_cannon_metrocomm4 27577 15.6 0.105 0.121 3.851 11.052 dbcsr_mm_accdrv_process 62242 16.2 5.403 6.078 10.324 10.995 mp_irecv_dv 69486 16.3 3.646 10.645 3.646 10.645 make_images_data 5014 15.6 0.066 0.075 8.304 10.419 hybrid_alltoall_any 5200 16.5 0.352 1.535 7.087 9.650 pw_transfer 1547 11.6 0.085 0.097 8.488 8.551 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.265 8.332 wfi_extrapolate 11 7.9 0.001 0.001 8.289 8.289 grid_integrate_task_list 128 12.3 7.149 7.520 7.149 7.520 fft_wrap_pw1pw2_140 523 13.2 1.213 1.232 7.348 7.435 density_rs2pw 128 9.7 0.006 0.007 6.960 7.397 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.287 7.067 cp_fm_cholesky_decompose 22 10.9 6.793 6.885 6.793 6.885 calculate_dm_sparse 128 9.5 0.001 0.001 6.483 6.621 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.525 6.535 mp_sum_l 7950 12.9 4.207 6.086 4.207 6.086 fft3d_ps 1291 14.7 2.841 3.010 5.824 5.871 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.655 5.858 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.385 5.461 grid_collocate_task_list 128 9.7 5.006 5.339 5.006 5.339 mp_allgather_i34 2507 14.6 1.873 5.011 1.873 5.011 mp_alltoall_d11v 2415 14.1 4.070 4.496 4.070 4.496 potential_pw2rs 128 12.3 0.016 0.019 4.393 4.406 mp_sum_d 4465 12.1 2.582 3.975 2.582 3.975 dbcsr_complete_redistribute 395 12.7 0.770 0.842 3.059 3.904 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=188.994000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=805.818182, yerr=3.638636 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1e516dface89e0648f7d6ff6212138fc2d72d23d_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430457200640 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986252263424 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 11613077360640 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239167967232 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239167967232 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 94.233007E+12 0.0% 0.0% 100.0% flops max/rank 2.951022E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383584 0.0% 0.0% 100.0% number of processed stacks 4017216 0.0% 0.0% 100.0% average stack size 0.0 0.0 1694.3 marketing flops 145.650931E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 961.110016E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1052064 MPI messages size (bytes): total size 2.737208E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.601750E+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 284544 37295751168 131072 < size <= 4194304 665856 1004519030784 4194304 < size <= 16777216 66080 937890820720 16777216 < size 28896 757491302400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 57629. MP_Allreduce 11184 996. MP_Sync 88 MP_Alltoall 1724 9394234. MP_SendRecv 7998 75008. MP_ISendRecv 7998 75008. MP_Wait 21986 MP_ISend 11836 275177. MP_IRecv 11836 275177. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.028 175.217 175.217 qs_mol_dyn_low 1 2.0 0.003 0.003 174.602 174.616 qs_forces 11 3.9 0.004 0.004 174.498 174.501 qs_energies 11 4.9 0.007 0.008 167.983 167.991 scf_env_do_scf 11 5.9 0.001 0.001 152.311 152.311 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 117.562 117.563 velocity_verlet 10 3.0 0.001 0.001 112.654 112.656 dbcsr_multiply_generic 2529 12.6 0.182 0.188 82.576 83.680 qs_scf_new_mos 118 7.6 0.001 0.001 80.450 80.781 qs_scf_loop_do_ot 118 8.6 0.001 0.001 80.449 80.780 ot_scf_mini 118 9.6 0.003 0.004 76.180 76.546 multiply_cannon 2529 13.6 0.517 0.534 63.143 67.586 multiply_cannon_loop 2529 14.6 1.139 1.196 59.893 62.662 ot_mini 118 10.6 0.001 0.001 42.684 43.059 init_scf_loop 11 6.9 0.000 0.000 34.652 34.653 mp_waitall_1 172006 16.6 24.386 33.291 24.386 33.291 prepare_preconditioner 11 7.9 0.000 0.000 30.683 30.737 make_preconditioner 11 8.9 0.000 0.000 30.683 30.737 rebuild_ks_matrix 129 8.3 0.001 0.001 30.232 30.696 qs_ks_build_kohn_sham_matrix 129 9.3 0.016 0.017 30.232 30.695 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.293 29.748 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.277 27.690 multiply_cannon_multrec 20232 15.6 13.248 16.464 22.606 25.676 multiply_cannon_metrocomm3 20232 15.6 0.063 0.067 15.378 24.605 qs_ot_get_derivative 118 11.6 0.002 0.002 23.062 23.432 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.745 20.692 apply_single 129 13.6 0.001 0.001 19.745 20.691 ot_diis_step 118 11.6 0.018 0.018 19.505 19.505 qs_ot_get_p 129 10.4 0.001 0.001 18.712 19.207 make_m2s 5058 13.6 0.079 0.087 14.388 15.808 make_images 5058 14.6 1.160 1.261 14.153 15.567 multiply_cannon_sync_h2d 20232 15.6 13.829 15.080 13.829 15.080 qs_ot_p2m_diag 84 11.4 0.269 0.276 14.438 14.449 cp_fm_cholesky_invert 11 10.9 14.159 14.168 14.159 14.168 sum_up_and_integrate 129 10.3 0.002 0.002 13.863 13.884 integrate_v_rspace 129 11.3 0.003 0.004 13.804 13.826 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.482 13.519 calculate_rho_elec 129 8.7 0.131 0.146 13.481 13.519 cp_dbcsr_syevd 84 12.4 0.005 0.006 13.375 13.381 init_scf_run 11 5.9 0.000 0.001 10.640 10.640 scf_env_initial_rho_setup 11 6.9 0.000 0.001 10.640 10.640 make_images_data 5058 15.6 0.062 0.073 8.511 10.336 cp_fm_diag_elpa 84 13.4 0.000 0.001 10.190 10.209 cp_fm_diag_elpa_base 84 14.4 9.779 9.943 10.187 10.206 hybrid_alltoall_any 5245 16.5 0.452 2.048 7.547 9.627 multiply_cannon_metrocomm4 17703 15.6 0.067 0.078 3.466 9.438 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.108 9.368 mp_irecv_dv 50659 16.2 3.336 9.171 3.336 9.171 dbcsr_mm_accdrv_process 41846 16.2 5.642 5.963 8.814 8.961 pw_transfer 1559 11.6 0.086 0.103 8.629 8.742 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 8.403 8.520 