=== 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: 5080151e8b1ac516ab82332683adfb2cbf2d479f ################# 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, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.2.2, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.2, # SIRIUS 7.4.3, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Last update: 21.06.2023 # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2023.05.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.2 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5080151e8b1ac516ab82332683adfb2cbf2d479f_performance_tests/01 job id: 48379777 --- 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/5080151e8b1ac516ab82332683adfb2cbf2d479f_performance_tests/02 job id: 48379778 --- 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/5080151e8b1ac516ab82332683adfb2cbf2d479f_performance_tests/03 job id: 48379779 --- 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/5080151e8b1ac516ab82332683adfb2cbf2d479f_performance_tests/04 job id: 48379780 --- 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/5080151e8b1ac516ab82332683adfb2cbf2d479f_performance_tests/05 job id: 48379781 --- 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/5080151e8b1ac516ab82332683adfb2cbf2d479f_performance_tests/06 job id: 48379782 --- 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/5080151e8b1ac516ab82332683adfb2cbf2d479f_performance_tests/07 job id: 48379783 --- 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/5080151e8b1ac516ab82332683adfb2cbf2d479f_performance_tests/08 job id: 48379784 --- 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/5080151e8b1ac516ab82332683adfb2cbf2d479f_performance_tests/09 job id: 48379785 --- 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/5080151e8b1ac516ab82332683adfb2cbf2d479f_performance_tests/10 job id: 48379786 --- 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/5080151e8b1ac516ab82332683adfb2cbf2d479f_performance_tests/11 job id: 48379787 --- 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/5080151e8b1ac516ab82332683adfb2cbf2d479f_performance_tests/12 job id: 48379788 --- 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/5080151e8b1ac516ab82332683adfb2cbf2d479f_performance_tests/13 job id: 48379789 --- 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/5080151e8b1ac516ab82332683adfb2cbf2d479f_performance_tests/14 job id: 48379790 --- 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/5080151e8b1ac516ab82332683adfb2cbf2d479f_performance_tests/15 job id: 48379791 --- 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/5080151e8b1ac516ab82332683adfb2cbf2d479f_performance_tests/16 job id: 48379792 --- 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/5080151e8b1ac516ab82332683adfb2cbf2d479f_performance_tests/17 job id: 48379793 --- 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/5080151e8b1ac516ab82332683adfb2cbf2d479f_performance_tests/18 job id: 48379794 --- 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/5080151e8b1ac516ab82332683adfb2cbf2d479f_performance_tests/19 job id: 48379795 --- 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/5080151e8b1ac516ab82332683adfb2cbf2d479f_performance_tests/20 job id: 48379796 --- 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/5080151e8b1ac516ab82332683adfb2cbf2d479f_performance_tests/21 job id: 48379797 --- 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/5080151e8b1ac516ab82332683adfb2cbf2d479f_performance_tests/22 job id: 48379798 --- 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/5080151e8b1ac516ab82332683adfb2cbf2d479f_performance_tests/23 job id: 48379799 --- 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/5080151e8b1ac516ab82332683adfb2cbf2d479f_performance_tests/24 job id: 48379800 --- 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/5080151e8b1ac516ab82332683adfb2cbf2d479f_performance_tests/25 job id: 48379801 --- 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/5080151e8b1ac516ab82332683adfb2cbf2d479f_performance_tests/26 job id: 48379802 --- 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/5080151e8b1ac516ab82332683adfb2cbf2d479f_performance_tests/27 job id: 48379803 --- 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/5080151e8b1ac516ab82332683adfb2cbf2d479f_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 0.033 0.041 134.808 134.816 farming_run 1 2.0 133.601 133.618 134.735 134.781 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.496736E+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 0.007 0.026 114.995 114.995 qs_energies 1 2.0 0.000 0.000 114.739 114.740 mp2_main 1 3.0 0.000 0.000 112.672 112.673 mp2_gpw_main 1 4.0 0.027 0.033 111.681 111.683 mp2_ri_gpw_compute_in 1 5.0 0.173 0.176 92.674 93.120 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 54.746 55.191 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.165 41.130 46.261 get_2c_integrals 1 6.0 0.008 0.009 37.185 37.754 integrate_v_rspace 273 8.0 0.440 0.452 24.695 29.574 pw_transfer 6555 10.6 0.372 0.389 26.971 27.686 fft_wrap_pw1pw2 5465 11.4 0.044 0.047 25.585 26.236 grid_integrate_task_list 273 9.0 20.593 25.942 20.593 25.942 fft_wrap_pw1pw2_100 2178 12.4 1.171 1.473 23.146 23.786 compute_2c_integrals 1 7.0 0.015 0.017 19.497 19.499 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.727 19.220 mp2_eri_2c_integrate_gpw 1 9.0 2.354 2.415 18.724 19.216 rpa_ri_compute_en 1 5.0 0.021 0.022 18.890 19.156 cp_fm_cholesky_decompose 12 8.2 17.619 18.205 17.619 18.205 cholesky_decomp 1 7.0 0.000 0.000 16.535 17.126 fft3d_s 5443 13.4 16.193 16.485 16.215 16.507 ao_to_mo_and_store_B_mult_1 272 7.0 10.775 15.301 10.775 15.301 calculate_wavefunction 272 8.0 5.406 5.562 12.343 12.981 rpa_num_int 1 6.0 0.000 0.002 10.647 10.647 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.596 10.643 calc_mat_Q 8 8.0 0.000 0.000 9.462 9.578 contract_S_to_Q 8 9.0 0.000 0.000 8.883 8.999 calc_potential_gpw 544 9.5 0.005 0.006 8.285 8.764 parallel_gemm_fm 14 9.1 0.000 0.000 8.481 8.564 parallel_gemm_fm_cosma 14 10.1 8.481 8.564 8.481 8.564 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 8.161 8.487 potential_pw2rs 545 10.0 0.107 0.110 7.522 8.258 create_integ_mat 1 6.0 0.023 0.028 7.704 7.705 collocate_single_gaussian 272 10.0 0.039 0.042 7.402 7.662 array2fm 1 7.0 0.000 0.000 6.625 7.118 pw_scatter_s 2720 13.7 4.369 4.476 4.369 4.476 pw_gather_s 2722 13.2 3.454 3.797 3.454 3.797 array2fm_buffer_send 1 8.0 2.889 3.108 2.889 3.108 pw_poisson_solve 545 10.5 1.116 1.192 2.212 2.478 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.680552, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2792.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5080151e8b1ac516ab82332683adfb2cbf2d479f_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 0.030 0.041 423.879 423.880 farming_run 1 2.0 422.570 422.690 423.835 423.842 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827141120 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788822 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.238766E+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 23183820. 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 0.032 217.373 217.373 qs_energies 1 2.0 0.002 0.007 217.098 217.120 scf_env_do_scf 1 3.0 0.000 0.000 114.962 114.962 qs_ks_update_qs_env 5 5.0 0.000 0.000 114.059 114.065 rebuild_ks_matrix 4 6.0 0.000 0.000 114.058 114.064 qs_ks_build_kohn_sham_matrix 4 7.0 0.055 0.061 114.058 114.064 hfx_ks_matrix 4 8.0 0.001 0.001 113.698 113.702 integrate_four_center 4 9.0 0.154 0.466 113.697 113.702 integrate_four_center_main 4 10.0 0.137 0.681 101.933 105.403 integrate_four_center_bin 264 11.0 101.796 105.324 101.796 105.324 mp2_main 1 3.0 0.001 0.007 101.843 101.866 mp2_gpw_main 1 4.0 0.058 0.129 100.941 100.967 init_scf_loop 1 4.0 0.000 0.000 97.055 97.055 mp2_ri_gpw_compute_in 1 5.0 0.082 0.186 73.957 75.014 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.020 53.839 54.909 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.161 41.617 46.588 integrate_v_rspace 95 8.0 0.396 0.569 28.103 32.938 pw_transfer 2240 10.6 0.144 0.164 29.510 29.903 fft_wrap_pw1pw2 1868 11.4 0.018 0.020 28.516 28.903 mp2_ri_gpw_compute_en 1 5.0 0.080 0.103 26.761 28.773 grid_integrate_task_list 95 9.0 23.507 28.567 23.507 28.567 ao_to_mo_and_store_B_mult_1 91 7.0 10.546 28.200 10.546 28.200 fft_wrap_pw1pw2_100 730 12.4 1.283 1.477 26.279 26.680 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.832 1.876 25.036 25.047 get_2c_integrals 1 6.0 0.000 0.000 20.016 20.041 compute_2c_integrals 1 7.0 0.002 0.003 18.997 19.004 fft3d_s 1823 13.4 18.502 18.906 18.516 18.920 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 18.628 18.869 mp2_eri_2c_integrate_gpw 1 9.0 1.736 1.870 18.627 18.868 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 17.906 17.906 calculate_wavefunction 91 8.0 2.003 2.032 9.615 9.856 mp2_ri_gpw_compute_en_expansio 172 7.0 0.557 0.575 8.881 9.301 potential_pw2rs 186 10.0 0.034 0.035 8.451 9.068 local_gemm 172 8.0 8.324 8.738 8.324 8.738 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.151 8.515 mp2_ri_gpw_compute_en_comm 22 7.0 0.502 0.516 7.952 8.279 calc_potential_gpw 182 9.5 0.002 0.002 7.887 8.090 collocate_single_gaussian 91 10.0 0.019 0.067 7.756 7.977 integrate_four_center_load 4 10.0 0.000 0.000 6.775 6.779 hfx_load_balance 1 11.0 0.000 0.000 6.774 6.779 mp2_ri_gpw_compute_en_ener 172 7.0 6.326 6.387 6.326 6.387 mp_sendrecv_dm3 2068 8.0 5.975 6.318 5.975 6.318 mp_sync 37 10.5 3.718 6.116 3.718 6.116 pw_gather_s 912 13.2 4.490 4.989 4.490 4.989 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=100.949475, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1499.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5080151e8b1ac516ab82332683adfb2cbf2d479f_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 453.574656E+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 0.030 0.103 54.228 54.232 qs_mol_dyn_low 1 2.0 0.003 0.004 53.743 53.775 qs_forces 11 3.9 0.002 0.003 53.681 53.683 qs_energies 11 4.9 0.015 0.016 52.204 52.214 scf_env_do_scf 11 5.9 0.000 0.002 45.034 45.034 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 42.894 42.894 dbcsr_multiply_generic 2286 12.5 0.093 0.099 32.926 33.237 qs_scf_new_mos 108 7.5 0.000 0.001 32.731 33.058 qs_scf_loop_do_ot 108 8.5 0.000 0.001 32.731 33.058 ot_scf_mini 108 9.5 0.002 0.002 31.156 31.342 multiply_cannon 2286 13.5 0.190 0.200 26.039 27.398 multiply_cannon_loop 2286 14.5 1.824 1.940 25.344 26.725 velocity_verlet 10 3.0 0.002 0.002 25.616 25.617 ot_mini 108 10.5 0.001 0.001 18.864 19.113 qs_ot_get_derivative 108 11.5 0.001 0.001 15.968 16.159 mp_waitall_1 245248 16.5 7.918 14.133 7.918 14.133 multiply_cannon_metrocomm3 54864 15.5 0.073 0.082 5.663 12.436 multiply_cannon_multrec 54864 15.5 3.689 5.802 7.823 11.219 qs_ot_get_p 119 10.4 0.001 0.001 7.895 8.173 rebuild_ks_matrix 119 8.3 0.000 0.000 7.953 8.119 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 7.953 8.119 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.003 7.160 mp_sum_l 7287 12.8 4.888 6.362 4.888 6.362 multiply_cannon_sync_h2d 54864 15.5 5.198 6.106 5.198 6.106 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.504 5.989 dbcsr_mm_accdrv_process 76910 16.1 1.835 2.920 4.046 5.831 init_scf_run 11 5.9 0.000 0.001 5.662 5.662 scf_env_initial_rho_setup 11 6.9 0.005 0.005 5.662 5.662 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.259 5.366 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.184 5.215 sum_up_and_integrate 119 10.3 0.012 0.014 4.544 4.550 integrate_v_rspace 119 11.3 0.002 0.003 4.532 4.539 cp_dbcsr_syevd 50 12.0 0.002 0.003 4.425 4.425 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.237 4.339 calculate_rho_elec 119 8.7 0.011 0.017 4.236 4.339 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.216 4.217 cp_fm_redistribute_end 50 14.0 2.152 4.191 2.157 4.194 cp_fm_diag_elpa_base 50 14.0 2.031 4.090 2.035 4.098 calculate_first_density_matrix 1 7.0 0.000 0.000 3.425 3.433 calculate_dm_sparse 119 9.5 0.000 0.000 2.959 3.117 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.902 3.090 apply_single 119 13.6 0.000 0.000 2.902 3.089 acc_transpose_blocks 54864 15.5 0.237 0.265 2.378 2.934 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.840 2.842 jit_kernel_multiply 13 15.8 2.146 2.818 2.146 2.818 ot_diis_step 108 11.5 0.006 0.006 2.675 2.676 multiply_cannon_metrocomm1 54864 15.5 0.057 0.063 1.538 2.604 density_rs2pw 119 9.7 0.004 0.004 2.251 2.349 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.215 2.257 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.149 2.190 wfi_extrapolate 11 7.9 0.001 0.001 2.158 2.158 grid_integrate_task_list 119 12.3 2.017 2.125 2.017 2.125 init_scf_loop 11 6.9 0.000 0.000 2.122 2.123 potential_pw2rs 119 12.3 0.004 0.004 1.897 1.917 mp_sum_d 4135 12.0 1.254 1.871 1.254 1.871 pw_transfer 1439 11.6 0.052 0.057 1.729 1.804 make_m2s 4572 13.5 0.054 0.057 1.737 1.788 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.653 1.730 make_images 4572 14.5 0.133 0.138 1.654 1.704 transfer_rs2pw 487 10.6 0.005 0.006 1.505 1.682 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.547 1.566 mp_alltoall_d11v 2130 13.8 1.364 1.511 1.364 1.511 mp_waitany 12084 13.8 1.344 1.469 1.344 1.469 acc_transpose_blocks_sync 164592 16.5 1.204 1.457 1.204 1.457 fft3d_ps 1201 14.6 0.370 0.475 1.369 1.441 grid_collocate_task_list 119 9.7 1.348 1.409 1.348 1.409 transfer_pw2rs 487 13.2 0.006 0.007 1.388 1.403 fft_wrap_pw1pw2_140 487 13.2 0.140 0.154 1.281 1.359 acc_transpose_blocks_kernels 54864 16.5 0.249 0.385 0.891 1.258 parallel_gemm_fm 81 9.0 0.000 0.000 1.166 1.178 parallel_gemm_fm_cosma 81 10.0 1.165 1.177 1.165 1.177 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=54.232000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=432.000000, yerr=1.414214 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5080151e8b1ac516ab82332683adfb2cbf2d479f_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 487.538688E+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 1644636. 