grid_integrate_task_list 129 12.3 7.385 7.867 7.385 7.867 fft_wrap_pw1pw2_140 527 13.2 1.311 1.347 7.490 7.618 wfi_extrapolate 11 7.9 0.001 0.001 7.428 7.428 cp_fm_upper_to_full 106 14.8 5.810 7.351 5.810 7.351 cp_fm_cholesky_decompose 22 10.9 7.276 7.306 7.276 7.306 density_rs2pw 129 9.7 0.006 0.006 6.821 7.281 dbcsr_complete_redistribute 397 12.7 1.167 1.196 4.592 6.446 calculate_dm_sparse 129 9.5 0.001 0.001 5.864 5.975 fft3d_ps 1301 14.7 2.781 3.008 5.761 5.853 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.686 5.691 grid_collocate_task_list 129 9.7 5.197 5.590 5.197 5.590 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.662 5.336 copy_fm_to_dbcsr 210 11.7 0.002 0.002 3.445 5.297 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.786 4.941 mp_sum_l 8016 12.9 3.188 4.914 3.188 4.914 mp_allgather_i34 2529 14.6 1.732 4.813 1.732 4.813 mp_alltoall_d11v 2429 14.1 4.169 4.673 4.169 4.673 potential_pw2rs 129 12.3 0.020 0.022 4.267 4.275 transfer_fm_to_dbcsr 11 9.9 0.018 0.025 2.371 4.147 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.945 3.973 mp_alltoall_i22 720 14.1 2.003 3.967 2.003 3.967 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 3.736 3.737 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.455 3.505 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=175.217000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=906.727273, yerr=12.934035 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1e516dface89e0648f7d6ff6212138fc2d72d23d_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 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796578E+12 0.0% 0.0% 100.0% flops max/rank 4.320337E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500608 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 1.178546E+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 660837763760 16777216 < size 30480 532676608000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4003 58189. MP_Allreduce 11085 1083. MP_Sync 86 MP_Alltoall 1700 12496381. MP_SendRecv 5842 75008. MP_ISendRecv 5842 75008. MP_Wait 22272 MP_ISend 14840 244848. MP_IRecv 14840 244848. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.030 186.821 186.823 qs_mol_dyn_low 1 2.0 0.003 0.003 186.110 186.154 qs_forces 11 3.9 0.004 0.004 185.977 185.987 qs_energies 11 4.9 0.002 0.002 178.904 178.913 scf_env_do_scf 11 5.9 0.001 0.001 161.514 161.523 velocity_verlet 10 3.0 0.019 0.022 122.885 122.888 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 115.706 115.707 dbcsr_multiply_generic 2485 12.5 0.187 0.192 80.686 81.473 qs_scf_new_mos 116 7.6 0.001 0.001 79.983 80.330 qs_scf_loop_do_ot 116 8.6 0.001 0.001 79.983 80.329 ot_scf_mini 116 9.6 0.003 0.004 75.328 75.663 multiply_cannon 2485 13.5 0.550 0.579 55.875 59.010 multiply_cannon_loop 2485 14.5 1.807 1.912 52.182 54.021 init_scf_loop 11 6.9 0.000 0.000 45.688 45.689 ot_mini 116 10.6 0.001 0.001 42.724 43.026 prepare_preconditioner 11 7.9 0.000 0.000 41.636 41.670 make_preconditioner 11 8.9 0.000 0.000 41.636 41.670 make_full_inverse_cholesky 11 9.9 0.011 0.022 35.344 40.298 multiply_cannon_multrec 29820 15.5 13.622 19.255 26.704 31.998 rebuild_ks_matrix 127 8.3 0.001 0.001 28.889 29.177 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.018 28.888 29.177 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.040 26.295 mp_waitall_1 146592 16.7 16.758 26.186 16.758 26.186 qs_ot_get_derivative 116 11.6 0.001 0.002 23.226 23.567 make_m2s 4970 13.5 0.094 0.097 20.477 21.429 make_images 4970 14.5 1.952 2.359 20.172 21.120 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 18.934 19.532 apply_single 127 13.6 0.001 0.001 18.934 19.532 ot_diis_step 116 11.6 0.017 0.018 19.374 19.375 qs_ot_get_p 127 10.4 0.001 0.001 18.818 19.118 cp_fm_upper_to_full 104 14.8 11.359 16.687 11.359 16.687 cp_fm_cholesky_invert 11 10.9 15.833 15.841 15.833 15.841 qs_ot_p2m_diag 82 11.4 0.338 0.385 14.649 14.703 multiply_cannon_metrocomm3 29820 15.5 0.050 0.054 6.231 14.495 sum_up_and_integrate 127 10.3 0.002 0.003 13.759 13.793 integrate_v_rspace 127 11.3 0.003 0.004 13.699 13.735 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.531 13.586 calculate_rho_elec 127 8.7 0.171 0.187 13.530 13.585 cp_dbcsr_syevd 82 12.4 0.005 0.005 13.266 13.267 dbcsr_mm_accdrv_process 61748 16.2 8.395 9.086 12.655 13.175 dbcsr_complete_redistribute 393 12.7 1.495 1.607 8.988 12.703 make_images_data 4970 15.5 0.065 0.071 10.783 12.280 init_scf_run 11 5.9 0.000 0.001 11.418 11.419 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.417 11.419 hybrid_alltoall_any 5155 16.4 0.526 2.207 9.683 11.411 multiply_cannon_sync_h2d 29820 15.5 10.427 11.353 10.427 11.353 copy_fm_to_dbcsr 208 11.6 0.001 0.002 7.620 11.297 cp_fm_diag_elpa 82 13.4 0.000 0.000 10.111 10.129 cp_fm_diag_elpa_base 82 14.4 9.173 9.474 10.104 10.120 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.795 10.034 transfer_fm_to_dbcsr 11 9.9 0.001 0.010 6.269 9.854 mp_alltoall_i22 712 14.1 5.578 9.309 5.578 9.309 pw_transfer 1535 11.6 0.084 0.102 8.773 8.849 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.549 8.631 grid_integrate_task_list 127 12.3 7.479 7.874 7.479 7.874 fft_wrap_pw1pw2_140 519 13.2 1.435 1.465 7.638 7.729 wfi_extrapolate 11 7.9 0.001 0.001 7.668 7.669 cp_fm_cholesky_decompose 22 10.9 7.358 7.439 7.358 7.439 density_rs2pw 127 9.7 0.006 0.006 6.595 7.029 multiply_cannon_metrocomm4 24850 15.5 0.083 0.091 2.769 6.685 calculate_dm_sparse 127 9.5 0.001 0.001 6.340 6.425 mp_irecv_dv 75445 16.2 2.615 6.415 2.615 6.415 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.001 6.063 fft3d_ps 1281 14.7 2.831 2.931 5.748 5.801 grid_collocate_task_list 127 9.7 5.234 5.673 5.234 5.673 mp_alltoall_d11v 2401 14.1 4.979 5.347 4.979 5.347 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.883 4.981 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.495 4.597 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 4.410 4.412 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.246 4.318 potential_pw2rs 127 12.3 0.023 0.023 4.146 4.155 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=186.823000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1096.454545, yerr=26.783182 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1e516dface89e0648f7d6ff6212138fc2d72d23d_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 5.865089E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 1960712 0.0% 0.0% 100.0% average stack size 0.0 0.0 3445.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.530180E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 240672 MPI messages size (bytes): total size 1.331455E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.532238E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 113904 59718500352 4194304 < size <= 16777216 104976 550376570880 16777216 < size 20208 721350232272 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8931 51. MP_Alltoall 9654 799394. MP_ISend 40068 2102573. MP_IRecv 40068 2101676. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4020 57947. MP_Allreduce 11127 1165. MP_Sync 87 MP_Alltoall 1712 18838222. MP_SendRecv 3840 122880. MP_ISendRecv 3840 122880. MP_Wait 16122 MP_ISend 10680 423556. MP_IRecv 10680 423556. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.027 0.054 168.817 168.819 qs_mol_dyn_low 1 2.0 0.003 0.003 168.384 168.424 qs_forces 11 3.9 0.004 0.004 168.232 168.233 qs_energies 11 4.9 0.001 0.002 160.851 160.859 scf_env_do_scf 11 5.9 0.001 0.001 143.675 143.683 velocity_verlet 10 3.0 0.001 0.001 110.876 110.883 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 108.364 108.367 dbcsr_multiply_generic 2507 12.6 0.182 0.193 72.211 72.649 qs_scf_new_mos 117 7.6 0.001 0.001 71.913 72.016 qs_scf_loop_do_ot 117 8.6 0.001 0.001 71.912 72.016 ot_scf_mini 117 9.6 0.003 0.004 67.510 67.641 multiply_cannon 2507 13.6 0.563 0.584 53.785 56.861 multiply_cannon_loop 2507 14.6 0.807 0.845 50.654 51.679 ot_mini 117 10.6 0.001 0.001 37.723 37.824 init_scf_loop 11 6.9 0.000 0.000 35.161 35.163 prepare_preconditioner 11 7.9 0.000 0.000 31.356 31.386 make_preconditioner 11 8.9 0.000 0.000 31.356 31.386 mp_waitall_1 125778 16.7 24.707 31.228 24.707 31.228 make_full_inverse_cholesky 11 9.9 0.017 0.028 29.355 29.623 rebuild_ks_matrix 128 8.3 0.001 0.001 28.785 28.917 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.018 28.784 28.917 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.246 26.373 multiply_cannon_multrec 10028 15.6 10.332 16.019 17.885 22.197 qs_ot_get_derivative 117 11.6 0.001 0.002 20.521 20.645 multiply_cannon_metrocomm3 10028 15.6 0.026 0.027 12.799 20.256 cp_fm_cholesky_invert 11 10.9 17.946 17.952 17.946 17.952 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 17.272 17.419 apply_single 128 13.6 0.001 0.001 17.271 17.418 qs_ot_get_p 128 10.4 0.001 0.001 17.068 17.199 ot_diis_step 117 11.6 0.020 0.020 17.131 17.131 make_m2s 5014 13.6 0.066 0.067 14.597 15.567 make_images 5014 14.6 2.170 2.586 14.289 15.254 sum_up_and_integrate 128 10.3 0.002 0.002 13.993 14.036 integrate_v_rspace 128 11.3 0.004 0.004 13.932 13.978 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.885 13.918 calculate_rho_elec 128 8.7 0.256 0.267 13.885 13.917 qs_ot_p2m_diag 83 11.4 0.497 0.525 13.361 13.396 cp_dbcsr_syevd 83 12.4 0.005 0.006 12.199 12.199 multiply_cannon_sync_h2d 10028 15.6 10.731 11.111 10.731 11.111 init_scf_run 11 5.9 0.000 0.001 10.391 10.392 scf_env_initial_rho_setup 11 6.9 0.000 0.001 10.391 10.391 make_images_data 5014 15.6 0.055 0.065 8.350 9.746 hybrid_alltoall_any 5200 16.5 0.843 3.779 8.194 9.349 cp_fm_diag_elpa 83 13.4 0.000 0.000 9.314 9.325 cp_fm_diag_elpa_base 83 14.4 9.077 9.152 9.311 9.321 pw_transfer 1547 11.6 0.085 0.094 9.190 9.214 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.967 9.001 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 8.017 8.112 fft_wrap_pw1pw2_140 523 13.2 1.805 1.842 7.961 7.994 grid_integrate_task_list 128 12.3 7.770 7.987 7.770 7.987 cp_fm_cholesky_decompose 22 10.9 7.793 7.942 7.793 7.942 dbcsr_mm_accdrv_process 20762 16.1 3.431 4.448 7.205 7.930 multiply_cannon_metrocomm1 10028 15.6 0.030 0.031 4.344 7.405 wfi_extrapolate 11 7.9 0.001 0.001 7.229 7.229 density_rs2pw 128 9.7 0.006 0.006 6.628 6.945 calculate_dm_sparse 128 9.5 0.001 0.001 6.039 6.120 multiply_cannon_metrocomm4 7521 15.6 0.026 0.030 1.875 5.869 grid_collocate_task_list 128 9.7 5.570 5.811 5.570 5.811 mp_irecv_dv 28860 15.9 1.835 5.777 1.835 5.777 fft3d_ps 1291 14.7 2.770 2.837 5.694 5.724 dbcsr_complete_redistribute 395 12.7 2.102 2.136 5.045 5.457 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.261 5.267 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.233 5.235 mp_alltoall_d11v 2415 14.1 4.426 4.682 4.426 4.682 mp_allgather_i34 2507 14.6 1.319 4.606 1.319 4.606 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.206 4.259 potential_pw2rs 128 12.3 0.026 0.027 4.170 4.175 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.584 3.900 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.675 3.755 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.657 3.711 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.429 3.560 copy_fm_to_dbcsr 209 11.7 0.001 0.002 3.221 3.547 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.438 3.453 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=168.819000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1454.909091, yerr=5.712101 