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 0.032 0.061 39.464 39.513 qs_mol_dyn_low 1 2.0 0.003 0.003 39.108 39.129 qs_forces 11 3.9 0.003 0.003 39.045 39.049 qs_energies 11 4.9 0.002 0.007 37.371 37.378 scf_env_do_scf 11 5.9 0.000 0.001 31.823 31.824 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 29.400 29.401 dbcsr_multiply_generic 2286 12.5 0.099 0.102 22.236 22.573 qs_scf_new_mos 108 7.5 0.001 0.001 20.771 21.000 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.770 20.999 ot_scf_mini 108 9.5 0.002 0.003 19.847 20.008 multiply_cannon 2286 13.5 0.210 0.218 17.038 18.480 velocity_verlet 10 3.0 0.001 0.002 18.391 18.392 multiply_cannon_loop 2286 14.5 1.194 1.256 15.866 17.362 ot_mini 108 10.5 0.001 0.001 12.225 12.453 mp_waitall_1 200699 16.5 5.538 10.692 5.538 10.692 qs_ot_get_derivative 108 11.5 0.001 0.001 9.853 10.018 multiply_cannon_metrocomm3 27432 15.5 0.072 0.075 4.004 9.391 multiply_cannon_multrec 27432 15.5 1.823 4.034 6.209 9.006 rebuild_ks_matrix 119 8.3 0.000 0.000 6.781 6.929 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 6.780 6.929 dbcsr_mm_accdrv_process 47894 16.0 3.575 5.864 4.305 6.409 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.962 6.096 qs_ot_get_p 119 10.4 0.001 0.001 4.693 4.912 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.693 4.608 init_scf_run 11 5.9 0.000 0.001 4.254 4.254 scf_env_initial_rho_setup 11 6.9 0.001 0.004 4.254 4.254 mp_sum_l 7287 12.8 2.125 4.145 2.125 4.145 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.077 4.057 apply_single 119 13.6 0.000 0.000 3.077 4.056 sum_up_and_integrate 119 10.3 0.025 0.027 3.711 3.715 integrate_v_rspace 119 11.3 0.002 0.003 3.686 3.692 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.612 3.637 calculate_rho_elec 119 8.7 0.021 0.024 3.611 3.636 qs_ot_p2m_diag 50 11.0 0.009 0.013 3.036 3.056 make_m2s 4572 13.5 0.052 0.054 2.633 2.960 make_images 4572 14.5 0.206 0.246 2.546 2.876 calculate_first_density_matrix 1 7.0 0.001 0.004 2.728 2.731 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.637 2.638 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.338 2.424 init_scf_loop 11 6.9 0.000 0.000 2.402 2.402 ot_diis_step 108 11.5 0.011 0.011 2.322 2.322 calculate_dm_sparse 119 9.5 0.000 0.001 2.232 2.308 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.307 2.307 cp_fm_redistribute_end 50 14.0 1.168 2.279 1.172 2.281 multiply_cannon_sync_h2d 27432 15.5 1.702 2.265 1.702 2.265 cp_fm_diag_elpa_base 50 14.0 1.077 2.188 1.105 2.223 acc_transpose_blocks 27432 15.5 0.117 0.122 1.722 2.068 density_rs2pw 119 9.7 0.004 0.004 1.912 2.004 grid_integrate_task_list 119 12.3 1.831 1.934 1.831 1.934 pw_transfer 1439 11.6 0.064 0.069 1.840 1.877 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.873 1.876 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.749 1.788 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.671 1.711 jit_kernel_multiply 8 16.5 0.671 1.612 0.671 1.612 make_images_data 4572 15.5 0.048 0.054 1.255 1.591 prepare_preconditioner 11 7.9 0.000 0.000 1.523 1.548 make_preconditioner 11 8.9 0.000 0.000 1.523 1.548 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.421 1.477 wfi_extrapolate 11 7.9 0.001 0.001 1.468 1.468 potential_pw2rs 119 12.3 0.006 0.006 1.452 1.461 hybrid_alltoall_any 4725 16.4 0.054 0.115 1.114 1.460 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.421 1.432 fft_wrap_pw1pw2_140 487 13.2 0.159 0.168 1.383 1.421 fft3d_ps 1201 14.6 0.515 0.571 1.385 1.419 grid_collocate_task_list 119 9.7 1.290 1.346 1.290 1.346 mp_alltoall_d11v 2130 13.8 1.205 1.308 1.205 1.308 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.212 1.257 transfer_rs2pw 487 10.6 0.005 0.005 1.123 1.240 mp_sum_d 4135 12.0 0.544 1.026 0.544 1.026 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.018 1.023 mp_allgather_i34 2286 14.5 0.594 1.009 0.594 1.009 acc_transpose_blocks_kernels 27432 16.5 0.188 0.278 0.764 0.986 acc_transpose_blocks_sync 82296 16.5 0.815 0.944 0.815 0.944 transfer_pw2rs 487 13.2 0.004 0.005 0.917 0.922 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.875 0.889 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=39.513000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=464.181818, yerr=1.748671 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5080151e8b1ac516ab82332683adfb2cbf2d479f_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 518.328320E+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 0.045 0.100 32.788 32.791 qs_mol_dyn_low 1 2.0 0.003 0.003 32.193 32.201 qs_forces 11 3.9 0.002 0.002 32.135 32.135 qs_energies 11 4.9 0.011 0.043 30.555 30.562 scf_env_do_scf 11 5.9 0.001 0.001 25.719 25.719 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 23.299 23.300 dbcsr_multiply_generic 2286 12.5 0.094 0.098 17.009 17.098 qs_scf_new_mos 108 7.5 0.001 0.001 15.599 15.623 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.598 15.622 velocity_verlet 10 3.0 0.001 0.002 15.251 15.252 ot_scf_mini 108 9.5 0.002 0.003 14.834 14.853 multiply_cannon 2286 13.5 0.195 0.200 13.653 14.390 multiply_cannon_loop 2286 14.5 0.864 0.918 12.893 13.691 ot_mini 108 10.5 0.001 0.001 9.175 9.195 qs_ot_get_derivative 108 11.5 0.001 0.001 7.673 7.691 multiply_cannon_multrec 18288 15.5 1.866 2.884 6.940 7.217 rebuild_ks_matrix 119 8.3 0.000 0.000 5.983 6.006 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.982 6.005 dbcsr_mm_accdrv_process 38222 16.0 4.923 5.727 4.981 5.793 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.270 5.291 mp_waitall_1 158411 16.6 2.660 3.851 2.660 3.851 sum_up_and_integrate 119 10.3 0.030 0.031 3.573 3.577 integrate_v_rspace 119 11.3 0.003 0.003 3.542 3.550 init_scf_run 11 5.9 0.000 0.001 3.500 3.500 scf_env_initial_rho_setup 11 6.9 0.000 0.001 3.499 3.500 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.885 3.479 qs_ot_get_p 119 10.4 0.001 0.001 3.400 3.435 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.286 3.299 calculate_rho_elec 119 8.7 0.031 0.031 3.286 3.298 multiply_cannon_metrocomm3 18288 15.5 0.047 0.048 1.472 2.556 init_scf_loop 11 6.9 0.001 0.003 2.403 2.405 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.002 2.361 apply_single 119 13.6 0.000 0.000 2.001 2.361 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.280 2.286 calculate_first_density_matrix 1 7.0 0.002 0.010 2.255 2.256 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.974 1.974 make_m2s 4572 13.5 0.045 0.047 1.822 1.951 density_rs2pw 119 9.7 0.004 0.004 1.838 1.923 grid_integrate_task_list 119 12.3 1.804 1.912 1.804 1.912 calculate_dm_sparse 119 9.5 0.000 0.000 1.859 1.875 make_images 4572 14.5 0.192 0.204 1.736 1.865 pw_transfer 1439 11.6 0.065 0.069 1.845 1.855 acc_transpose_blocks 18288 15.5 0.082 0.085 1.652 1.777 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.759 1.768 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.753 1.765 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.720 1.729 cp_fm_diag_elpa_base 50 14.0 1.697 1.707 1.718 1.728 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.681 1.683 mp_sum_l 7287 12.8 1.178 1.663 1.178 1.663 prepare_preconditioner 11 7.9 0.000 0.000 1.648 1.651 make_preconditioner 11 8.9 0.000 0.001 1.648 1.650 make_full_inverse_cholesky 11 9.9 0.000 0.001 1.510 1.591 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.517 1.525 ot_diis_step 108 11.5 0.011 0.012 1.481 1.481 fft_wrap_pw1pw2_140 487 13.2 0.212 0.217 1.401 1.410 potential_pw2rs 119 12.3 0.007 0.008 1.353 1.360 fft3d_ps 1201 14.6 0.526 0.545 1.320 1.334 grid_collocate_task_list 119 9.7 1.235 1.307 1.235 1.307 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.190 1.202 wfi_extrapolate 11 7.9 0.001 0.001 1.192 1.192 transfer_rs2pw 487 10.6 0.005 0.005 1.053 1.157 multiply_cannon_sync_h2d 18288 15.5 1.010 1.129 1.010 1.129 qs_energies_init_hamiltonians 11 5.9 0.007 0.015 1.041 1.042 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.966 0.983 make_images_data 4572 15.5 0.048 0.052 0.781 0.920 acc_transpose_blocks_sync 54864 16.5 0.761 0.882 0.761 0.882 mp_alltoall_d11v 2130 13.8 0.689 0.860 0.689 0.860 hybrid_alltoall_any 4725 16.4 0.058 0.115 0.668 0.841 transfer_pw2rs 487 13.2 0.004 0.004 0.821 0.825 acc_transpose_blocks_kernels 18288 16.5 0.219 0.226 0.789 0.799 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.783 0.786 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.003 0.678 0.746 mp_alltoall_z22v 1201 16.6 0.643 0.711 0.643 0.711 cp_fm_cholesky_invert 11 10.9 0.697 0.701 0.697 0.701 transfer_rs2pw_140 130 11.5 0.118 0.121 0.583 0.686 mp_waitany 9880 13.7 0.591 0.686 0.591 0.686 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=32.791000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=494.363636, yerr=1.067940 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5080151e8b1ac516ab82332683adfb2cbf2d479f_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 558.264320E+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 0.076 0.149 38.226 38.237 qs_mol_dyn_low 1 2.0 0.003 0.006 37.497 37.660 qs_forces 11 3.9 0.003 0.005 37.433 37.434 qs_energies 11 4.9 0.004 0.009 35.703 35.707 scf_env_do_scf 11 5.9 0.001 0.002 30.201 30.202 scf_env_do_scf_inner_loop 108 6.5 0.003 0.008 26.637 26.638 dbcsr_multiply_generic 2286 12.5 0.099 0.104 20.163 20.260 velocity_verlet 10 3.0 0.002 0.002 18.956 18.958 qs_scf_new_mos 108 7.5 0.001 0.001 18.343 18.396 qs_scf_loop_do_ot 108 8.5 0.001 0.001 18.343 18.395 ot_scf_mini 108 9.5 0.002 0.003 17.292 17.340 multiply_cannon 2286 13.5 0.227 0.245 16.301 16.569 multiply_cannon_loop 2286 14.5 1.536 1.618 15.295 15.696 ot_mini 108 10.5 0.001 0.001 10.506 10.567 multiply_cannon_multrec 27432 15.5 2.450 3.124 9.073 9.367 qs_ot_get_derivative 108 11.5 0.001 0.001 8.619 8.667 dbcsr_mm_accdrv_process 47916 15.9 6.052 7.787 6.519 7.849 rebuild_ks_matrix 119 8.3 0.000 0.000 6.473 6.518 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.472 6.518 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.737 5.775 init_scf_run 11 5.9 0.000 0.001 4.097 4.097 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.097 4.097 qs_ot_get_p 119 10.4 0.001 0.001 3.903 3.969 sum_up_and_integrate 119 10.3 0.035 0.037 3.638 3.646 integrate_v_rspace 119 11.3 0.003 0.003 3.603 3.612 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.210 3.609 init_scf_loop 11 6.9 0.001 0.004 3.540 3.541 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.473 3.516 calculate_rho_elec 119 8.7 0.040 0.046 3.473 3.516 prepare_preconditioner 11 7.9 0.000 0.000 2.687 2.696 make_preconditioner 11 8.9 0.002 0.009 2.687 2.696 calculate_first_density_matrix 1 7.0 0.002 0.009 2.632 2.634 make_full_inverse_cholesky 11 9.9 0.001 0.003 2.290 2.622 mp_waitall_1 137007 16.6 1.978 2.604 1.978 2.604 make_m2s 4572 13.5 0.055 0.057 2.410 2.547 acc_transpose_blocks 27432 15.5 0.124 0.129 2.417 2.498 make_images 4572 14.5 0.274 0.336 2.302 2.436 qs_ot_p2m_diag 50 11.0 0.016 0.024 2.421 2.430 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.397 2.419 calculate_dm_sparse 119 9.5 0.000 0.000 2.201 2.260 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.127 2.250 apply_single 119 13.6 0.000 0.000 2.127 2.249 density_rs2pw 119 9.7 0.004 0.004 1.909 2.040 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.007 2.008 pw_transfer 1439 11.6 0.065 0.069 1.970 2.008 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.933 1.934 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.878 1.919 grid_integrate_task_list 119 12.3 1.834 1.909 1.834 1.909 ot_diis_step 108 11.5 0.012 0.013 1.846 1.847 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.725 1.738 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.725 1.735 cp_fm_diag_elpa_base 50 14.0 1.690 1.708 1.723 1.732 fft_wrap_pw1pw2_140 487 13.2 0.245 0.260 1.530 1.574 acc_transpose_blocks_sync 82296 16.5 1.440 1.521 1.440 1.521 multiply_cannon_metrocomm3 27432 15.5 0.040 0.041 0.778 1.440 fft3d_ps 1201 14.6 0.553 0.606 1.397 1.423 potential_pw2rs 119 12.3 0.009 0.010 1.406 1.409 wfi_extrapolate 11 7.9 0.001 0.001 1.406 1.406 grid_collocate_task_list 119 9.7 1.249 1.339 1.249 1.339 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.319 1.332 mp_sum_l 7287 12.8 1.046 1.329 1.046 1.329 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.282 1.292 transfer_rs2pw 487 10.6 0.004 0.005 1.028 1.182 cp_fm_upper_to_full 72 14.2 0.811 1.157 0.811 1.157 dbcsr_complete_redistribute 329 12.2 0.125 0.162 0.854 1.130 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.129 1.129 jit_kernel_multiply 6 16.3 0.400 1.096 0.400 1.096 make_images_data 4572 15.5 0.048 0.053 0.901 1.040 mp_alltoall_d11v 2130 13.8 0.831 0.980 0.831 0.980 hybrid_alltoall_any 4725 16.4 0.065 0.154 0.786 0.960 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 0.801 0.876 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.602 0.871 transfer_pw2rs 487 13.2 0.004 0.004 0.833 0.836 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.829 0.834 acc_transpose_blocks_kernels 27432 16.5 0.273 0.280 0.824 0.830 cp_fm_cholesky_invert 11 10.9 0.812 0.815 0.812 0.815 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=38.237000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=530.363636, yerr=3.364865 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5080151e8b1ac516ab82332683adfb2cbf2d479f_performance_tests/07/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 117.977176E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 1384136 0.0% 0.0% 100.0% average stack size 0.0 0.0 106.2 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 622.366720E+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 0.019 0.083 28.848 28.849 qs_mol_dyn_low 1 2.0 0.003 0.008 28.402 28.422 qs_forces 11 3.9 0.003 0.003 28.300 28.301 qs_energies 11 4.9 0.001 0.001 26.587 26.590 scf_env_do_scf 11 5.9 0.001 0.001 21.734 21.734 scf_env_do_scf_inner_loop 108 6.5 0.007 0.018 19.277 19.278 velocity_verlet 10 3.0 0.002 0.002 14.703 14.708 dbcsr_multiply_generic 2286 12.5 0.091 0.096 12.905 12.972 qs_scf_new_mos 108 7.5 0.001 0.001 11.688 11.710 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.687 11.709 ot_scf_mini 108 9.5 0.002 0.002 10.979 11.009 multiply_cannon 2286 13.5 0.225 0.234 10.338 10.787 multiply_cannon_loop 2286 14.5 0.640 0.664 9.446 9.603 ot_mini 108 10.5 0.001 0.001 6.257 6.292 multiply_cannon_multrec 9144 15.5 1.664 1.856 5.941 6.169 rebuild_ks_matrix 119 8.3 0.000 0.000 5.743 5.762 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.742 5.762 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.107 5.125 qs_ot_get_derivative 108 11.5 0.001 0.001 4.963 4.991 dbcsr_mm_accdrv_process 12550 15.8 3.686 4.206 4.165 4.249 sum_up_and_integrate 119 10.3 0.037 0.041 3.428 3.431 integrate_v_rspace 119 11.3 0.003 0.003 3.390 3.395 init_scf_run 11 5.9 0.000 0.001 3.384 3.384 scf_env_initial_rho_setup 11 6.9 0.001 0.002 3.384 3.384 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.312 3.317 calculate_rho_elec 119 8.7 0.060 0.061 3.312 3.317 qs_ot_get_p 119 10.4 0.001 0.001 2.846 2.880 init_scf_loop 11 6.9 0.000 0.000 2.436 2.436 calculate_first_density_matrix 1 7.0 0.000 0.000 2.239 2.240 pw_transfer 1439 11.6 0.065 0.068 1.964 1.975 grid_integrate_task_list 119 12.3 1.865 1.932 1.865 1.932 make_m2s 4572 13.5 0.034 0.035 1.773 1.905 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.871 1.882 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.848 1.850 mp_waitall_1 115863 16.7 1.324 1.823 1.324 1.823 make_images 4572 14.5 0.269 0.299 1.685 1.816 density_rs2pw 119 9.7 0.004 0.004 1.712 1.797 calculate_dm_sparse 119 9.5 0.000 0.000 1.746 1.763 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.736 1.760 prepare_preconditioner 11 7.9 0.000 0.000 1.710 1.715 make_preconditioner 11 8.9 0.000 0.000 1.710 1.715 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.601 1.626 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.602 1.603 fft_wrap_pw1pw2_140 487 13.2 0.323 0.330 1.547 1.560 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.520 1.521 acc_transpose_blocks 9144 15.5 0.042 0.044 1.418 1.455 grid_collocate_task_list 119 9.7 1.296 1.384 1.296 1.384 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.356 1.365 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.346 1.361 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.332 1.339 cp_fm_diag_elpa_base 50 14.0 1.305 1.320 1.330 1.338 fft3d_ps 1201 14.6 0.558 0.571 1.285 1.297 ot_diis_step 108 11.5 0.013 0.013 1.279 1.280 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.227 1.269 apply_single 119 13.6 0.000 0.000 1.227 1.268 qs_energies_init_hamiltonians 11 5.9 0.031 0.037 1.254 1.255 potential_pw2rs 119 12.3 0.010 0.011 1.210 1.212 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.134 1.138 wfi_extrapolate 11 7.9 0.001 0.001 1.094 1.094 jit_kernel_multiply 6 15.3 0.440 1.027 0.440 1.027 hybrid_alltoall_any 4725 16.4 0.065 0.176 0.757 0.943 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.881 0.931 make_images_data 4572 15.5 0.042 0.046 0.762 0.916 cp_fm_cholesky_invert 11 10.9 0.874 0.877 0.874 0.877 transfer_rs2pw 487 10.6 0.004 0.004 0.792 0.876 mp_alltoall_d11v 2130 13.8 0.757 0.872 0.757 0.872 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.819 0.827 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.759 0.761 acc_transpose_blocks_sync 27432 16.5 0.715 0.748 0.715 0.748 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.665 0.714 multiply_cannon_metrocomm3 9144 15.5 0.021 0.021 0.332 0.692 acc_transpose_blocks_kernels 9144 16.5 0.118 0.119 0.645 0.649 transfer_pw2rs 487 13.2 0.003 0.004 0.647 0.649 mp_alltoall_z22v 1201 16.6 0.600 0.630 0.600 0.630 mp_allgather_i34 2286 14.5 0.209 0.618 0.209 0.618 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=28.849000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=590.363636, yerr=3.574473 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5080151e8b1ac516ab82332683adfb2cbf2d479f_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 786.628608E+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 0.030 0.043 44.571 44.572 qs_mol_dyn_low 1 2.0 0.003 0.004 44.120 44.128 qs_forces 11 3.9 0.002 0.003 44.059 44.062 qs_energies 11 4.9 0.002 0.002 42.025 42.031 scf_env_do_scf 11 5.9 0.001 0.001 35.795 35.795 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 27.701 27.704 velocity_verlet 10 3.0 0.002 0.002 24.733 24.738 dbcsr_multiply_generic 2286 12.5 0.100 0.103 19.726 19.916 qs_scf_new_mos 108 7.5 0.001 0.001 18.050 18.146 qs_scf_loop_do_ot 108 8.5 0.001 0.001 18.049 18.145 ot_scf_mini 108 9.5 0.002 0.002 16.862 16.964 multiply_cannon 2286 13.5 0.335 0.365 15.454 16.428 multiply_cannon_loop 2286 14.5 0.866 0.885 14.090 15.098 ot_mini 108 10.5 0.001 0.001 10.121 10.240 multiply_cannon_multrec 9144 15.5 3.418 4.818 8.915 8.953 qs_ot_get_derivative 108 11.5 0.001 0.001 7.998 8.097 init_scf_loop 11 6.9 0.000 0.000 8.059 8.065 rebuild_ks_matrix 119 8.3 0.000 0.000 7.303 7.450 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.303 7.449 prepare_preconditioner 11 7.9 0.000 0.000 7.063 7.077 make_preconditioner 11 8.9 0.000 0.000 7.063 7.077 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.622 6.954 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.573 6.705 dbcsr_mm_accdrv_process 12550 15.8 4.550 6.082 5.355 6.638 cp_fm_upper_to_full 72 14.2 3.207 4.592 3.207 4.592 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.132 4.138 calculate_rho_elec 119 8.7 0.118 0.121 4.131 4.138 init_scf_run 11 5.9 0.000 0.001 4.100 4.101 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.100 4.100 sum_up_and_integrate 119 10.3 0.065 0.066 3.937 3.945 integrate_v_rspace 119 11.3 0.004 0.004 3.872 3.880 qs_ot_get_p 119 10.4 0.001 0.001 3.733 3.869 mp_waitall_1 94719 16.7 2.480 3.543 2.480 3.543 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.896 3.324 dbcsr_complete_redistribute 329 12.2 0.286 0.290 2.048 2.924 make_m2s 4572 13.5 0.038 0.038 2.585 2.772 pw_transfer 1439 11.6 0.068 0.069 2.704 2.708 make_images 4572 14.5 0.353 0.389 2.465 2.650 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.606 2.611 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.287 2.595 apply_single 119 13.6 0.000 0.000 2.286 2.595 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.703 2.571 calculate_first_density_matrix 1 7.0 0.000 0.000 2.569 2.570 calculate_dm_sparse 119 9.5 0.000 0.001 2.426 2.443 mp_alltoall_i22 627 13.8 1.486 2.378 1.486 2.378 multiply_cannon_metrocomm3 9144 15.5 0.021 0.022 1.383 2.334 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.436 2.299 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.246 2.248 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.169 2.220 density_rs2pw 119 9.7 0.004 0.004 2.149 2.175 fft_wrap_pw1pw2_140 487 13.2 0.575 0.578 2.166 2.172 grid_integrate_task_list 119 12.3 2.085 2.107 2.085 2.107 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.089 2.090 ot_diis_step 108 11.5 0.014 0.014 2.089 2.090 acc_transpose_blocks 9144 15.5 0.045 0.045 1.954 1.985 mp_sum_l 7287 12.8 1.140 1.916 1.140 1.916 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.866 1.867 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.803 1.805 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.742 1.767 fft3d_ps 1201 14.6 0.593 0.607 1.702 1.705 cp_fm_cholesky_invert 11 10.9 1.599 1.603 1.599 1.603 grid_collocate_task_list 119 9.7 1.525 1.556 1.525 1.556 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.538 1.539 cp_fm_diag_elpa_base 50 14.0 1.389 1.445 1.536 1.537 wfi_extrapolate 11 7.9 0.001 0.001 1.468 1.468 hybrid_alltoall_any 4725 16.4 0.090 0.150 1.158 1.434 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.412 1.421 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.359 1.384 potential_pw2rs 119 12.3 0.014 0.014 1.372 1.375 make_images_data 4572 15.5 0.047 0.049 1.114 1.370 mp_alltoall_d11v 2130 13.8 1.244 1.286 1.244 1.286 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.132 1.153 acc_transpose_blocks_sync 27432 16.5 1.105 1.126 1.105 1.126 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 0.995 1.042 mp_alltoall_z22v 1201 16.6 0.972 0.995 0.972 0.995 jit_kernel_multiply 5 15.4 0.775 0.985 0.775 0.985 qs_create_task_list 11 7.9 0.019 0.019 0.955 0.963 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.944 0.959 generate_qs_task_list 11 8.9 0.367 0.386 0.936 0.945 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=44.572000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=739.363636, yerr=13.452088 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5080151e8b1ac516ab82332683adfb2cbf2d479f_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 500.813824E+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 1419942. 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 0.043 0.110 83.037 83.043 qs_mol_dyn_low 1 2.0 0.003 0.007 82.221 82.231 qs_forces 11 3.9 0.003 0.005 82.146 82.150 qs_energies 11 4.9 0.004 0.017 79.254 79.268 scf_env_do_scf 11 5.9 0.001 0.001 70.264 70.267 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 64.807 64.808 dbcsr_multiply_generic 2055 12.4 0.106 0.109 51.414 51.797 qs_scf_new_mos 99 7.5 0.000 0.001 47.747 47.916 qs_scf_loop_do_ot 99 8.5 0.001 0.001 47.747 47.916 ot_scf_mini 99 9.5 0.002 0.003 45.284 45.389 multiply_cannon 2055 13.4 0.184 0.191 42.834 43.801 velocity_verlet 10 3.0 0.001 0.002 43.618 43.619 multiply_cannon_loop 2055 14.4 1.785 1.833 41.886 42.893 ot_mini 99 10.5 0.001 0.001 26.625 26.757 qs_ot_get_derivative 99 11.5 0.001 0.002 19.769 19.883 multiply_cannon_multrec 49320 15.4 11.371 11.971 17.342 18.084 rebuild_ks_matrix 110 8.3 0.000 0.001 14.577 14.758 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.017 14.577 14.757 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.752 12.919 mp_waitall_1 220248 16.4 11.148 11.877 11.148 11.877 multiply_cannon_sync_h2d 49320 15.4 9.605 10.193 9.605 10.193 qs_ot_get_p 110 10.4 0.001 0.002 9.987 10.172 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.664 8.194 multiply_cannon_metrocomm3 49320 15.4 0.083 0.086 6.761 7.923 apply_preconditioner_dbcsr 110 12.6 0.000 0.001 7.249 7.804 apply_single 110 13.6 0.000 0.001 7.249 7.803 sum_up_and_integrate 110 10.3 0.036 0.043 7.046 7.062 integrate_v_rspace 110 11.3 0.003 0.004 7.009 7.034 init_scf_run 11 5.9 0.000 0.001 6.798 6.798 scf_env_initial_rho_setup 11 6.9 0.001 0.003 6.797 6.798 qs_ot_p2m_diag 48 11.0 0.013 0.018 6.596 6.637 ot_diis_step 99 11.5 0.006 0.007 6.627 6.628 qs_rho_update_rho_low 110 7.6 0.001 0.002 6.309 6.442 calculate_rho_elec 110 8.6 0.021 0.025 6.308 6.441 dbcsr_mm_accdrv_process 87628 16.1 3.024 3.123 5.841 6.136 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.676 5.676 init_scf_loop 11 6.9 0.001 0.003 5.431 5.432 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.287 5.361 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.109 5.132 cp_fm_diag_elpa_base 48 14.0 5.095 5.120 5.107 5.130 mp_sum_l 6594 12.7 4.268 4.995 4.268 4.995 wfi_extrapolate 11 7.9 0.002 0.012 4.043 4.043 make_m2s 4110 13.4 0.061 0.066 3.920 4.027 calculate_dm_sparse 110 9.5 0.001 0.001 3.827 3.939 make_images 4110 14.4 0.176 0.189 3.824 3.935 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.532 3.536 density_rs2pw 110 9.6 0.004 0.005 3.339 3.484 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.396 3.482 multiply_cannon_metrocomm1 49320 15.4 0.065 0.068 2.522 3.471 grid_integrate_task_list 110 12.3 3.256 3.393 3.256 3.393 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.188 3.268 prepare_preconditioner 11 7.9 0.000 0.000 3.219 3.243 make_preconditioner 11 8.9 0.000 0.002 3.219 3.243 pw_transfer 1331 11.6 0.055 0.071 3.099 3.187 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.010 3.096 make_full_inverse_cholesky 11 9.9 0.000 0.001 2.983 3.040 mp_alltoall_d11v 2046 13.8 2.085 2.676 2.085 2.676 fft_wrap_pw1pw2_140 451 13.1 0.389 0.430 2.577 2.666 calculate_first_density_matrix 1 7.0 0.004 0.015 2.653 2.660 potential_pw2rs 110 12.3 0.006 0.006 2.623 2.645 acc_transpose_blocks 49320 15.4 0.219 0.230 2.555 2.621 jit_kernel_multiply 13 15.9 2.534 2.548 2.534 2.548 fft3d_ps 1111 14.6 0.790 0.884 2.307 2.374 grid_collocate_task_list 110 9.6 2.154 2.294 2.154 2.294 mp_waitany 14300 13.8 1.896 2.136 1.896 2.136 transfer_rs2pw 451 10.6 0.005 0.006 1.962 2.127 mp_sum_d 3889 11.9 1.447 2.100 1.447 2.100 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.957 1.971 make_images_data 4110 15.4 0.044 0.047 1.791 1.916 hybrid_alltoall_any 4261 16.3 0.084 0.481 1.543 1.803 transfer_pw2rs 451 13.1 0.006 0.007 1.756 1.767 cp_fm_cholesky_invert 11 10.9 1.749 1.753 1.749 1.753 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=83.043000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=475.363636, yerr=2.568081 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5080151e8b1ac516ab82332683adfb2cbf2d479f_performance_tests/10/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 390.715586E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 5019072 0.0% 0.0% 100.0% average stack size 0.0 0.0 196.1 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 594.030592E+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 3592521. 