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1e516dface89e0648f7d6ff6212138fc2d72d23d_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 1975684956160 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 1992006770688 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 15239182565376 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239182565376 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 94.245913E+12 0.0% 0.0% 100.0% flops max/rank 11.787674E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806580192 0.0% 0.0% 100.0% number of processed stacks 1982496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3433.3 marketing flops 145.663816E+12 ------------------------------------------------------------------------------- # multiplications 2535 max memory usage/rank 3.182998E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101400 MPI messages size (bytes): total size 1.145171E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.293599E+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 45888 35634806784 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592540656 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4057 58363. MP_Allreduce 11219 1499. 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 0.026 0.053 287.414 287.416 qs_mol_dyn_low 1 2.0 0.003 0.003 286.881 286.902 qs_forces 11 3.9 0.005 0.005 286.787 286.790 qs_energies 11 4.9 0.002 0.002 278.046 278.056 scf_env_do_scf 11 5.9 0.001 0.001 255.754 255.758 velocity_verlet 10 3.0 0.002 0.002 208.816 208.824 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 132.719 132.721 init_scf_loop 11 6.9 0.000 0.000 122.758 122.764 prepare_preconditioner 11 7.9 0.000 0.000 118.002 118.019 make_preconditioner 11 8.9 0.000 0.000 118.002 118.019 make_full_inverse_cholesky 11 9.9 0.038 0.039 94.774 115.185 qs_scf_new_mos 118 7.6 0.001 0.001 89.278 89.327 qs_scf_loop_do_ot 118 8.6 0.001 0.001 89.277 89.327 ot_scf_mini 118 9.6 0.004 0.004 84.381 84.391 dbcsr_multiply_generic 2535 12.6 0.234 0.245 82.084 82.492 cp_fm_upper_to_full 106 14.8 52.929 75.097 52.929 75.097 multiply_cannon 2535 13.6 0.699 0.747 59.142 60.728 multiply_cannon_loop 2535 14.6 1.062 1.092 55.093 56.285 ot_mini 118 10.6 0.001 0.001 44.113 44.119 dbcsr_complete_redistribute 397 12.7 3.996 4.038 29.113 41.327 copy_fm_to_dbcsr 210 11.7 0.001 0.002 25.641 37.842 transfer_fm_to_dbcsr 11 9.9 0.030 0.031 23.185 35.201 rebuild_ks_matrix 129 8.3 0.001 0.001 33.921 33.924 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.018 33.921 33.923 mp_alltoall_i22 720 14.1 20.945 33.280 20.945 33.280 cp_fm_cholesky_invert 11 10.9 32.971 32.977 32.971 32.977 mp_waitall_1 104820 16.8 27.848 32.265 27.848 32.265 qs_ks_update_qs_env 129 7.6 0.001 0.001 31.513 31.524 qs_ot_get_p 129 10.4 0.001 0.002 25.044 25.078 qs_ot_get_derivative 118 11.6 0.002 0.002 24.664 24.676 qs_ot_p2m_diag 84 11.4 0.891 0.898 21.017 21.050 multiply_cannon_metrocomm3 10140 15.6 0.025 0.027 18.817 19.978 make_m2s 5070 13.6 0.077 0.078 18.452 19.950 make_images 5070 14.6 3.086 3.283 17.967 19.464 ot_diis_step 118 11.6 0.022 0.022 19.425 19.426 cp_dbcsr_syevd 84 12.4 0.006 0.006 19.211 19.213 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 18.717 18.792 apply_single 129 13.6 0.001 0.001 18.716 18.792 multiply_cannon_multrec 10140 15.6 10.369 12.219 18.205 18.314 qs_rho_update_rho_low 129 7.7 0.001 0.001 16.429 16.437 calculate_rho_elec 129 8.7 0.484 0.485 16.428 16.436 sum_up_and_integrate 129 10.3 0.002 0.002 16.005 16.097 integrate_v_rspace 129 11.3 0.005 0.005 15.942 16.034 cp_fm_diag_elpa 84 13.4 0.000 0.000 16.029 16.029 cp_fm_diag_elpa_base 84 14.4 11.592 13.113 16.025 16.025 multiply_cannon_sync_h2d 10140 15.6 14.408 14.452 14.408 14.452 hybrid_alltoall_any 5257 16.5 1.323 3.092 10.670 12.982 make_images_data 5070 15.6 0.065 0.071 10.383 12.548 init_scf_run 11 5.9 0.000 0.001 12.019 12.020 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.019 12.020 pw_transfer 1559 11.6 0.096 0.097 11.215 11.226 fft_wrap_pw1pw2 1301 12.7 0.012 0.012 10.975 10.986 fft_wrap_pw1pw2_140 527 13.2 3.064 3.117 9.813 9.824 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 9.741 9.749 dbcsr_mm_accdrv_process 20958 16.1 4.213 6.166 7.583 9.451 wfi_extrapolate 11 7.9 0.001 0.001 8.923 8.923 grid_integrate_task_list 129 12.3 8.621 8.815 8.621 8.815 cp_fm_cholesky_decompose 22 10.9 8.780 8.804 8.780 8.804 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 8.137 8.138 density_rs2pw 129 9.7 0.005 0.006 7.398 7.448 mp_alltoall_d11v 2429 14.1 6.811 6.969 6.811 6.969 calculate_dm_sparse 129 9.5 0.001 0.001 6.750 6.805 grid_collocate_task_list 129 9.7 6.463 6.535 6.463 6.535 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.366 6.384 fft3d_ps 1301 14.7 2.877 2.886 6.191 6.226 copy_dbcsr_to_fm 187 11.8 0.004 0.004 6.115 6.152 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=287.416000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2886.000000, yerr=158.525593 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1e516dface89e0648f7d6ff6212138fc2d72d23d_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 1.260143E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255646. MP_Allreduce 3139 6114. MP_Sync 4 MP_Alltoall 54 MP_SendRecv 285 19200. MP_ISendRecv 285 19200. MP_Wait 1017 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.040 84.676 84.677 qs_energies 1 2.0 0.000 0.000 84.192 84.204 ls_scf 1 3.0 0.000 0.000 83.298 83.310 dbcsr_multiply_generic 111 6.7 0.024 0.031 72.342 72.527 multiply_cannon 111 7.7 0.017 0.020 55.731 57.046 multiply_cannon_loop 111 8.7 0.227 0.243 52.368 53.765 ls_scf_main 1 4.0 0.000 0.000 52.125 52.126 