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 0.063 0.198 70.069 70.071 qs_mol_dyn_low 1 2.0 0.003 0.003 69.454 69.480 qs_forces 11 3.9 0.003 0.003 69.102 69.103 qs_energies 11 4.9 0.008 0.022 65.767 65.771 scf_env_do_scf 11 5.9 0.001 0.002 56.957 56.960 scf_env_do_scf_inner_loop 99 6.5 0.002 0.008 49.333 49.334 dbcsr_multiply_generic 2055 12.4 0.114 0.121 37.946 38.091 velocity_verlet 10 3.0 0.001 0.002 36.452 36.459 qs_scf_new_mos 99 7.5 0.001 0.001 33.406 33.528 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.405 33.528 multiply_cannon 2055 13.4 0.227 0.257 31.289 32.341 ot_scf_mini 99 9.5 0.003 0.004 31.737 31.851 multiply_cannon_loop 2055 14.4 1.163 1.190 30.042 30.967 ot_mini 99 10.5 0.001 0.001 18.426 18.545 multiply_cannon_multrec 24660 15.4 6.975 8.713 14.019 15.649 rebuild_ks_matrix 110 8.3 0.000 0.000 13.582 13.699 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.016 13.581 13.698 qs_ot_get_derivative 99 11.5 0.001 0.001 12.646 12.761 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.948 12.054 mp_waitall_1 176588 16.5 7.533 9.908 7.533 9.908 multiply_cannon_metrocomm3 24660 15.4 0.072 0.075 5.100 7.857 init_scf_loop 11 6.9 0.001 0.003 7.587 7.587 multiply_cannon_sync_h2d 24660 15.4 6.394 7.402 6.394 7.402 dbcsr_mm_accdrv_process 52282 16.1 5.535 6.278 6.873 7.210 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.469 7.178 apply_single 110 13.6 0.000 0.001 6.469 7.178 qs_ot_get_p 110 10.4 0.001 0.001 6.667 6.822 sum_up_and_integrate 110 10.3 0.053 0.058 6.422 6.433 integrate_v_rspace 110 11.3 0.003 0.003 6.369 6.383 init_scf_run 11 5.9 0.000 0.001 6.209 6.210 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.209 6.210 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.952 5.963 calculate_rho_elec 110 8.6 0.039 0.047 5.952 5.963 ot_diis_step 99 11.5 0.010 0.011 5.732 5.733 prepare_preconditioner 11 7.9 0.000 0.001 5.546 5.564 make_preconditioner 11 8.9 0.000 0.002 5.546 5.564 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.746 5.464 make_full_inverse_cholesky 11 9.9 0.000 0.001 5.117 5.270 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.754 4.776 make_m2s 4110 13.4 0.057 0.060 4.254 4.665 make_images 4110 14.4 0.409 0.468 4.146 4.553 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.284 4.285 pw_transfer 1331 11.6 0.066 0.076 3.591 3.734 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.701 3.714 cp_fm_diag_elpa_base 48 14.0 3.651 3.664 3.698 3.710 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.484 3.628 wfi_extrapolate 11 7.9 0.001 0.001 3.514 3.514 density_rs2pw 110 9.6 0.004 0.005 3.192 3.374 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.305 3.372 grid_integrate_task_list 110 12.3 3.165 3.302 3.165 3.302 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.295 3.296 fft_wrap_pw1pw2_140 451 13.1 0.459 0.472 2.959 3.102 calculate_dm_sparse 110 9.5 0.001 0.001 2.974 3.000 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.944 2.975 hybrid_alltoall_any 4261 16.3 0.108 0.464 1.979 2.770 fft3d_ps 1111 14.6 1.110 1.346 2.582 2.750 make_images_data 4110 15.4 0.049 0.055 2.300 2.744 cp_fm_cholesky_invert 11 10.9 2.597 2.604 2.597 2.604 calculate_first_density_matrix 1 7.0 0.001 0.005 2.591 2.594 mp_sum_l 6594 12.7 1.798 2.462 1.798 2.462 grid_collocate_task_list 110 9.6 2.167 2.349 2.167 2.349 potential_pw2rs 110 12.3 0.008 0.009 2.271 2.281 jit_kernel_multiply 11 16.2 0.972 2.133 0.972 2.133 qs_energies_init_hamiltonians 11 5.9 0.016 0.049 2.052 2.053 acc_transpose_blocks 24660 15.4 0.118 0.121 1.976 2.003 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.966 1.988 mp_alltoall_d11v 2046 13.8 1.754 1.941 1.754 1.941 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.794 1.811 transfer_rs2pw 451 10.6 0.006 0.007 1.520 1.765 multiply_cannon_metrocomm4 22605 15.4 0.078 0.082 0.772 1.563 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.544 1.556 mp_allgather_i34 2055 14.4 0.554 1.543 0.554 1.543 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.389 1.493 mp_irecv_dv 57340 16.2 0.644 1.443 0.644 1.443 mp_waitany 10164 13.8 1.256 1.438 1.256 1.438 dbcsr_complete_redistribute 325 12.2 0.239 0.302 1.134 1.409 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=70.071000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=561.181818, yerr=8.031961 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5080151e8b1ac516ab82332683adfb2cbf2d479f_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 668.270592E+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 0.029 0.037 60.199 60.200 qs_mol_dyn_low 1 2.0 0.003 0.004 59.583 59.593 qs_forces 11 3.9 0.003 0.004 59.503 59.508 qs_energies 11 4.9 0.001 0.002 56.302 56.310 scf_env_do_scf 11 5.9 0.000 0.001 48.132 48.132 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 39.638 39.639 velocity_verlet 10 3.0 0.002 0.002 32.403 32.405 dbcsr_multiply_generic 2055 12.4 0.105 0.110 28.652 28.895 qs_scf_new_mos 99 7.5 0.001 0.001 25.074 25.174 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.073 25.174 ot_scf_mini 99 9.5 0.002 0.002 23.826 23.936 multiply_cannon 2055 13.4 0.214 0.222 22.423 23.600 multiply_cannon_loop 2055 14.4 0.810 0.865 21.309 22.351 ot_mini 99 10.5 0.001 0.001 13.625 13.735 rebuild_ks_matrix 110 8.3 0.000 0.000 12.100 12.238 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.099 12.238 multiply_cannon_multrec 16440 15.4 3.716 4.731 9.939 10.876 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.649 10.780 mp_waitall_1 139946 16.5 6.802 9.789 6.802 9.789 qs_ot_get_derivative 99 11.5 0.001 0.001 9.216 9.325 init_scf_loop 11 6.9 0.000 0.000 8.461 8.462 multiply_cannon_metrocomm3 16440 15.4 0.046 0.047 4.219 7.047 prepare_preconditioner 11 7.9 0.000 0.000 6.726 6.744 make_preconditioner 11 8.9 0.000 0.000 6.726 6.744 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.074 6.409 sum_up_and_integrate 110 10.3 0.061 0.063 6.291 6.306 dbcsr_mm_accdrv_process 34862 16.1 5.441 5.688 6.068 6.246 integrate_v_rspace 110 11.3 0.003 0.003 6.230 6.245 init_scf_run 11 5.9 0.000 0.001 5.734 5.734 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.734 5.734 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.693 5.705 calculate_rho_elec 110 8.6 0.059 0.059 5.692 5.705 qs_ot_get_p 110 10.4 0.001 0.001 5.390 5.529 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.920 5.325 apply_single 110 13.6 0.000 0.000 4.920 5.325 make_m2s 4110 13.4 0.050 0.051 4.164 4.495 ot_diis_step 99 11.5 0.011 0.011 4.380 4.380 make_images 4110 14.4 0.397 0.515 4.050 4.379 multiply_cannon_sync_h2d 16440 15.4 3.258 3.982 3.258 3.982 qs_ot_p2m_diag 48 11.0 0.041 0.044 3.760 3.765 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.131 3.761 pw_transfer 1331 11.6 0.065 0.073 3.558 3.564 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.451 3.459 grid_integrate_task_list 110 12.3 3.191 3.417 3.191 3.417 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.394 3.394 density_rs2pw 110 9.6 0.004 0.005 2.937 3.163 fft_wrap_pw1pw2_140 451 13.1 0.579 0.586 2.978 2.988 wfi_extrapolate 11 7.9 0.001 0.001 2.951 2.951 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.872 2.874 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.860 2.869 cp_fm_diag_elpa_base 48 14.0 2.796 2.823 2.858 2.867 make_images_data 4110 15.4 0.046 0.050 2.344 2.768 hybrid_alltoall_any 4261 16.3 0.109 0.380 2.106 2.720 calculate_first_density_matrix 1 7.0 0.000 0.000 2.692 2.694 cp_fm_cholesky_invert 11 10.9 2.639 2.644 2.639 2.644 calculate_dm_sparse 110 9.5 0.001 0.001 2.533 2.568 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.520 2.566 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.441 2.510 multiply_cannon_metrocomm4 14385 15.4 0.048 0.051 0.869 2.429 fft3d_ps 1111 14.6 1.090 1.097 2.392 2.405 grid_collocate_task_list 110 9.6 2.220 2.378 2.220 2.378 mp_irecv_dv 48980 15.7 0.795 2.301 0.795 2.301 mp_alltoall_d11v 2046 13.8 1.792 2.145 1.792 2.145 potential_pw2rs 110 12.3 0.011 0.012 2.115 2.120 mp_sum_l 6594 12.7 1.451 2.033 1.451 2.033 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.998 2.004 dbcsr_complete_redistribute 325 12.2 0.325 0.361 1.435 1.894 cp_fm_upper_to_full 70 14.2 1.359 1.713 1.359 1.713 acc_transpose_blocks 16440 15.4 0.080 0.082 1.578 1.661 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.644 1.659 cp_fm_cholesky_decompose 22 10.9 1.568 1.588 1.568 1.588 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.366 1.507 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.472 1.487 mp_allgather_i34 2055 14.4 0.426 1.478 0.426 1.478 transfer_rs2pw 451 10.6 0.005 0.006 1.255 1.462 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.972 1.409 mp_waitany 17072 13.8 1.132 1.322 1.132 1.322 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.275 1.287 rs_gather_matrices 110 12.3 0.138 0.152 0.874 1.204 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=60.200000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=634.272727, yerr=7.688136 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5080151e8b1ac516ab82332683adfb2cbf2d479f_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 728.076288E+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 0.049 0.183 69.088 69.090 qs_mol_dyn_low 1 2.0 0.003 0.003 68.173 68.537 qs_forces 11 3.9 0.005 0.008 67.558 67.559 qs_energies 11 4.9 0.002 0.009 64.123 64.128 scf_env_do_scf 11 5.9 0.001 0.002 55.351 55.354 scf_env_do_scf_inner_loop 99 6.5 0.003 0.006 43.197 43.198 velocity_verlet 10 3.0 0.002 0.002 39.008 39.032 dbcsr_multiply_generic 2055 12.4 0.112 0.119 31.030 31.260 qs_scf_new_mos 99 7.5 0.001 0.001 28.238 28.345 qs_scf_loop_do_ot 99 8.5 0.001 0.001 28.238 28.344 ot_scf_mini 99 9.5 0.003 0.003 26.531 26.638 multiply_cannon 2055 13.4 0.237 0.254 23.548 24.353 multiply_cannon_loop 2055 14.4 1.414 1.509 22.074 22.773 ot_mini 99 10.5 0.001 0.001 15.093 15.211 multiply_cannon_multrec 24660 15.4 4.090 6.716 12.997 14.164 rebuild_ks_matrix 110 8.3 0.000 0.000 12.200 12.300 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.200 12.300 init_scf_loop 11 6.9 0.001 0.002 12.107 12.108 qs_ot_get_derivative 99 11.5 0.001 0.001 10.883 10.980 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.782 10.874 prepare_preconditioner 11 7.9 0.000 0.000 10.292 10.309 make_preconditioner 11 8.9 0.000 0.001 10.291 10.309 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.487 9.951 dbcsr_mm_accdrv_process 52304 16.0 7.764 8.927 8.750 9.724 sum_up_and_integrate 110 10.3 0.068 0.071 6.432 6.445 qs_ot_get_p 110 10.4 0.001 0.001 6.243 6.406 integrate_v_rspace 110 11.3 0.003 0.003 6.364 6.378 mp_waitall_1 121746 16.5 4.429 6.234 4.429 6.234 make_m2s 4110 13.4 0.060 0.062 5.750 6.115 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.059 6.074 calculate_rho_elec 110 8.6 0.078 0.081 6.059 6.073 make_images 4110 14.4 0.577 0.700 5.608 5.969 init_scf_run 11 5.9 0.000 0.001 5.865 5.865 scf_env_initial_rho_setup 11 6.9 0.012 0.013 5.864 5.865 cp_fm_upper_to_full 70 14.2 3.338 4.846 3.338 4.846 qs_ot_p2m_diag 48 11.0 0.055 0.064 4.259 4.275 ot_diis_step 99 11.5 0.011 0.012 4.165 4.167 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.069 4.152 apply_single 110 13.6 0.000 0.000 4.069 4.152 dbcsr_complete_redistribute 325 12.2 0.415 0.457 2.784 3.937 pw_transfer 1331 11.6 0.065 0.074 3.814 3.846 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.707 3.743 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.689 3.689 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.636 3.687 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.434 3.488 grid_integrate_task_list 110 12.3 3.276 3.425 3.276 3.425 copy_fm_to_dbcsr 174 11.2 0.001 0.003 2.201 3.351 hybrid_alltoall_any 4261 16.3 0.123 0.465 2.484 3.321 density_rs2pw 110 9.6 0.004 0.004 3.133 3.303 make_images_data 4110 15.4 0.049 0.053 2.893 3.285 fft_wrap_pw1pw2_140 451 13.1 0.608 0.630 3.138 3.173 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.126 3.136 cp_fm_diag_elpa_base 48 14.0 2.972 3.034 3.123 3.134 wfi_extrapolate 11 7.9 0.001 0.001 3.126 3.126 calculate_dm_sparse 110 9.5 0.001 0.001 3.008 3.050 multiply_cannon_metrocomm3 24660 15.4 0.038 0.039 1.304 2.986 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.796 2.923 cp_fm_cholesky_invert 11 10.9 2.890 2.900 2.890 2.900 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.866 2.868 mp_alltoall_i22 605 13.7 1.675 2.851 1.675 2.851 calculate_first_density_matrix 1 7.0 0.000 0.002 2.628 2.640 fft3d_ps 1111 14.6 1.085 1.118 2.591 2.606 acc_transpose_blocks 24660 15.4 0.117 0.121 2.429 2.568 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.463 2.501 multiply_cannon_sync_h2d 24660 15.4 2.349 2.445 2.349 2.445 grid_collocate_task_list 110 9.6 2.264 2.394 2.264 2.394 qs_energies_init_hamiltonians 11 5.9 0.062 0.083 2.338 2.341 potential_pw2rs 110 12.3 0.013 0.013 2.171 2.180 mp_alltoall_d11v 2046 13.8 1.979 2.146 1.979 2.146 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.820 1.855 cp_fm_cholesky_decompose 22 10.9 1.806 1.852 1.806 1.852 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.614 1.708 mp_sum_l 6594 12.7 1.107 1.703 1.107 1.703 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.663 1.674 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.548 1.563 acc_transpose_blocks_sync 73980 16.4 1.410 1.539 1.410 1.539 transfer_rs2pw 451 10.6 0.005 0.005 1.307 1.500 multiply_cannon_metrocomm4 20550 15.4 0.062 0.065 0.860 1.476 mp_irecv_dv 62702 16.1 0.755 1.387 0.755 1.387 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=69.090000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=690.818182, yerr=7.221410 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5080151e8b1ac516ab82332683adfb2cbf2d479f_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 858.124288E+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 0.037 0.039 56.448 56.450 qs_mol_dyn_low 1 2.0 0.003 0.003 55.828 55.838 qs_forces 11 3.9 0.003 0.004 55.762 55.763 qs_energies 11 4.9 0.001 0.001 52.089 52.093 scf_env_do_scf 11 5.9 0.000 0.001 43.502 43.503 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 35.641 35.642 velocity_verlet 10 3.0 0.002 0.002 31.593 31.596 dbcsr_multiply_generic 2055 12.4 0.103 0.108 23.464 23.654 qs_scf_new_mos 99 7.5 0.001 0.001 21.054 21.111 qs_scf_loop_do_ot 99 8.5 0.001 0.001 21.053 21.110 ot_scf_mini 99 9.5 0.002 0.002 19.788 19.812 multiply_cannon 2055 13.4 0.249 0.267 17.754 19.299 multiply_cannon_loop 2055 14.4 0.599 0.620 16.457 16.884 rebuild_ks_matrix 110 8.3 0.000 0.000 11.678 11.700 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 11.678 11.700 ot_mini 99 10.5 0.001 0.001 10.919 10.935 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.379 10.398 multiply_cannon_multrec 8220 15.4 3.182 4.349 7.641 8.523 init_scf_loop 11 6.9 0.000 0.000 7.812 7.813 mp_waitall_1 103326 16.6 5.964 7.490 5.964 7.490 qs_ot_get_derivative 99 11.5 0.001 0.001 7.210 7.232 sum_up_and_integrate 110 10.3 0.080 0.081 6.262 6.275 integrate_v_rspace 110 11.3 0.003 0.004 6.182 6.195 prepare_preconditioner 11 7.9 0.000 0.000 6.141 6.145 make_preconditioner 11 8.9 0.000 0.000 6.141 6.145 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.979 5.995 calculate_rho_elec 110 8.6 0.114 0.114 5.978 5.995 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.714 5.787 init_scf_run 11 5.9 0.000 0.001 5.360 5.360 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.359 5.359 dbcsr_mm_accdrv_process 17442 15.9 3.174 4.117 4.321 5.238 qs_ot_get_p 110 10.4 0.001 0.001 4.890 4.908 make_m2s 4110 13.4 0.038 0.039 4.306 4.588 multiply_cannon_metrocomm3 8220 15.4 0.019 0.020 2.994 4.567 make_images 4110 14.4 0.650 0.709 4.177 4.459 pw_transfer 1331 11.6 0.065 0.072 3.864 3.885 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.757 3.781 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.736 3.772 apply_single 110 13.6 0.000 0.000 3.736 3.772 ot_diis_step 99 11.5 0.012 0.012 3.686 3.686 grid_integrate_task_list 110 12.3 3.374 3.536 3.374 3.536 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.503 3.508 fft_wrap_pw1pw2_140 451 13.1 0.775 0.786 3.242 3.268 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.162 3.163 density_rs2pw 110 9.6 0.004 0.004 2.949 3.087 cp_fm_cholesky_invert 11 10.9 2.957 2.960 2.957 2.960 hybrid_alltoall_any 4261 16.3 0.200 0.854 2.322 2.782 wfi_extrapolate 11 7.9 0.001 0.001 2.767 2.767 qs_energies_init_hamiltonians 11 5.9 0.010 0.014 2.750 2.751 make_images_data 4110 15.4 0.041 0.047 2.317 2.678 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.578 2.585 cp_fm_diag_elpa_base 48 14.0 2.522 2.548 2.576 2.583 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.579 2.580 grid_collocate_task_list 110 9.6 2.367 2.565 2.367 2.565 calculate_dm_sparse 110 9.5 0.001 0.001 2.487 2.525 calculate_first_density_matrix 1 7.0 0.000 0.000 2.497 2.498 multiply_cannon_sync_h2d 8220 15.4 2.381 2.461 2.381 2.461 fft3d_ps 1111 14.6 1.137 1.163 2.423 2.442 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.215 2.227 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.066 2.093 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.791 2.001 potential_pw2rs 110 12.3 0.015 0.016 1.993 1.999 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.955 1.969 mp_alltoall_d11v 2046 13.8 1.785 1.959 1.785 1.959 cp_fm_cholesky_decompose 22 10.9 1.712 1.741 1.712 1.741 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.573 1.687 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.627 1.632 mp_allgather_i34 2055 14.4 0.457 1.578 0.457 1.578 dbcsr_complete_redistribute 325 12.2 0.560 0.588 1.478 1.572 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.432 1.442 acc_transpose_blocks 8220 15.4 0.040 0.041 1.348 1.370 transfer_rs2pw 451 10.6 0.005 0.005 1.182 1.357 qs_create_task_list 11 7.9 0.002 0.003 1.213 1.310 generate_qs_task_list 11 8.9 0.373 0.438 1.211 1.309 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.236 1.260 mp_waitany 9240 13.8 1.089 1.249 1.089 1.249 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=56.450000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=810.272727, yerr=10.913635 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5080151e8b1ac516ab82332683adfb2cbf2d479f_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 1.384022E+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 0.022 0.050 87.482 87.489 qs_mol_dyn_low 1 2.0 0.003 0.003 86.934 86.944 qs_forces 11 3.9 0.003 0.004 86.729 86.730 qs_energies 11 4.9 0.001 0.001 82.595 82.598 scf_env_do_scf 11 5.9 0.001 0.001 72.451 72.451 velocity_verlet 10 3.0 0.002 0.002 55.755 55.778 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 44.190 44.192 dbcsr_multiply_generic 2055 12.4 0.118 0.124 29.623 29.725 init_scf_loop 11 6.9 0.000 0.000 28.189 28.193 qs_scf_new_mos 99 7.5 0.001 0.001 26.902 26.950 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.901 26.950 prepare_preconditioner 11 7.9 0.000 0.000 26.199 26.207 make_preconditioner 11 8.9 0.000 0.000 26.198 26.207 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.409 25.669 ot_scf_mini 99 9.5 0.002 0.002 25.085 25.126 multiply_cannon 2055 13.4 0.333 0.355 22.616 23.307 multiply_cannon_loop 2055 14.4 0.804 0.833 20.849 21.147 cp_fm_upper_to_full 70 14.2 12.646 18.154 12.646 18.154 ot_mini 99 10.5 0.001 0.001 14.174 14.215 rebuild_ks_matrix 110 8.3 0.000 0.001 13.646 13.701 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 13.646 13.700 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.373 12.423 dbcsr_complete_redistribute 325 12.2 1.023 1.054 7.355 10.570 multiply_cannon_multrec 8220 15.4 4.057 4.234 9.766 9.877 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.338 9.564 qs_ot_get_derivative 99 11.5 0.001 0.001 9.514 9.558 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.774 8.979 mp_alltoall_i22 605 13.7 5.410 8.606 5.410 8.606 mp_waitall_1 84994 16.7 7.553 8.432 7.553 8.432 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.054 7.082 calculate_rho_elec 110 8.6 0.224 0.225 7.054 7.081 sum_up_and_integrate 110 10.3 0.150 0.151 6.847 6.862 integrate_v_rspace 110 11.3 0.004 0.004 6.697 6.711 init_scf_run 11 5.9 0.000 0.001 5.859 5.859 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.859 5.859 dbcsr_mm_accdrv_process 11614 15.7 3.930 4.125 5.561 5.773 make_m2s 4110 13.4 0.044 0.044 5.245 5.721 qs_ot_get_p 110 10.4 0.001 0.001 5.601 5.657 make_images 4110 14.4 0.887 0.944 5.056 5.532 cp_fm_cholesky_invert 11 10.9 5.255 5.259 5.255 5.259 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 4.982 5.258 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.773 5.237 apply_single 110 13.6 0.000 0.000 4.773 5.237 pw_transfer 1331 11.6 0.074 0.075 4.884 4.891 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 4.767 4.774 ot_diis_step 99 11.5 0.015 0.016 4.637 4.638 fft_wrap_pw1pw2_140 451 13.1 1.277 1.281 4.169 4.179 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.039 4.047 grid_integrate_task_list 110 12.3 3.687 3.748 3.687 3.748 qs_energies_init_hamiltonians 11 5.9 0.002 0.004 3.678 3.679 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.523 3.523 density_rs2pw 110 9.6 0.004 0.004 3.502 3.518 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.009 3.465 hybrid_alltoall_any 4261 16.3 0.263 0.562 2.809 3.432 calculate_dm_sparse 110 9.5 0.001 0.001 3.308 3.336 make_images_data 4110 15.4 0.046 0.049 2.718 3.333 wfi_extrapolate 11 7.9 0.001 0.001 3.311 3.311 multiply_cannon_sync_h2d 8220 15.4 3.159 3.185 3.159 3.185 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.934 2.934 cp_fm_diag_elpa_base 48 14.0 2.392 2.593 2.932 2.932 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.887 2.892 fft3d_ps 1111 14.6 1.297 1.300 2.821 2.831 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.672 2.696 grid_collocate_task_list 110 9.6 2.675 2.685 2.675 2.685 calculate_first_density_matrix 1 7.0 0.000 0.000 2.432 2.436 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.379 2.399 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.200 2.243 potential_pw2rs 110 12.3 0.021 0.022 2.237 2.241 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.108 2.180 mp_alltoall_d11v 2046 13.8 2.044 2.106 2.044 2.106 cp_fm_cholesky_decompose 22 10.9 1.984 1.999 1.984 1.999 qs_create_task_list 11 7.9 0.001 0.001 1.877 1.919 generate_qs_task_list 11 8.9 0.730 0.781 1.876 1.918 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.874 1.879 acc_transpose_blocks 8220 15.4 0.041 0.041 1.774 1.824 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.765 1.811 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.790 1.805 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=87.489000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1240.272727, yerr=64.458713 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5080151e8b1ac516ab82332683adfb2cbf2d479f_performance_tests/15/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 1992001093632 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 15239162695680 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239162695680 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 94.232994E+12 0.0% 0.0% 100.0% flops max/rank 1.103378E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806382528 0.0% 0.0% 100.0% number of processed stacks 12049536 0.0% 0.0% 100.0% average stack size 0.0 0.0 564.9 marketing flops 145.651870E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 632.307712E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10439712 MPI messages size (bytes): total size 4.526311E+12 min size 0.000000E+00 max size 4.537280E+06 average size 433.566688E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3611304 96569262080 32768 < size <= 131072 1307392 74803314688 131072 < size <= 4194304 5190114 3200149557592 4194304 < size <= 16777216 263934 1154745839544 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 57182. MP_Allreduce 11186 792. MP_Sync 88 MP_Alltoall 2242 2917195. 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 0.039 0.084 208.789 208.791 qs_mol_dyn_low 1 2.0 0.015 0.026 207.793 207.808 qs_forces 11 3.9 0.005 0.005 207.440 207.442 qs_energies 11 4.9 0.008 0.041 201.728 201.765 scf_env_do_scf 11 5.9 0.001 0.002 185.220 185.224 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 164.753 164.754 dbcsr_multiply_generic 2529 12.6 0.178 0.184 126.413 126.798 qs_scf_new_mos 118 7.6 0.001 0.001 125.168 125.362 qs_scf_loop_do_ot 118 8.6 0.001 0.001 125.168 125.361 velocity_verlet 10 3.0 0.003 0.017 125.101 125.106 ot_scf_mini 118 9.6 0.003 0.004 118.387 118.606 multiply_cannon 2529 13.6 0.243 0.252 102.300 103.665 multiply_cannon_loop 2529 14.6 2.426 2.485 100.063 101.632 ot_mini 118 10.6 0.001 0.001 66.761 66.968 multiply_cannon_multrec 60696 15.6 31.962 33.327 41.795 43.504 qs_ot_get_derivative 118 11.6 0.001 0.001 41.861 42.035 rebuild_ks_matrix 129 8.3 0.001 0.001 33.849 34.147 qs_ks_build_kohn_sham_matrix 129 9.3 0.031 0.138 33.848 34.146 mp_waitall_1 269446 16.5 28.986 32.120 28.986 32.120 qs_ks_update_qs_env 129 7.6 0.001 0.001 30.297 30.574 qs_ot_get_p 129 10.4 0.001 0.001 29.387 29.654 multiply_cannon_sync_h2d 60696 15.6 26.606 28.466 26.606 28.466 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 24.454 25.126 apply_single 129 13.6 0.001 0.001 24.453 25.126 ot_diis_step 118 11.6 0.008 0.008 24.665 24.666 qs_ot_p2m_diag 84 11.4 0.080 0.092 22.758 22.851 init_scf_loop 11 6.9 0.001 0.003 20.393 20.394 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 20.175 20.360 cp_dbcsr_syevd 84 12.4 0.005 0.005 19.855 19.855 multiply_cannon_metrocomm3 60696 15.6 0.118 0.123 16.181 18.865 cp_fm_diag_elpa 84 13.4 0.000 0.000 16.897 16.937 cp_fm_diag_elpa_base 84 14.4 16.830 16.884 16.894 16.934 prepare_preconditioner 11 7.9 0.000 0.000 15.822 15.852 make_preconditioner 11 8.9 0.000 0.001 15.822 15.852 make_full_inverse_cholesky 11 9.9 0.000 0.001 15.052 15.242 make_m2s 5058 13.6 0.105 0.112 13.919 14.290 sum_up_and_integrate 129 10.3 0.090 0.107 14.097 14.197 make_images 5058 14.6 0.399 0.418 13.736 14.118 integrate_v_rspace 129 11.3 0.004 0.005 14.007 14.113 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.127 13.296 calculate_rho_elec 129 8.7 0.045 0.064 13.126 13.295 init_scf_run 11 5.9 0.000 0.001 12.309 12.310 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.309 12.310 mp_sum_l 8016 12.9 8.940 10.097 8.940 10.097 dbcsr_mm_accdrv_process 125516 16.2 4.748 4.906 9.396 9.904 wfi_extrapolate 11 7.9 0.001 0.001 9.061 9.062 cp_fm_cholesky_invert 11 10.9 9.015 9.022 9.015 9.022 calculate_dm_sparse 129 9.5 0.001 0.001 8.588 8.699 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.292 8.415 qs_ot_get_orbitals 118 10.6 0.001 0.001 8.094 8.219 multiply_cannon_metrocomm1 60696 15.6 0.096 0.101 6.260 8.204 pw_transfer 1559 11.6 0.074 0.095 7.672 7.917 make_images_data 5058 15.6 0.068 0.075 6.902 7.745 fft_wrap_pw1pw2 1301 12.7 0.010 0.013 7.468 7.707 grid_integrate_task_list 129 12.3 7.151 7.506 7.151 7.506 hybrid_alltoall_any 5245 16.5 0.297 2.277 6.027 7.250 density_rs2pw 129 9.7 0.006 0.007 6.712 7.175 fft_wrap_pw1pw2_140 527 13.2 1.145 1.186 6.553 6.798 cp_dbcsr_sm_fm_multiply 37 9.5 0.003 0.004 6.622 6.635 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.790 5.852 fft3d_ps 1301 14.7 2.208 2.853 5.448 5.834 mp_alltoall_d11v 2429 14.1 4.288 5.592 4.288 5.592 grid_collocate_task_list 129 9.7 4.880 5.232 4.880 5.232 cp_fm_cholesky_decompose 22 10.9 4.627 4.642 4.627 4.642 potential_pw2rs 129 12.3 0.009 0.010 4.512 4.619 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=208.791000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=597.272727, yerr=6.510005 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5080151e8b1ac516ab82332683adfb2cbf2d479f_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 840.237056E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100943E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834845264 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 58214. MP_Allreduce 11083 960. MP_Sync 87 MP_Alltoall 1969 5138969. 