density_matrix_trs4 2 5.0 0.002 0.003 46.590 46.659 ls_scf_init_scf 1 4.0 0.000 0.000 28.125 28.126 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.015 27.071 mp_waitall_1 11031 10.9 22.297 25.240 22.297 25.240 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 24.930 24.946 multiply_cannon_multrec 2664 9.7 8.139 8.912 15.527 17.296 multiply_cannon_sync_h2d 2664 9.7 13.582 15.789 13.582 15.789 make_m2s 222 7.7 0.009 0.011 12.974 13.440 make_images 222 8.7 0.098 0.109 12.952 13.419 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.665 12.698 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.397 8.514 make_images_data 222 9.7 0.004 0.005 7.568 8.030 dbcsr_mm_accdrv_process 4760 10.4 0.594 0.712 7.004 7.962 hybrid_alltoall_any 227 10.6 0.216 1.834 6.520 7.610 dbcsr_mm_accdrv_process_sort 4760 11.4 6.212 7.085 6.212 7.085 calculate_norms 4752 9.8 5.496 6.225 5.496 6.225 apply_matrix_preconditioner 6 5.3 0.001 0.003 5.002 5.138 mp_sum_l 887 5.1 3.087 4.767 3.087 4.767 make_images_sizes 222 9.7 0.000 0.000 0.673 3.737 mp_alltoall_i44 222 10.7 0.673 3.736 0.673 3.736 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.345 3.690 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.036 3.432 mp_irecv_dv 6231 10.9 2.020 3.401 2.020 3.401 arnoldi_extremal 4 6.8 0.000 0.000 3.169 3.192 arnoldi_normal_ev 4 7.8 0.001 0.003 3.169 3.192 build_subspace 16 8.4 0.009 0.012 3.079 3.081 ls_scf_post 1 4.0 0.000 0.000 3.048 3.059 ls_scf_store_result 1 5.0 0.000 0.000 2.862 2.903 dbcsr_special_finalize 555 9.7 0.005 0.006 2.381 2.746 dbcsr_merge_single_wm 555 10.7 0.460 0.574 2.372 2.737 make_images_pack 222 9.7 2.205 2.621 2.207 2.623 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.305 2.546 dbcsr_sort_data 658 11.4 2.171 2.487 2.171 2.487 dbcsr_matrix_vector_mult_local 304 10.0 2.064 2.453 2.066 2.455 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.320 2.415 buffer_matrices_ensure_size 222 8.7 1.756 2.093 1.756 2.093 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.761 1.763 rebuild_ks_matrix 3 7.3 0.000 0.000 1.752 1.753 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.752 1.753 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=84.677000, 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/1e516dface89e0648f7d6ff6212138fc2d72d23d_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 2.172142E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266673. MP_Allreduce 3138 10075. MP_Sync 4 MP_Alltoall 47 15335933. MP_SendRecv 141 57600. MP_ISendRecv 141 57600. MP_Wait 687 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.039 89.678 89.678 qs_energies 1 2.0 0.000 0.000 89.220 89.224 ls_scf 1 3.0 0.000 0.000 87.884 87.888 dbcsr_multiply_generic 111 6.7 0.015 0.016 74.012 74.333 multiply_cannon 111 7.7 0.027 0.041 52.677 56.932 ls_scf_main 1 4.0 0.000 0.000 54.294 54.300 multiply_cannon_loop 111 8.7 0.135 0.146 49.855 52.705 density_matrix_trs4 2 5.0 0.002 0.003 48.684 48.881 ls_scf_init_scf 1 4.0 0.000 0.000 29.859 29.861 mp_waitall_1 9105 10.9 20.568 29.106 20.568 29.106 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.682 28.761 multiply_cannon_multrec 1332 9.7 13.292 17.215 22.583 27.639 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.286 26.295 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.435 20.199 make_m2s 222 7.7 0.006 0.007 14.943 15.524 make_images 222 8.7 1.368 1.695 14.912 15.494 dbcsr_mm_accdrv_process 4041 10.4 0.352 0.548 8.886 10.475 dbcsr_mm_accdrv_process_sort 4041 11.4 8.401 9.926 8.401 9.926 make_images_data 222 9.7 0.004 0.004 8.676 9.643 hybrid_alltoall_any 227 10.6 0.544 2.559 8.047 9.414 mp_sum_l 887 5.1 4.917 7.803 4.917 7.803 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.205 7.762 mp_irecv_dv 3311 11.0 3.185 7.707 3.185 7.707 calculate_norms 2376 9.8 6.066 6.722 6.066 6.722 multiply_cannon_sync_h2d 1332 9.7 4.958 6.499 4.958 6.499 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.754 6.401 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.879 5.089 arnoldi_extremal 4 6.8 0.000 0.000 4.641 4.656 arnoldi_normal_ev 4 7.8 0.001 0.004 4.641 4.656 build_subspace 16 8.4 0.014 0.021 4.395 4.398 ls_scf_post 1 4.0 0.000 0.000 3.730 3.734 ls_scf_store_result 1 5.0 0.000 0.000 3.386 3.555 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.129 3.360 dbcsr_matrix_vector_mult_local 304 10.0 2.750 3.231 2.752 3.233 mp_allgather_i34 111 8.7 0.819 3.058 0.819 3.058 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.603 2.678 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.107 2.521 dbcsr_data_new 4174 10.1 2.114 2.401 2.114 2.401 make_images_pack 222 9.7 1.836 2.139 1.839 2.141 dbcsr_sort_data 436 11.2 1.827 2.051 1.827 2.051 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.846 1.849 rebuild_ks_matrix 3 7.3 0.000 0.000 1.834 1.836 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.834 1.836 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=89.678000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1773.