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 0.062 0.138 189.397 189.399 qs_mol_dyn_low 1 2.0 0.007 0.029 188.725 188.739 qs_forces 11 3.9 0.009 0.046 188.616 188.621 qs_energies 11 4.9 0.007 0.047 181.750 181.766 scf_env_do_scf 11 5.9 0.001 0.002 165.166 165.176 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 132.558 132.560 velocity_verlet 10 3.0 0.001 0.002 118.475 118.476 dbcsr_multiply_generic 2507 12.6 0.186 0.193 97.048 98.336 qs_scf_new_mos 117 7.6 0.001 0.001 94.208 94.819 qs_scf_loop_do_ot 117 8.6 0.001 0.001 94.207 94.818 ot_scf_mini 117 9.6 0.004 0.005 89.414 90.073 multiply_cannon 2507 13.6 0.505 0.562 76.829 80.998 multiply_cannon_loop 2507 14.6 1.574 1.642 73.889 76.869 ot_mini 117 10.6 0.001 0.001 50.393 51.020 mp_waitall_1 214728 16.6 24.125 37.750 24.125 37.750 multiply_cannon_multrec 30084 15.6 20.798 25.645 31.439 36.765 rebuild_ks_matrix 128 8.3 0.001 0.001 32.213 32.896 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.019 32.213 32.895 init_scf_loop 11 6.9 0.001 0.004 32.518 32.521 qs_ks_update_qs_env 128 7.6 0.001 0.001 28.870 29.495 qs_ot_get_derivative 117 11.6 0.001 0.002 28.395 29.069 multiply_cannon_metrocomm3 30084 15.6 0.097 0.102 15.913 28.501 prepare_preconditioner 11 7.9 0.000 0.001 28.218 28.299 make_preconditioner 11 8.9 0.000 0.002 28.218 28.299 make_full_inverse_cholesky 11 9.9 0.000 0.001 26.882 27.412 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.261 23.511 apply_single 128 13.6 0.001 0.001 22.260 23.511 qs_ot_get_p 128 10.4 0.001 0.002 21.247 21.998 ot_diis_step 117 11.6 0.015 0.015 21.820 21.822 multiply_cannon_sync_h2d 30084 15.6 17.805 20.228 17.805 20.228 cp_fm_cholesky_invert 11 10.9 16.489 16.502 16.489 16.502 qs_ot_p2m_diag 83 11.4 0.188 0.217 16.330 16.364 make_m2s 5014 13.6 0.089 0.093 14.074 15.347 make_images 5014 14.6 1.173 1.368 13.866 15.139 cp_dbcsr_syevd 83 12.4 0.006 0.006 15.126 15.127 sum_up_and_integrate 128 10.3 0.117 0.134 14.121 14.163 integrate_v_rspace 128 11.3 0.004 0.005 14.004 14.051 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.329 13.370 calculate_rho_elec 128 8.7 0.087 0.104 13.328 13.369 cp_fm_diag_elpa 83 13.4 0.000 0.001 11.975 12.008 cp_fm_diag_elpa_base 83 14.4 11.721 11.806 11.969 11.997 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 11.122 11.636 init_scf_run 11 5.9 0.000 0.001 11.618 11.619 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.618 11.619 multiply_cannon_metrocomm4 27577 15.6 0.105 0.117 3.939 10.845 dbcsr_mm_accdrv_process 62242 16.2 5.543 6.525 10.094 10.841 mp_irecv_dv 69486 16.3 3.733 10.439 3.733 10.439 make_images_data 5014 15.6 0.068 0.076 8.242 9.815 hybrid_alltoall_any 5200 16.5 0.354 1.589 6.982 9.176 pw_transfer 1547 11.6 0.085 0.101 8.662 8.724 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.438 8.504 wfi_extrapolate 11 7.9 0.001 0.001 8.300 8.300 fft_wrap_pw1pw2_140 523 13.2 1.223 1.248 7.490 7.578 grid_integrate_task_list 128 12.3 7.158 7.479 7.158 7.479 density_rs2pw 128 9.7 0.006 0.007 7.025 7.463 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.381 7.376 cp_fm_cholesky_decompose 22 10.9 6.859 6.940 6.859 6.940 calculate_dm_sparse 128 9.5 0.001 0.001 6.447 6.589 mp_sum_l 7950 12.9 4.387 6.469 4.387 6.469 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.004 6.108 6.118 fft3d_ps 1291 14.7 2.836 2.998 5.986 6.036 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.375 5.448 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.180 5.392 grid_collocate_task_list 128 9.7 5.001 5.314 5.001 5.314 mp_alltoall_d11v 2415 14.1 4.183 4.894 4.183 4.894 potential_pw2rs 128 12.3 0.016 0.018 4.557 4.596 mp_allgather_i34 2507 14.6 1.461 4.400 1.461 4.400 mp_sum_d 4470 12.1 2.661 4.096 2.661 4.096 dbcsr_complete_redistribute 395 12.7 0.788 0.900 3.170 4.024 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=189.399000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=801.181818, yerr=1.695887 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5080151e8b1ac516ab82332683adfb2cbf2d479f_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420241154048 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528896499712 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514757E+12 0.0% 0.0% 100.0% flops max/rank 2.928533E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755939872 0.0% 0.0% 100.0% number of processed stacks 3984192 0.0% 0.0% 100.0% average stack size 0.0 0.0 1695.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 956.481536E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1042912 MPI messages size (bytes): total size 2.716210E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.604448E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 281856 36943429632 131072 < size <= 4194304 660064 996105256960 4194304 < size <= 16777216 65632 931531083728 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58351. MP_Allreduce 11057 1000. MP_Sync 87 MP_Alltoall 1712 9388896. MP_SendRecv 7936 75008. MP_ISendRecv 7936 75008. MP_Wait 21820 MP_ISend 11748 275205. MP_IRecv 11748 275205. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.040 0.092 175.596 175.598 qs_mol_dyn_low 1 2.0 0.003 0.003 174.852 174.867 qs_forces 11 3.9 0.032 0.171 174.764 174.766 qs_energies 11 4.9 0.020 0.154 168.144 168.175 scf_env_do_scf 11 5.9 0.001 0.002 152.295 152.301 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 117.647 117.648 velocity_verlet 10 3.0 0.002 0.002 112.137 112.139 dbcsr_multiply_generic 2507 12.6 0.177 0.183 81.898 83.066 qs_scf_new_mos 117 7.6 0.001 0.001 80.556 80.968 qs_scf_loop_do_ot 117 8.6 0.001 0.001 80.555 80.967 ot_scf_mini 117 9.6 0.003 0.004 76.321 76.774 multiply_cannon 2507 13.6 0.506 0.525 62.022 67.350 multiply_cannon_loop 2507 14.6 1.126 1.195 59.313 62.090 ot_mini 117 10.6 0.001 0.001 42.567 43.005 init_scf_loop 11 6.9 0.001 0.003 34.547 34.548 mp_waitall_1 170520 16.6 24.992 34.391 24.992 34.391 rebuild_ks_matrix 128 8.3 0.001 0.001 30.223 30.750 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.017 30.222 30.750 prepare_preconditioner 11 7.9 0.000 0.000 30.550 30.602 make_preconditioner 11 8.9 0.000 0.001 30.550 30.602 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.241 29.604 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.254 27.731 multiply_cannon_multrec 20056 15.6 12.978 16.432 22.347 25.868 multiply_cannon_metrocomm3 20056 15.6 0.063 0.067 15.571 24.930 qs_ot_get_derivative 117 11.6 0.002 0.002 22.834 23.268 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.824 20.872 apply_single 128 13.6 0.001 0.001 19.824 20.872 qs_ot_get_p 128 10.4 0.001 0.001 19.084 19.645 ot_diis_step 117 11.6 0.018 0.018 19.614 19.614 make_m2s 5014 13.6 0.080 0.085 14.798 15.565 multiply_cannon_sync_h2d 20056 15.6 13.621 15.442 13.621 15.442 make_images 5014 14.6 1.150 1.240 14.566 15.332 qs_ot_p2m_diag 83 11.4 0.266 0.290 14.819 14.849 cp_fm_cholesky_invert 11 10.9 14.261 14.270 14.261 14.270 sum_up_and_integrate 128 10.3 0.133 0.144 14.005 14.030 integrate_v_rspace 128 11.3 0.004 0.004 13.871 13.901 cp_dbcsr_syevd 83 12.4 0.005 0.006 13.763 13.765 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.540 13.572 calculate_rho_elec 128 8.7 0.130 0.145 13.539 13.571 cp_fm_diag_elpa 83 13.4 0.000 0.001 10.568 10.587 cp_fm_diag_elpa_base 83 14.4 10.166 10.319 10.564 10.584 init_scf_run 11 5.9 0.000 0.001 10.509 10.510 scf_env_initial_rho_setup 11 6.9 0.002 0.015 10.509 10.509 make_images_data 5014 15.6 0.064 0.073 8.942 10.094 multiply_cannon_metrocomm4 17549 15.6 0.067 0.077 3.516 9.564 hybrid_alltoall_any 5200 16.5 0.451 2.053 7.729 9.390 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.013 9.336 mp_irecv_dv 50230 16.2 3.386 9.305 3.386 9.305 dbcsr_mm_accdrv_process 41502 16.2 5.618 5.958 8.835 8.993 pw_transfer 1547 11.6 0.085 0.103 8.688 8.790 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 8.464 8.569 grid_integrate_task_list 128 12.3 7.297 7.727 7.297 7.727 fft_wrap_pw1pw2_140 523 13.2 1.312 1.361 7.513 7.635 wfi_extrapolate 11 7.9 0.001 0.001 7.445 7.445 density_rs2pw 128 9.7 0.006 0.007 6.885 7.289 cp_fm_cholesky_decompose 22 10.9 7.259 7.279 7.259 7.279 cp_fm_upper_to_full 105 14.8 5.630 7.105 5.630 7.105 dbcsr_complete_redistribute 395 12.7 1.169 1.202 4.558 6.314 calculate_dm_sparse 128 9.5 0.001 0.001 5.827 5.911 fft3d_ps 1291 14.7 2.755 2.968 5.828 5.893 grid_collocate_task_list 128 9.7 5.164 5.554 5.164 5.554 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.465 5.472 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.576 5.237 mp_alltoall_d11v 2415 14.1 4.314 5.163 4.314 5.163 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.381 5.135 mp_sum_l 7950 12.9 3.261 4.801 3.261 4.801 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.551 4.732 potential_pw2rs 128 12.3 0.021 0.023 4.376 4.389 mp_allgather_i34 2507 14.6 1.210 4.331 1.210 4.331 transfer_fm_to_dbcsr 11 9.9 0.020 0.026 2.289 4.032 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.989 4.027 qs_energies_init_hamiltonians 11 5.9 0.002 0.004 3.926 3.955 mp_alltoall_i22 716 14.1 1.909 3.840 1.909 3.840 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.466 3.522 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=175.598000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=899.636364, yerr=10.890135 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5080151e8b1ac516ab82332683adfb2cbf2d479f_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420241154048 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528896499712 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514757E+12 0.0% 0.0% 100.0% flops max/rank 4.353788E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755939872 0.0% 0.0% 100.0% number of processed stacks 5977344 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.142399E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1143192 MPI messages size (bytes): total size 2.023815E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.770319E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 319024 36042702848 131072 < size <= 4194304 715736 785529176064 4194304 < size <= 16777216 70320 665379345520 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58348. MP_Allreduce 11057 1083. MP_Sync 87 MP_Alltoall 1712 12503107. MP_SendRecv 5888 75008. MP_ISendRecv 5888 75008. MP_Wait 22442 MP_ISend 14952 244818. MP_IRecv 14952 244818. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.046 0.116 187.535 187.538 qs_mol_dyn_low 1 2.0 0.003 0.010 186.898 186.912 qs_forces 11 3.9 0.022 0.048 186.785 186.795 qs_energies 11 4.9 0.003 0.010 179.677 179.702 scf_env_do_scf 11 5.9 0.001 0.002 162.547 162.558 velocity_verlet 10 3.0 0.002 0.002 123.737 123.740 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 116.698 116.699 dbcsr_multiply_generic 2507 12.6 0.186 0.192 80.880 81.882 qs_scf_new_mos 117 7.6 0.001 0.001 80.566 80.906 qs_scf_loop_do_ot 117 8.6 0.001 0.001 80.566 80.905 ot_scf_mini 117 9.6 0.003 0.004 76.015 76.350 multiply_cannon 2507 13.6 0.554 0.587 56.041 59.724 multiply_cannon_loop 2507 14.6 1.828 1.909 52.491 54.770 init_scf_loop 11 6.9 0.001 0.003 45.725 45.726 ot_mini 117 10.6 0.001 0.001 43.299 43.629 prepare_preconditioner 11 7.9 0.000 0.001 41.607 41.636 make_preconditioner 11 8.9 0.000 0.002 41.607 41.636 make_full_inverse_cholesky 11 9.9 0.012 0.023 35.352 40.286 multiply_cannon_multrec 30084 15.6 13.589 19.018 26.783 31.805 rebuild_ks_matrix 128 8.3 0.001 0.001 29.267 29.644 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.019 29.267 29.643 mp_waitall_1 147882 16.7 17.354 27.210 17.354 27.210 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.410 26.749 qs_ot_get_derivative 117 11.6 0.002 0.002 23.627 23.963 make_m2s 5014 13.6 0.097 0.103 20.362 21.379 make_images 5014 14.6 1.957 2.155 20.053 21.066 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 19.027 19.700 apply_single 128 13.6 0.001 0.001 19.026 19.699 ot_diis_step 117 11.6 0.018 0.019 19.554 19.555 qs_ot_get_p 128 10.4 0.001 0.001 18.885 19.256 cp_fm_upper_to_full 105 14.8 11.385 16.690 11.385 16.690 cp_fm_cholesky_invert 11 10.9 15.807 15.815 15.807 15.815 multiply_cannon_metrocomm3 30084 15.6 0.050 0.052 6.359 15.587 qs_ot_p2m_diag 83 11.4 0.343 0.390 14.672 14.726 sum_up_and_integrate 128 10.3 0.141 0.151 14.027 14.052 integrate_v_rspace 128 11.3 0.004 0.004 13.886 13.915 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.622 13.667 calculate_rho_elec 128 8.7 0.173 0.189 13.622 13.666 dbcsr_mm_accdrv_process 62264 16.2 8.747 9.429 12.764 13.281 cp_dbcsr_syevd 83 12.4 0.005 0.006 13.277 13.278 dbcsr_complete_redistribute 395 12.7 1.522 1.639 9.061 12.814 make_images_data 5014 15.6 0.068 0.075 10.990 12.463 hybrid_alltoall_any 5200 16.5 0.522 2.184 9.731 11.702 multiply_cannon_sync_h2d 30084 15.6 10.420 11.555 10.420 11.555 copy_fm_to_dbcsr 209 11.7 0.002 0.002 7.701 11.424 init_scf_run 11 5.9 0.000 0.001 10.901 10.902 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.901 10.902 cp_fm_diag_elpa 83 13.4 0.000 0.001 10.090 10.108 cp_fm_diag_elpa_base 83 14.4 9.135 9.455 10.084 10.100 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.780 10.028 transfer_fm_to_dbcsr 11 9.9 0.001 0.005 6.233 9.882 mp_alltoall_i22 716 14.1 5.573 9.433 5.573 9.433 pw_transfer 1547 11.6 0.086 0.101 8.921 9.002 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.695 8.782 grid_integrate_task_list 128 12.3 7.502 7.936 7.502 7.936 fft_wrap_pw1pw2_140 523 13.2 1.455 1.484 7.764 7.862 wfi_extrapolate 11 7.9 0.001 0.001 7.655 7.655 cp_fm_cholesky_decompose 22 10.9 7.444 7.533 7.444 7.533 density_rs2pw 128 9.7 0.006 0.006 6.661 6.921 multiply_cannon_metrocomm4 25070 15.6 0.084 0.093 2.822 6.656 mp_irecv_dv 76098 16.2 2.666 6.383 2.666 6.383 calculate_dm_sparse 128 9.5 0.001 0.001 6.253 6.364 fft3d_ps 1291 14.7 2.845 2.937 5.858 5.931 grid_collocate_task_list 128 9.7 5.287 5.822 5.287 5.822 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.605 5.661 mp_alltoall_d11v 2415 14.1 4.885 5.405 4.885 5.405 qs_energies_init_hamiltonians 11 5.9 0.009 0.012 4.551 4.567 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.466 4.558 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.419 4.530 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.204 4.270 potential_pw2rs 128 12.3 0.023 0.023 4.211 4.224 mp_allgather_i34 2507 14.6 1.633 3.945 1.633 3.945 mp_sum_l 7950 12.9 2.568 3.871 2.568 3.871 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=187.538000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1074.181818, yerr=20.121532 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5080151e8b1ac516ab82332683adfb2cbf2d479f_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456039424 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613072052224 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176077312 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176077312 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.233020E+12 0.0% 0.0% 100.0% flops max/rank 5.910120E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383904 0.0% 0.0% 100.0% number of processed stacks 1976928 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.9 marketing flops 145.650931E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 1.513337E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 242784 MPI messages size (bytes): total size 1.341806E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.526748E+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 115008 60297314304 4194304 < size <= 16777216 105840 554906419200 16777216 < size 20352 726592466352 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 9010 51. MP_Alltoall 9724 794507. MP_ISend 40420 2100460. MP_IRecv 40420 2099564. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 57624. MP_Allreduce 11184 1163. MP_Sync 88 MP_Alltoall 1724 18848034. MP_SendRecv 3870 122880. MP_ISendRecv 3870 122880. MP_Wait 16244 MP_ISend 10760 423501. MP_IRecv 10760 423501. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.024 0.063 168.917 168.941 qs_mol_dyn_low 1 2.0 0.003 0.004 168.326 168.343 qs_forces 11 3.9 0.005 0.008 168.165 168.171 qs_energies 11 4.9 0.002 0.007 160.728 160.737 scf_env_do_scf 11 5.9 0.001 0.002 143.260 143.269 velocity_verlet 10 3.0 0.002 0.002 111.240 111.247 scf_env_do_scf_inner_loop 118 6.6 0.003 0.009 108.189 108.190 dbcsr_multiply_generic 2529 12.6 0.180 0.190 72.125 72.526 qs_scf_new_mos 118 7.6 0.001 0.001 71.288 71.418 qs_scf_loop_do_ot 118 8.6 0.001 0.001 71.287 71.418 ot_scf_mini 118 9.6 0.003 0.004 66.838 66.957 multiply_cannon 2529 13.6 0.568 0.604 53.537 56.916 multiply_cannon_loop 2529 14.6 0.813 0.862 50.635 51.464 ot_mini 118 10.6 0.001 0.001 37.475 37.580 init_scf_loop 11 6.9 0.000 0.000 34.921 34.924 prepare_preconditioner 11 7.9 0.000 0.000 31.105 31.138 make_preconditioner 11 8.9 0.000 0.000 31.105 31.138 mp_waitall_1 126876 16.7 24.646 30.924 24.646 30.924 make_full_inverse_cholesky 11 9.9 0.017 0.027 29.087 29.360 rebuild_ks_matrix 129 8.3 0.001 0.001 29.177 29.286 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.018 29.177 29.286 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.622 26.724 multiply_cannon_multrec 10116 15.6 10.469 14.922 18.142 21.355 qs_ot_get_derivative 118 11.6 0.001 0.002 20.392 20.508 multiply_cannon_metrocomm3 10116 15.6 0.026 0.027 12.555 19.891 cp_fm_cholesky_invert 11 10.9 17.704 17.709 17.704 17.709 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 17.054 17.225 apply_single 129 13.6 0.001 0.001 17.054 17.225 ot_diis_step 118 11.6 0.020 0.021 17.015 17.016 qs_ot_get_p 129 10.4 0.001 0.001 16.584 16.733 make_m2s 5058 13.6 0.067 0.072 14.748 15.531 make_images 5058 14.6 2.191 2.620 14.440 15.220 sum_up_and_integrate 129 10.3 0.183 0.193 14.218 14.264 integrate_v_rspace 129 11.3 0.004 0.004 14.035 14.085 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.010 14.051 calculate_rho_elec 129 8.7 0.258 0.269 14.010 14.050 qs_ot_p2m_diag 84 11.4 0.502 0.507 12.851 12.869 cp_dbcsr_syevd 84 12.4 0.006 0.006 11.679 11.680 multiply_cannon_sync_h2d 10116 15.6 10.955 11.370 10.955 11.370 init_scf_run 11 5.9 0.000 0.001 10.556 10.557 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.556 10.557 hybrid_alltoall_any 5245 16.5 0.852 3.812 8.345 9.772 make_images_data 5058 15.6 0.057 0.066 8.494 9.744 pw_transfer 1559 11.6 0.085 0.094 9.251 9.271 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 9.028 9.056 cp_fm_diag_elpa 84 13.4 0.000 0.000 8.820 8.833 cp_fm_diag_elpa_base 84 14.4 8.582 8.658 8.816 8.830 grid_integrate_task_list 129 12.3 7.838 8.211 7.838 8.211 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 8.040 8.116 fft_wrap_pw1pw2_140 527 13.2 1.830 1.874 8.032 8.061 dbcsr_mm_accdrv_process 20934 16.1 3.209 4.535 7.320 8.013 cp_fm_cholesky_decompose 22 10.9 7.770 7.869 7.770 7.869 multiply_cannon_metrocomm1 10116 15.6 0.030 0.031 4.410 7.371 wfi_extrapolate 11 7.9 0.001 0.001 7.340 7.340 density_rs2pw 129 9.7 0.006 0.006 6.652 6.959 calculate_dm_sparse 129 9.5 0.001 0.001 6.136 6.208 grid_collocate_task_list 129 9.7 5.610 5.883 5.610 5.883 fft3d_ps 1301 14.7 2.781 2.848 5.714 5.743 dbcsr_complete_redistribute 397 12.7 2.145 2.187 5.111 5.517 qs_energies_init_hamiltonians 11 5.9 0.034 0.043 5.342 5.346 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.209 5.228 mp_alltoall_d11v 2429 14.1 4.483 5.066 4.483 5.066 mp_allgather_i34 2529 14.6 1.074 4.531 1.074 4.531 potential_pw2rs 129 12.3 0.027 0.028 4.191 4.204 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.094 4.158 multiply_cannon_metrocomm4 7587 15.6 0.027 0.029 1.769 4.088 mp_irecv_dv 29102 15.9 1.730 4.007 1.730 4.007 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.595 3.919 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.592 3.633 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.552 3.629 copy_fm_to_dbcsr 210 11.7 0.001 0.002 3.278 3.608 copy_dbcsr_to_fm 187 11.8 0.004 0.004 3.439 3.547 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.471 3.483 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=168.941000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1424.090909, yerr=27.625596 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5080151e8b1ac516ab82332683adfb2cbf2d479f_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420241154048 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528896499712 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514757E+12 0.0% 0.0% 100.0% flops max/rank 11.696233E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755939872 0.0% 0.0% 100.0% number of processed stacks 1964048 0.0% 0.0% 100.0% average stack size 0.0 0.0 3439.8 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 3.185930E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 100280 MPI messages size (bytes): total size 1.136195E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.330227E+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 45208 35089547264 4194304 < size <= 16777216 44352 379752284160 16777216 < size 10104 721350154512 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 59279. MP_Allreduce 11057 1504. MP_Sync 87 MP_Alltoall 1712 36974159. MP_SendRecv 1792 218624. MP_ISendRecv 1792 218624. MP_Wait 9802 MP_ISend 6408 1080322. MP_IRecv 6408 1080322. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.073 0.137 289.725 289.745 qs_mol_dyn_low 1 2.0 0.003 0.003 288.926 288.983 qs_forces 11 3.9 0.005 0.005 288.820 288.824 qs_energies 11 4.9 0.002 0.002 280.049 280.055 scf_env_do_scf 11 5.9 0.002 0.015 257.496 257.501 velocity_verlet 10 3.0 0.019 0.020 208.932 208.939 scf_env_do_scf_inner_loop 117 6.6 0.004 0.009 134.649 134.651 init_scf_loop 11 6.9 0.000 0.000 122.550 122.551 prepare_preconditioner 11 7.9 0.000 0.000 117.701 117.731 make_preconditioner 11 8.9 0.000 0.000 117.701 117.731 make_full_inverse_cholesky 11 9.9 0.037 0.039 94.351 114.838 qs_scf_new_mos 117 7.6 0.001 0.001 91.017 91.166 qs_scf_loop_do_ot 117 8.6 0.001 0.001 91.016 91.165 ot_scf_mini 117 9.6 0.004 0.004 86.072 86.168 dbcsr_multiply_generic 2507 12.6 0.211 0.223 83.211 84.027 cp_fm_upper_to_full 105 14.8 51.894 74.204 51.894 74.204 multiply_cannon 2507 13.6 0.686 0.737 59.584 60.598 multiply_cannon_loop 2507 14.6 1.047 1.060 55.329 57.319 ot_mini 117 10.6 0.001 0.001 45.112 45.237 dbcsr_complete_redistribute 395 12.7 4.033 4.113 29.373 41.908 copy_fm_to_dbcsr 209 11.7 0.001 0.002 25.838 38.329 transfer_fm_to_dbcsr 11 9.9 0.030 0.031 23.308 35.580 rebuild_ks_matrix 128 8.3 0.001 0.001 34.260 34.400 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.018 34.259 34.399 mp_waitall_1 103674 16.8 28.653 33.926 28.653 33.926 mp_alltoall_i22 716 14.1 21.116 33.803 21.116 33.803 cp_fm_cholesky_invert 11 10.9 33.298 33.304 33.298 33.304 qs_ks_update_qs_env 128 7.6 0.001 0.001 31.814 31.941 qs_ot_get_p 128 10.4 0.001 0.001 25.680 25.793 qs_ot_get_derivative 117 11.6 0.002 0.002 24.874 24.967 qs_ot_p2m_diag 83 11.4 0.879 0.885 21.553 21.585 multiply_cannon_metrocomm3 10028 15.6 0.025 0.026 19.276 21.343 make_m2s 5014 13.6 0.078 0.080 18.761 20.438 ot_diis_step 117 11.6 0.022 0.022 20.191 20.192 make_images 5014 14.6 3.069 3.253 18.279 19.958 cp_dbcsr_syevd 83 12.4 0.006 0.006 19.742 19.747 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.361 19.695 apply_single 128 13.6 0.001 0.001 19.361 19.694 multiply_cannon_multrec 10028 15.6 10.337 12.108 18.100 18.166 cp_fm_diag_elpa 83 13.4 0.000 0.000 16.562 16.563 cp_fm_diag_elpa_base 83 14.4 12.181 13.791 16.559 16.559 qs_rho_update_rho_low 128 7.7 0.001 0.001 16.434 16.446 calculate_rho_elec 128 8.7 0.480 0.481 16.433 16.445 sum_up_and_integrate 128 10.3 0.321 0.323 16.262 16.350 integrate_v_rspace 128 11.3 0.005 0.005 15.941 16.027 multiply_cannon_sync_h2d 10028 15.6 14.311 14.345 14.311 14.345 hybrid_alltoall_any 5200 16.5 1.321 3.091 10.917 13.353 make_images_data 5014 15.6 0.065 0.071 10.669 13.017 init_scf_run 11 5.9 0.000 0.001 12.219 12.220 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.219 12.219 pw_transfer 1547 11.6 0.093 0.093 11.290 11.297 fft_wrap_pw1pw2 1291 12.7 0.011 0.011 11.054 11.061 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.868 9.943 fft_wrap_pw1pw2_140 523 13.2 3.060 3.096 9.878 9.890 dbcsr_mm_accdrv_process 20762 16.1 4.237 5.977 7.514 9.361 wfi_extrapolate 11 7.9 0.001 0.001 9.060 9.060 cp_fm_cholesky_decompose 22 10.9 8.955 8.987 8.955 8.987 grid_integrate_task_list 128 12.3 8.558 8.758 8.558 8.758 qs_energies_init_hamiltonians 11 5.9 0.009 0.044 8.151 8.153 density_rs2pw 128 9.7 0.006 0.006 7.469 7.517 mp_alltoall_d11v 2415 14.1 6.887 7.013 6.887 7.013 calculate_dm_sparse 128 9.5 0.001 0.001 6.824 6.888 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.473 6.516 grid_collocate_task_list 128 9.7 6.414 6.494 6.414 6.494 fft3d_ps 1291 14.7 2.851 2.860 6.286 6.326 copy_dbcsr_to_fm 186 11.8 0.004 0.004 6.229 6.281 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=289.745000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2861.363636, yerr=161.761481 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5080151e8b1ac516ab82332683adfb2cbf2d479f_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 1.259839E+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 18986711. 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 0.050 0.064 85.138 85.140 qs_energies 1 2.0 0.000 0.001 84.469 84.489 ls_scf 1 3.0 0.000 0.001 83.545 83.564 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.539 72.704 multiply_cannon 111 7.7 0.026 0.035 55.808 56.892 multiply_cannon_loop 111 8.7 0.227 0.245 52.439 53.675 ls_scf_main 1 4.0 0.002 0.039 52.253 52.253 density_matrix_trs4 2 5.0 0.002 0.014 46.681 46.763 ls_scf_init_scf 1 4.0 0.001 0.005 28.221 28.222 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.098 27.143 mp_waitall_1 11031 10.9 22.342 25.272 22.342 25.272 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 24.958 24.977 multiply_cannon_multrec 2664 9.7 8.118 8.888 15.613 17.467 multiply_cannon_sync_h2d 2664 9.7 13.523 15.627 13.523 15.627 make_m2s 222 7.7 0.009 0.011 13.029 13.448 make_images 222 8.7 0.099 0.109 13.006 13.426 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.675 12.603 make_images_data 222 9.7 0.004 0.005 7.599 8.147 dbcsr_mm_accdrv_process 4760 10.4 0.591 0.706 7.115 8.145 hybrid_alltoall_any 227 10.6 0.216 1.837 6.567 7.441 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.403 7.441 dbcsr_mm_accdrv_process_sort 4760 11.4 6.321 7.255 6.321 7.255 calculate_norms 4752 9.8 5.508 6.180 5.508 6.180 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.105 5.247 mp_sum_l 887 5.1 3.182 4.522 3.182 4.522 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.057 3.900 mp_irecv_dv 6231 10.9 2.041 3.868 2.041 3.868 make_images_sizes 222 9.7 0.000 0.000 0.680 3.612 mp_alltoall_i44 222 10.7 0.679 3.611 0.679 3.611 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.344 3.383 arnoldi_extremal 4 6.8 0.000 0.000 3.256 3.286 arnoldi_normal_ev 4 7.8 0.001 0.003 3.256 3.286 build_subspace 16 8.4 0.009 0.012 3.164 3.166 ls_scf_post 1 4.0 0.002 0.016 3.071 3.090 ls_scf_store_result 1 5.0 0.000 0.000 2.867 2.911 dbcsr_special_finalize 555 9.7 0.005 0.006 2.395 2.726 dbcsr_merge_single_wm 555 10.7 0.456 0.591 2.386 2.717 make_images_pack 222 9.7 2.207 2.620 2.209 2.621 dbcsr_matrix_vector_mult 304 9.0 0.006 0.014 2.325 2.562 dbcsr_sort_data 658 11.4 2.187 2.468 2.187 2.468 dbcsr_matrix_vector_mult_local 304 10.0 2.071 2.457 2.073 2.458 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.303 2.385 buffer_matrices_ensure_size 222 8.7 1.744 2.026 1.744 2.026 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.771 1.772 rebuild_ks_matrix 3 7.3 0.000 0.000 1.761 1.762 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.761 1.762 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.140000, 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/5080151e8b1ac516ab82332683adfb2cbf2d479f_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 2.172625E+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 0.088 0.166 90.636 90.641 qs_energies 1 2.0 0.000 0.001 89.681 89.685 ls_scf 1 3.0 0.000 0.001 88.306 88.310 dbcsr_multiply_generic 111 6.7 0.018 0.025 74.108 74.420 multiply_cannon 111 7.7 0.028 0.045 52.474 57.326 ls_scf_main 1 4.0 0.003 0.036 54.290 54.294 multiply_cannon_loop 111 8.7 0.136 0.146 49.679 53.083 density_matrix_trs4 2 5.0 0.003 0.004 48.574 48.789 ls_scf_init_scf 1 4.0 0.001 0.005 30.281 30.283 mp_waitall_1 9105 10.9 20.236 29.142 20.236 29.142 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.903 28.989 multiply_cannon_multrec 1332 9.7 13.381 17.050 22.786 27.588 matrix_sqrt_Newton_Schulz 2 6.5 0.002 0.006 26.477 26.523 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.071 20.235 make_m2s 222 7.7 0.007 0.009 15.072 15.613 make_images 222 8.7 1.370 1.695 15.041 15.583 dbcsr_mm_accdrv_process 4041 10.4 0.344 0.512 8.997 10.579 dbcsr_mm_accdrv_process_sort 4041 11.4 8.515 10.048 8.515 10.048 make_images_data 222 9.7 0.004 0.006 8.690 9.554 hybrid_alltoall_any 227 10.6 0.544 2.565 7.996 9.488 mp_sum_l 887 5.1 5.094 8.137 5.094 8.137 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.207 7.668 mp_irecv_dv 3311 11.0 3.186 7.609 3.186 7.609 calculate_norms 2376 9.8 6.070 6.721 6.070 6.721 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.847 6.598 multiply_cannon_sync_h2d 1332 9.7 4.918 6.322 4.918 6.322 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.964 5.172 arnoldi_extremal 4 6.8 0.000 0.003 4.674 4.739 arnoldi_normal_ev 4 7.8 0.001 0.006 4.673 4.735 build_subspace 16 8.4 0.015 0.021 4.361 4.365 ls_scf_post 1 4.0 0.000 0.001 3.734 3.739 ls_scf_store_result 1 5.0 0.000 0.000 3.404 3.506 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.130 3.356 dbcsr_matrix_vector_mult_local 304 10.0 2.735 3.218 2.737 3.220 mp_allgather_i34 111 8.7 0.783 3.007 0.783 3.007 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.601 2.693 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.125 2.485 dbcsr_data_new 4174 10.1 2.113 2.399 2.113 2.399 make_images_pack 222 9.7 1.821 2.140 1.823 2.142 dbcsr_sort_data 436 11.2 1.839 2.092 1.839 2.092 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.922 1.925 rebuild_ks_matrix 3 7.3 0.000 0.000 1.910 1.912 qs_ks_build_kohn_sham_matrix 3 8.3 0.018 0.058 1.910 1.912 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=90.641000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1779.