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1e516dface89e0648f7d6ff6212138fc2d72d23d_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 2.925146E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265448. MP_Allreduce 3138 10896. MP_Sync 4 MP_Alltoall 47 23526250. MP_SendRecv 93 57600. MP_ISendRecv 93 57600. MP_Wait 639 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.024 0.042 91.404 91.405 qs_energies 1 2.0 0.000 0.000 90.836 90.843 ls_scf 1 3.0 0.000 0.000 89.430 89.435 dbcsr_multiply_generic 111 6.7 0.016 0.016 74.058 74.379 ls_scf_main 1 4.0 0.000 0.000 55.546 55.550 multiply_cannon 111 7.7 0.036 0.104 51.506 54.812 multiply_cannon_loop 111 8.7 0.116 0.130 48.756 52.771 density_matrix_trs4 2 5.0 0.002 0.003 49.611 49.846 mp_waitall_1 7281 11.0 22.815 32.662 22.815 32.662 ls_scf_init_scf 1 4.0 0.000 0.000 30.235 30.238 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.055 29.123 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.716 26.730 multiply_cannon_multrec 888 9.7 12.652 15.500 21.222 24.687 multiply_cannon_metrocomm3 888 9.7 0.004 0.005 10.584 22.702 make_m2s 222 7.7 0.006 0.007 15.981 16.775 make_images 222 8.7 1.581 1.856 15.943 16.736 make_images_data 222 9.7 0.004 0.004 9.247 10.422 hybrid_alltoall_any 227 10.6 0.639 2.935 8.856 9.975 dbcsr_mm_accdrv_process 3754 10.4 0.319 0.570 8.108 9.411 dbcsr_mm_accdrv_process_sort 3754 11.4 7.665 8.841 7.665 8.841 mp_sum_l 887 5.1 4.876 8.006 4.876 8.006 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.452 7.295 mp_irecv_dv 2335 11.1 2.438 7.245 2.438 7.245 multiply_cannon_sync_h2d 888 9.7 6.062 7.222 6.062 7.222 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.624 7.128 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.764 6.224 arnoldi_extremal 4 6.8 0.000 0.000 5.156 5.173 arnoldi_normal_ev 4 7.8 0.001 0.005 5.156 5.173 apply_matrix_preconditioner 6 5.3 0.000 0.001 4.744 5.012 build_subspace 16 8.4 0.014 0.020 4.853 4.860 calculate_norms 1584 9.8 4.346 4.811 4.346 4.811 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.452 3.779 ls_scf_post 1 4.0 0.000 0.000 3.649 3.654 dbcsr_matrix_vector_mult_local 304 10.0 3.034 3.608 3.036 3.610 mp_allgather_i34 111 8.7 0.869 3.583 0.869 3.583 ls_scf_store_result 1 5.0 0.000 0.000 3.382 3.470 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.883 3.016 dbcsr_data_new 4116 9.9 2.102 2.441 2.102 2.441 dbcsr_sort_data 325 11.1 1.866 2.168 1.866 2.168 make_images_sizes 222 9.7 0.000 0.001 0.912 2.091 mp_alltoall_i44 222 10.7 0.911 2.090 0.911 2.090 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.906 1.908 dbcsr_finalize 304 7.8 0.026 0.032 1.601 1.899 make_images_pack 222 9.7 1.625 1.890 1.628 1.892 rebuild_ks_matrix 3 7.3 0.000 0.000 1.889 1.890 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.889 1.890 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=91.405000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2240.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1e516dface89e0648f7d6ff6212138fc2d72d23d_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 3.293340E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266673. MP_Allreduce 3138 13030. MP_Sync 4 MP_Alltoall 47 30278988. MP_SendRecv 69 86400. MP_ISendRecv 69 86400. MP_Wait 531 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.028 0.086 97.521 97.522 qs_energies 1 2.0 0.000 0.000 96.874 96.882 ls_scf 1 3.0 0.000 0.000 95.210 95.217 dbcsr_multiply_generic 111 6.7 0.017 0.019 79.015 79.285 ls_scf_main 1 4.0 0.000 0.000 59.339 59.340 multiply_cannon 111 7.7 0.044 0.140 51.854 56.304 density_matrix_trs4 2 5.0 0.002 0.003 53.213 53.404 multiply_cannon_loop 111 8.7 0.152 0.168 46.834 50.421 ls_scf_init_scf 1 4.0 0.000 0.000 32.622 32.624 ls_scf_init_matrix_S 1 5.0 0.000 0.001 31.399 31.486 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.954 28.970 mp_waitall_1 6369 11.0 22.764 28.934 22.764 28.934 multiply_cannon_multrec 1332 9.7 14.166 17.308 22.079 24.927 make_m2s 222 7.7 0.006 0.008 21.151 22.567 make_images 222 8.7 3.149 3.615 21.101 22.519 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.187 17.655 make_images_data 222 9.7 0.004 0.004 11.718 13.197 hybrid_alltoall_any 227 10.6 0.798 3.755 11.129 12.705 dbcsr_mm_accdrv_process 3641 10.4 0.293 0.474 7.545 9.085 dbcsr_mm_accdrv_process_sort 3641 11.4 7.105 8.589 7.105 8.589 mp_sum_l 887 5.1 4.423 8.281 4.423 8.281 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.391 6.553 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.099 6.081 mp_irecv_dv 3229 10.9 2.075 6.000 2.075 6.000 multiply_cannon_sync_h2d 1332 9.7 5.498 5.898 5.498 5.898 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.691 5.555 arnoldi_extremal 4 6.8 0.000 0.000 5.214 5.230 arnoldi_normal_ev 4 7.8 0.001 0.005 5.214 5.230 build_subspace 16 8.4 0.014 0.021 4.879 4.888 apply_matrix_preconditioner 6 5.3 0.000 0.001 4.567 4.761 mp_allgather_i34 111 8.7 2.129 4.560 2.129 4.560 calculate_norms 2376 9.8 4.184 4.556 4.184 4.556 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.595 3.900 dbcsr_matrix_vector_mult_local 304 10.0 3.218 3.720 3.220 3.722 dbcsr_sort_data 658 11.4 3.116 3.541 3.116 3.541 dbcsr_special_finalize 555 9.7 0.006 0.007 2.867 3.277 dbcsr_merge_single_wm 555 10.7 0.542 0.670 2.859 3.268 ls_scf_post 1 4.0 0.000 0.000 3.249 3.257 ls_scf_store_result 1 5.0 0.000 0.000 2.972 3.056 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.981 3.015 dbcsr_data_release 10477 10.7 1.588 2.419 1.588 2.419 dbcsr_finalize 304 7.8 0.049 0.061 1.803 2.032 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.522000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2711.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1e516dface89e0648f7d6ff6212138fc2d72d23d_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 4.731998E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265536. MP_Allreduce 3129 15263. MP_Sync 4 MP_Alltoall 47 46208988. MP_SendRecv 45 115200. MP_ISendRecv 45 115200. MP_Wait 528 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.052 0.056 92.207 92.208 qs_energies 1 2.0 0.000 0.000 90.841 90.857 ls_scf 1 3.0 0.000 0.000 88.865 88.871 dbcsr_multiply_generic 111 6.7 0.017 0.019 70.451 70.656 ls_scf_main 1 4.0 0.000 0.000 56.204 56.204 multiply_cannon 111 7.7 0.099 0.184 52.370 55.507 multiply_cannon_loop 111 8.7 0.089 0.094 49.795 50.967 density_matrix_trs4 2 5.0 0.002 0.003 49.213 49.292 ls_scf_init_scf 1 4.0 0.000 0.000 29.264 29.265 mp_waitall_1 5436 11.0 24.136 29.149 24.136 29.149 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.034 28.065 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.946 25.960 multiply_cannon_multrec 444 9.7 13.631 16.296 20.743 21.987 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.665 15.385 make_m2s 222 7.7 0.005 0.005 13.406 14.357 make_images 222 8.7 2.044 2.496 13.339 14.288 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.131 13.602 make_images_data 222 9.7 0.003 0.004 8.157 9.724 hybrid_alltoall_any 227 10.6 0.800 3.817 8.048 9.388 multiply_cannon_sync_h2d 444 9.7 6.785 8.335 6.785 8.335 dbcsr_mm_accdrv_process 3003 10.4 0.383 0.439 6.793 7.910 dbcsr_mm_accdrv_process_sort 3003 11.4 6.409 7.471 6.409 7.471 arnoldi_extremal 4 6.8 0.000 0.000 5.857 5.872 arnoldi_normal_ev 4 7.8 0.002 0.005 5.856 5.872 build_subspace 16 8.4 0.015 0.019 5.459 5.468 apply_matrix_preconditioner 6 5.3 0.000 0.001 4.434 4.634 dbcsr_matrix_vector_mult 304 9.0 0.011 0.022 4.215 4.415 dbcsr_matrix_vector_mult_local 304 10.0 3.745 4.223 3.747 4.225 mp_sum_l 887 5.1 2.662 4.170 2.662 4.170 calculate_norms 792 9.8 3.623 3.715 3.623 3.715 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.480 3.701 mp_irecv_dv 1241 11.2 1.467 3.678 1.467 3.678 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.534 3.603 mp_allgather_i34 111 8.7 1.160 3.553 1.160 3.553 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.901 3.446 ls_scf_post 1 4.0 0.000 0.000 3.398 3.404 make_images_sizes 222 9.7 0.000 0.000 0.863 3.280 mp_alltoall_i44 222 10.7 0.863 3.279 0.863 3.279 ls_scf_store_result 1 5.0 0.000 0.000 3.190 3.221 dbcsr_finalize 304 7.8 0.062 0.077 2.203 2.301 dbcsr_data_new 4608 9.7 1.769 2.244 1.769 2.244 dbcsr_merge_all 275 8.9 0.483 0.528 2.061 2.142 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.003 2.004 rebuild_ks_matrix 3 7.3 0.000 0.000 1.971 1.972 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.000 1.971 1.972 qs_energies_init_hamiltonians 1 3.0 0.000 0.001 1.960 1.970 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=92.208000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3791.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1e516dface89e0648f7d6ff6212138fc2d72d23d_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 8.873906E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284089. MP_Allreduce 3123 21388. MP_Sync 4 MP_Alltoall 47 88727262. MP_SendRecv 42 732600. MP_ISendRecv 42 732600. MP_Wait 267 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.074 0.090 105.640 105.641 qs_energies 1 2.0 0.000 0.000 103.981 103.988 ls_scf 1 3.0 0.000 0.000 101.061 101.067 dbcsr_multiply_generic 111 6.7 0.024 0.028 74.712 74.843 ls_scf_main 1 4.0 0.000 0.000 63.485 63.485 density_matrix_trs4 2 5.0 0.002 0.003 54.545 54.603 multiply_cannon 111 7.7 0.146 0.223 48.406 50.476 multiply_cannon_loop 111 8.7 0.098 0.099 45.417 46.028 ls_scf_init_scf 1 4.0 0.000 0.000 33.807 33.808 ls_scf_init_matrix_S 1 5.0 0.000 0.001 32.314 32.338 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.609 29.619 mp_waitall_1 4527 11.1 21.723 25.111 21.723 25.111 make_m2s 222 7.7 0.005 0.005 22.684 23.762 make_images 222 8.7 3.581 3.878 22.576 23.654 multiply_cannon_multrec 444 9.7 17.833 18.454 22.661 23.414 hybrid_alltoall_any 227 10.6 1.654 3.635 12.858 15.798 make_images_data 222 9.7 0.003 0.004 13.092 15.183 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 9.948 10.698 multiply_cannon_sync_h2d 444 9.7 8.788 8.854 8.788 8.854 arnoldi_extremal 4 6.8 0.000 0.000 7.498 7.508 arnoldi_normal_ev 4 7.8 0.003 0.009 7.498 7.508 build_subspace 16 8.4 0.026 0.036 6.951 6.963 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.582 5.746 dbcsr_matrix_vector_mult_local 304 10.0 5.155 5.475 5.157 5.477 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.082 5.173 apply_matrix_preconditioner 6 5.3 0.000 0.001 4.855 5.109 dbcsr_mm_accdrv_process 1814 10.4 0.497 0.723 4.645 4.957 dbcsr_mm_accdrv_process_sort 1814 11.4 4.124 4.254 4.124 4.254 ls_scf_post 1 4.0 0.000 0.000 3.768 3.776 ls_scf_store_result 1 5.0 0.000 0.000 3.496 3.532 make_images_sizes 222 9.7 0.000 0.000 1.455 3.513 mp_alltoall_i44 222 10.7 1.455 3.512 1.455 3.512 mp_allgather_i34 111 8.7 1.086 3.483 1.086 3.483 calculate_norms 792 9.8 3.233 3.269 3.233 3.269 dbcsr_finalize 304 7.8 0.082 0.089 3.103 3.198 dbcsr_merge_all 275 8.9 0.885 0.924 2.870 2.953 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.888 2.888 dbcsr_complete_redistribute 5 7.6 1.428 1.473 2.740 2.855 matrix_ls_to_qs 2 6.0 0.000 0.000 2.408 2.533 dbcsr_sort_data 325 11.1 2.445 2.503 2.445 2.503 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.381 2.383 dbcsr_data_new 6591 9.6 1.856 2.332 1.856 2.332 rebuild_ks_matrix 3 7.3 0.000 0.000 2.315 2.317 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.315 2.317 dbcsr_new_transposed 4 7.5 0.241 0.251 2.282 2.292 dbcsr_frobenius_norm 74 6.6 2.056 2.132 2.203 2.233 dbcsr_add_d 103 6.2 0.000 0.000 2.144 2.229 dbcsr_add_anytype 103 7.2 0.858 0.890 2.144 2.229 dbcsr_data_release 12724 10.6 1.991 2.192 1.991 2.192 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=105.641000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=7063.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1e516dface89e0648f7d6ff6212138fc2d72d23d_performance_tests/27/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 7009386627072 0.0% 0.0% 100.0% flops 9 x 9 x 32 7335108845568 0.0% 0.0% 100.0% flops 9 x 22 x 32 9866241589248 0.0% 0.0% 100.0% flops 22 x 9 x 32 9884108906496 0.0% 0.0% 100.0% flops 22 x 22 x 32 13354440523776 0.0% 0.0% 100.0% flops 32 x 32 x 9 20607185977344 0.0% 0.0% 100.0% flops 32 x 32 x 22 25186560638976 0.0% 0.0% 100.0% flops 9 x 32 x 32 28458319085568 0.0% 0.0% 100.0% flops 22 x 32 x 32 34782389993472 0.0% 0.0% 100.0% flops 9 x 32 x 9 42881542373376 0.0% 0.0% 100.0% flops 22 x 32 x 9 55680402235392 0.0% 0.0% 100.0% flops 9 x 32 x 22 55680402235392 0.0% 0.0% 100.0% flops 22 x 32 x 22 72328573419520 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 383.054662E+12 0.0% 0.0% 100.0% flops max/rank 733.641090E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 26899403712 0.0% 0.0% 100.0% number of processed stacks 118860288 0.0% 0.0% 100.0% average stack size 0.0 0.0 226.3 marketing flops 780.439111E+12 ------------------------------------------------------------------------------- # multiplications 1445 max memory usage/rank 588.804096E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 102097920 MPI messages size (bytes): total size 37.227590E+12 min size 0.000000E+00 max size 4.551360E+06 average size 364.626312E+03 MPI breakdown and total messages size (bytes): size <= 128 731472 0 128 < size <= 8192 11922720 97670922240 8192 < size <= 32768 24718992 614677610496 32768 < size <= 131072 20000256 1970081366016 131072 < size <= 4194304 42515668 24886801223040 4194304 < size <= 16777216 2208812 9656099886720 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4640 78072. MP_Allreduce 13232 2081. MP_Sync 1064 MP_Alltoall 2588 1144059044. 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 0.011 0.033 226.881 226.882 qs_mol_dyn_low 1 2.0 0.003 0.003 225.964 225.986 qs_forces 5 3.8 0.004 0.004 225.901 225.903 qs_energies 5 4.8 0.001 0.001 222.971 222.989 scf_env_do_scf 5 5.8 0.000 0.001 207.552 207.555 scf_env_do_scf_inner_loop 105 6.6 0.002 0.006 181.375 181.376 qs_scf_new_mos 105 7.6 0.001 0.001 141.773 141.923 qs_scf_loop_do_ot 105 8.6 0.001 0.001 141.772 141.922 dbcsr_multiply_generic 1445 12.2 0.126 0.135 134.435 134.816 ot_scf_mini 105 9.6 0.003 0.003 131.878 132.018 multiply_cannon 1445 13.2 0.275 0.286 115.551 118.290 multiply_cannon_loop 1445 14.2 2.858 3.015 113.846 115.136 velocity_verlet 4 3.0 0.002 0.006 105.793 105.794 ot_mini 105 10.6 0.001 0.001 59.948 60.078 multiply_cannon_multrec 69360 15.2 29.634 34.863 39.852 44.811 qs_ot_get_p 112 10.4 0.001 0.001 41.855 42.166 mp_waitall_1 488190 16.1 34.903 41.873 34.903 41.873 qs_ot_get_derivative 55 11.6 0.001 0.001 38.236 38.399 multiply_cannon_sync_h2d 69360 15.2 29.060 33.317 29.060 33.317 multiply_cannon_metrocomm3 69360 15.2 0.202 0.213 25.816 33.084 qs_ot_p2m_diag 40 11.0 0.020 0.030 30.901 30.974 rebuild_ks_matrix 110 8.4 0.000 0.000 28.761 28.915 qs_ks_build_kohn_sham_matrix 110 9.4 0.011 0.012 28.761 28.915 cp_dbcsr_syevd 40 12.0 0.002 0.002 27.560 27.560 qs_ks_update_qs_env 112 7.6 0.001 0.001 26.403 26.543 init_scf_loop 7 6.6 0.000 0.000 26.149 26.150 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 22.999 23.267 apply_single 62 13.6 0.000 0.000 22.999 23.267 cp_fm_syevd 40 13.0 0.000 0.001 22.466 22.597 prepare_preconditioner 7 7.6 0.000 0.000 21.412 21.446 make_preconditioner 7 8.6 0.000 0.000 21.412 21.446 ot_new_cg_direction 55 11.6 0.001 0.001 20.989 20.990 cp_fm_redistribute_end 40 14.0 8.869 17.698 8.874 17.700 cp_fm_syevd_base 40 14.0 8.819 17.649 8.819 17.649 qs_rho_update_rho_low 110 7.6 0.001 0.001 16.520 16.865 calculate_rho_elec 110 8.6 0.029 0.032 16.519 16.865 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.840 14.987 make_full_inverse_cholesky 7 9.6 0.000 0.000 14.243 14.302 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 13.815 13.925 mp_sum_l 4764 12.2 12.072 12.907 12.072 12.907 init_scf_run 5 5.8 0.000 0.001 12.876 12.876 scf_env_initial_rho_setup 5 6.8 0.001 0.001 12.875 12.876 pw_transfer 1645 12.4 0.080 0.102 11.704 11.881 calculate_dm_sparse 110 9.5 0.000 0.001 11.713 11.848 fft_wrap_pw1pw2 1425 13.5 0.012 0.015 11.565 11.746 dbcsr_mm_accdrv_process 154766 15.8 6.214 6.407 10.087 11.608 density_rs2pw 110 9.6 0.005 0.006 10.882 11.352 qs_vxc_create 110 10.4 0.002 0.003 10.442 10.485 qs_ot_get_derivative_diag 18 12.0 0.000 0.000 10.330 10.392 fft_wrap_pw1pw2_240 915 15.0 1.144 1.233 10.139 10.339 cp_fm_cholesky_invert 7 10.6 9.905 9.913 9.905 9.913 check_diag 80 13.5 8.687 8.947 9.502 9.650 acc_transpose_blocks 69360 15.2 0.357 0.387 8.474 8.989 multiply_cannon_metrocomm1 69360 15.2 0.095 0.101 4.673 8.581 fft3d_pb 915 16.0 2.402 2.660 8.329 8.561 sum_up_and_integrate 60 10.3 0.001 0.002 8.332 8.343 integrate_v_rspace 60 11.3 0.002 0.002 8.315 8.327 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.001 8.237 8.254 calculate_first_density_matrix 1 7.0 0.000 0.001 7.980 7.999 transfer_rs2pw 445 10.6 0.007 0.008 7.525 7.960 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 7.859 7.907 xc_rho_set_and_dset_create 110 12.4 0.076 0.098 7.347 7.611 xc_vxc_pw_create 60 11.3 0.039 0.049 6.952 6.995 make_full_single_inverse 7 9.6 0.001 0.001 6.911 6.945 make_m2s 2890 13.2 0.078 0.087 6.310 6.871 make_images 2890 14.2 0.238 0.258 6.203 6.763 xc_pw_derive 510 13.4 0.006 0.007 6.202 6.270 acc_transpose_blocks_kernels 69360 16.2 0.849 0.895 5.699 6.115 mp_alltoall_z22v 2340 17.7 5.349 5.672 5.349 5.672 jit_kernel_transpose 5 15.0 4.850 5.241 4.850 5.241 mp_waitany 7680 13.5 4.336 4.911 4.336 4.911 multiply_cannon_metrocomm4 67915 15.2 0.183 0.206 2.025 4.724 potential_pw2rs 60 12.3 0.003 0.003 4.589 4.613 wfi_extrapolate 5 7.8 0.000 0.001 4.538 4.538 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=226.882000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=560.600000, yerr=1.959592 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 1e516dface89e0648f7d6ff6212138fc2d72d23d Summary: empty Status: OK