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5080151e8b1ac516ab82332683adfb2cbf2d479f_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 2.887340E+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 0.058 0.096 92.381 92.383 qs_energies 1 2.0 0.000 0.001 91.659 91.664 ls_scf 1 3.0 0.000 0.001 90.119 90.127 dbcsr_multiply_generic 111 6.7 0.017 0.021 74.684 75.055 multiply_cannon 111 7.7 0.035 0.145 51.683 56.342 ls_scf_main 1 4.0 0.000 0.002 56.163 56.167 multiply_cannon_loop 111 8.7 0.117 0.131 48.973 52.585 density_matrix_trs4 2 5.0 0.002 0.004 50.214 50.350 mp_waitall_1 7281 11.0 23.258 32.853 23.258 32.853 ls_scf_init_scf 1 4.0 0.000 0.001 30.312 30.315 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.124 29.199 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.778 26.797 multiply_cannon_multrec 888 9.7 12.632 15.065 21.301 24.350 multiply_cannon_metrocomm3 888 9.7 0.004 0.005 10.977 22.319 make_m2s 222 7.7 0.006 0.008 16.466 17.136 make_images 222 8.7 1.584 1.862 16.428 17.099 make_images_data 222 9.7 0.004 0.005 9.631 10.571 hybrid_alltoall_any 227 10.6 0.642 2.959 9.037 10.255 dbcsr_mm_accdrv_process 3754 10.4 0.292 0.495 8.207 9.460 dbcsr_mm_accdrv_process_sort 3754 11.4 7.772 8.965 7.772 8.965 mp_sum_l 887 5.1 4.851 7.995 4.851 7.995 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.548 7.352 multiply_cannon_sync_h2d 888 9.7 6.018 7.320 6.018 7.320 mp_irecv_dv 2335 11.1 2.532 7.285 2.532 7.285 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.292 6.621 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.682 6.281 arnoldi_extremal 4 6.8 0.000 0.000 5.132 5.153 arnoldi_normal_ev 4 7.8 0.001 0.005 5.132 5.153 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.795 4.991 calculate_norms 1584 9.8 4.376 4.870 4.376 4.870 build_subspace 16 8.4 0.014 0.020 4.828 4.834 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.431 3.770 mp_allgather_i34 111 8.7 0.829 3.742 0.829 3.742 ls_scf_post 1 4.0 0.000 0.001 3.644 3.651 dbcsr_matrix_vector_mult_local 304 10.0 3.013 3.586 3.015 3.588 ls_scf_store_result 1 5.0 0.000 0.000 3.388 3.469 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.840 2.940 dbcsr_data_new 4116 9.9 2.107 2.440 2.107 2.440 make_images_sizes 222 9.7 0.000 0.000 1.005 2.271 mp_alltoall_i44 222 10.7 1.004 2.271 1.004 2.271 dbcsr_sort_data 325 11.1 1.866 2.103 1.866 2.103 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.889 1.891 make_images_pack 222 9.7 1.626 1.876 1.629 1.878 rebuild_ks_matrix 3 7.3 0.000 0.000 1.871 1.873 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.871 1.873 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=92.383000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2247.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5080151e8b1ac516ab82332683adfb2cbf2d479f_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 3.322368E+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 0.068 0.158 100.610 100.610 qs_energies 1 2.0 0.011 0.045 99.498 99.532 ls_scf 1 3.0 0.000 0.001 97.796 97.801 dbcsr_multiply_generic 111 6.7 0.018 0.023 81.385 81.670 multiply_cannon 111 7.7 0.052 0.183 54.647 59.198 ls_scf_main 1 4.0 0.001 0.014 58.854 58.855 density_matrix_trs4 2 5.0 0.003 0.015 52.682 52.833 multiply_cannon_loop 111 8.7 0.153 0.168 49.579 52.090 ls_scf_init_scf 1 4.0 0.000 0.002 35.680 35.682 ls_scf_init_matrix_S 1 5.0 0.000 0.000 34.479 34.541 mp_waitall_1 6369 11.0 23.116 32.089 23.116 32.089 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 32.004 32.021 multiply_cannon_multrec 1332 9.7 14.157 16.859 22.168 25.005 make_m2s 222 7.7 0.007 0.008 21.126 22.521 make_images 222 8.7 3.133 3.595 21.076 22.473 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.653 18.506 make_images_data 222 9.7 0.004 0.005 11.799 13.493 hybrid_alltoall_any 227 10.6 0.796 3.746 11.135 13.115 dbcsr_mm_accdrv_process 3641 10.4 0.278 0.472 7.651 9.172 dbcsr_mm_accdrv_process_sort 3641 11.4 7.212 8.678 7.212 8.678 mp_sum_l 887 5.1 4.035 7.513 4.035 7.513 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.074 6.164 mp_irecv_dv 3229 10.9 2.049 6.092 2.049 6.092 multiply_cannon_sync_h2d 1332 9.7 5.423 6.033 5.423 6.033 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.063 5.866 arnoldi_extremal 4 6.8 0.000 0.000 5.335 5.350 arnoldi_normal_ev 4 7.8 0.001 0.005 5.335 5.350 build_subspace 16 8.4 0.014 0.021 4.992 5.001 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.486 4.836 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.569 4.767 mp_allgather_i34 111 8.7 2.183 4.558 2.183 4.558 calculate_norms 2376 9.8 4.191 4.450 4.191 4.450 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.617 3.942 compute_matrix_preconditioner 1 6.0 0.001 0.002 3.724 3.734 dbcsr_matrix_vector_mult_local 304 10.0 3.181 3.682 3.183 3.684 dbcsr_sort_data 658 11.4 3.099 3.356 3.099 3.356 ls_scf_post 1 4.0 0.000 0.001 3.262 3.268 acc_transpose_blocks 1332 9.7 0.008 0.009 2.741 3.255 acc_transpose_blocks_kernels 1332 10.7 0.018 0.019 2.630 3.145 dbcsr_special_finalize 555 9.7 0.006 0.007 2.853 3.139 dbcsr_merge_single_wm 555 10.7 0.541 0.645 2.845 3.130 jit_kernel_transpose 1 13.0 2.613 3.127 2.613 3.127 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.032 3.110 ls_scf_store_result 1 5.0 0.000 0.000 2.991 3.065 dbcsr_data_release 10477 10.7 1.572 2.437 1.572 2.437 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=100.610000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2724.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5080151e8b1ac516ab82332683adfb2cbf2d479f_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 4.761866E+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 0.085 0.090 92.194 92.195 qs_energies 1 2.0 0.000 0.000 91.150 91.155 ls_scf 1 3.0 0.000 0.000 89.133 89.137 dbcsr_multiply_generic 111 6.7 0.018 0.019 70.653 70.808 ls_scf_main 1 4.0 0.000 0.000 56.249 56.249 multiply_cannon 111 7.7 0.074 0.131 52.565 55.773 multiply_cannon_loop 111 8.7 0.088 0.094 50.018 50.980 density_matrix_trs4 2 5.0 0.002 0.003 49.259 49.340 ls_scf_init_scf 1 4.0 0.000 0.000 29.458 29.460 mp_waitall_1 5436 11.0 24.211 28.543 24.211 28.543 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.216 28.244 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.123 26.136 multiply_cannon_multrec 444 9.7 13.727 16.453 20.900 22.144 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.475 15.406 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.420 14.823 make_m2s 222 7.7 0.005 0.005 13.376 14.301 make_images 222 8.7 2.041 2.487 13.308 14.231 make_images_data 222 9.7 0.003 0.004 8.134 9.652 hybrid_alltoall_any 227 10.6 0.802 3.809 8.073 9.458 dbcsr_mm_accdrv_process 3003 10.4 0.351 0.406 6.871 8.000 multiply_cannon_sync_h2d 444 9.7 6.757 7.829 6.757 7.829 dbcsr_mm_accdrv_process_sort 3003 11.4 6.509 7.597 6.509 7.597 arnoldi_extremal 4 6.8 0.000 0.000 5.846 5.862 arnoldi_normal_ev 4 7.8 0.001 0.004 5.846 5.861 build_subspace 16 8.4 0.015 0.020 5.454 5.467 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.452 4.648 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.187 4.377 mp_sum_l 887 5.1 2.710 4.258 2.710 4.258 dbcsr_matrix_vector_mult_local 304 10.0 3.687 4.172 3.689 4.174 calculate_norms 792 9.8 3.632 3.732 3.632 3.732 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.541 3.621 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.499 3.616 mp_irecv_dv 1241 11.2 1.483 3.568 1.483 3.568 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.919 3.532 mp_allgather_i34 111 8.7 1.123 3.481 1.123 3.481 ls_scf_post 1 4.0 0.000 0.000 3.426 3.431 make_images_sizes 222 9.7 0.000 0.000 0.856 3.285 mp_alltoall_i44 222 10.7 0.855 3.285 0.855 3.285 ls_scf_store_result 1 5.0 0.000 0.000 3.221 3.252 dbcsr_data_new 4608 9.7 1.790 2.291 1.790 2.291 dbcsr_finalize 304 7.8 0.062 0.076 2.198 2.286 dbcsr_merge_all 275 8.9 0.481 0.530 2.056 2.122 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.034 2.034 rebuild_ks_matrix 3 7.3 0.000 0.000 2.002 2.002 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.002 2.002 qs_energies_init_hamiltonians 1 3.0 0.001 0.002 2.001 2.001 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=92.195000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3757.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5080151e8b1ac516ab82332683adfb2cbf2d479f_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 8.881099E+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 0.075 0.093 108.088 108.089 qs_energies 1 2.0 0.000 0.000 106.539 106.547 ls_scf 1 3.0 0.000 0.000 103.612 103.620 dbcsr_multiply_generic 111 6.7 0.024 0.028 77.448 77.599 ls_scf_main 1 4.0 0.000 0.000 63.624 63.624 density_matrix_trs4 2 5.0 0.002 0.003 54.678 54.734 multiply_cannon 111 7.7 0.175 0.285 51.268 53.721 multiply_cannon_loop 111 8.7 0.096 0.098 48.161 48.703 ls_scf_init_scf 1 4.0 0.000 0.000 36.205 36.206 ls_scf_init_matrix_S 1 5.0 0.000 0.000 34.702 34.726 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 32.017 32.029 mp_waitall_1 4527 11.1 22.094 25.579 22.094 25.579 make_m2s 222 7.7 0.005 0.005 22.586 23.802 make_images 222 8.7 3.573 3.865 22.478 23.694 multiply_cannon_multrec 444 9.7 17.840 18.427 22.543 23.072 hybrid_alltoall_any 227 10.6 1.657 3.640 12.762 15.642 make_images_data 222 9.7 0.003 0.004 12.997 15.215 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.516 13.098 multiply_cannon_sync_h2d 444 9.7 8.798 8.858 8.798 8.858 arnoldi_extremal 4 6.8 0.000 0.000 7.342 7.362 arnoldi_normal_ev 4 7.8 0.003 0.009 7.342 7.362 build_subspace 16 8.4 0.026 0.036 6.794 6.806 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.449 5.598 dbcsr_matrix_vector_mult_local 304 10.0 5.027 5.340 5.030 5.343 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.086 5.178 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.840 5.096 dbcsr_mm_accdrv_process 1814 10.4 0.290 0.361 4.522 4.658 dbcsr_mm_accdrv_process_sort 1814 11.4 4.186 4.320 4.186 4.320 mp_allgather_i34 111 8.7 1.225 3.945 1.225 3.945 ls_scf_post 1 4.0 0.000 0.000 3.783 3.791 make_images_sizes 222 9.7 0.000 0.000 1.461 3.585 mp_alltoall_i44 222 10.7 1.461 3.585 1.461 3.585 ls_scf_store_result 1 5.0 0.000 0.000 3.505 3.541 compute_matrix_preconditioner 1 6.0 0.002 0.002 3.421 3.426 calculate_norms 792 9.8 3.226 3.272 3.226 3.272 dbcsr_finalize 304 7.8 0.082 0.089 3.070 3.164 acc_transpose_blocks 444 9.7 0.003 0.003 2.604 2.949 dbcsr_merge_all 275 8.9 0.883 0.914 2.856 2.941 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.896 2.896 acc_transpose_blocks_kernels 444 10.7 0.006 0.006 2.524 2.868 dbcsr_complete_redistribute 5 7.6 1.442 1.485 2.747 2.865 jit_kernel_transpose 1 13.0 2.518 2.862 2.518 2.862 matrix_ls_to_qs 2 6.0 0.000 0.000 2.393 2.520 dbcsr_sort_data 325 11.1 2.438 2.497 2.438 2.497 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.402 2.404 rebuild_ks_matrix 3 7.3 0.000 0.000 2.336 2.338 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.336 2.338 dbcsr_new_transposed 4 7.5 0.242 0.253 2.273 2.286 dbcsr_data_new 6591 9.6 1.827 2.278 1.827 2.278 dbcsr_frobenius_norm 74 6.6 2.059 2.137 2.198 2.232 dbcsr_add_d 103 6.2 0.000 0.000 2.123 2.199 dbcsr_add_anytype 103 7.2 0.859 0.890 2.123 2.199 dbcsr_data_release 12724 10.6 1.977 2.188 1.977 2.188 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=108.089000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6987.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5080151e8b1ac516ab82332683adfb2cbf2d479f_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 588.701696E+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 4804984. 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 0.041 0.209 231.230 231.232 qs_mol_dyn_low 1 2.0 0.004 0.045 229.569 229.590 qs_forces 5 3.8 0.005 0.008 229.378 229.380 qs_energies 5 4.8 0.012 0.097 226.315 226.365 scf_env_do_scf 5 5.8 0.002 0.040 210.949 210.952 scf_env_do_scf_inner_loop 105 6.6 0.002 0.006 183.902 183.904 qs_scf_new_mos 105 7.6 0.000 0.001 143.152 143.348 qs_scf_loop_do_ot 105 8.6 0.001 0.001 143.152 143.348 dbcsr_multiply_generic 1445 12.2 0.123 0.134 134.210 134.678 ot_scf_mini 105 9.6 0.003 0.005 133.230 133.399 multiply_cannon 1445 13.2 0.276 0.288 114.726 116.848 multiply_cannon_loop 1445 14.2 2.851 2.993 112.967 114.139 velocity_verlet 4 3.0 0.002 0.020 106.670 106.671 ot_mini 105 10.6 0.001 0.002 60.580 60.719 multiply_cannon_multrec 69360 15.2 29.594 34.370 39.774 45.031 qs_ot_get_p 112 10.4 0.001 0.002 42.442 42.725 mp_waitall_1 488190 16.1 35.266 41.800 35.266 41.800 qs_ot_get_derivative 55 11.6 0.001 0.001 38.770 38.930 multiply_cannon_sync_h2d 69360 15.2 29.070 32.909 29.070 32.909 multiply_cannon_metrocomm3 69360 15.2 0.200 0.215 25.939 32.317 qs_ot_p2m_diag 40 11.0 0.020 0.032 31.364 31.461 rebuild_ks_matrix 110 8.4 0.000 0.000 29.241 29.415 qs_ks_build_kohn_sham_matrix 110 9.4 0.012 0.016 29.240 29.415 cp_dbcsr_syevd 40 12.0 0.002 0.002 28.232 28.233 init_scf_loop 7 6.6 0.000 0.004 27.016 27.018 qs_ks_update_qs_env 112 7.6 0.001 0.001 26.794 26.951 apply_preconditioner_dbcsr 62 12.6 0.000 0.001 23.109 23.389 apply_single 62 13.6 0.000 0.000 23.109 23.389 cp_fm_syevd 40 13.0 0.000 0.000 22.999 23.150 prepare_preconditioner 7 7.6 0.000 0.000 22.100 22.135 make_preconditioner 7 8.6 0.000 0.002 22.100 22.135 ot_new_cg_direction 55 11.6 0.001 0.002 21.089 21.090 cp_fm_redistribute_end 40 14.0 9.122 18.199 9.127 18.200 cp_fm_syevd_base 40 14.0 9.065 18.144 9.065 18.144 qs_rho_update_rho_low 110 7.6 0.001 0.002 17.352 17.705 calculate_rho_elec 110 8.6 0.030 0.032 17.352 17.705 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.999 15.209 make_full_inverse_cholesky 7 9.6 0.000 0.001 14.784 14.852 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 13.916 14.026 mp_sum_l 4764 12.2 12.571 13.293 12.571 13.293 init_scf_run 5 5.8 0.000 0.001 12.490 12.491 scf_env_initial_rho_setup 5 6.8 0.002 0.005 12.490 12.491 density_rs2pw 110 9.6 0.005 0.007 11.636 12.236 pw_transfer 1645 12.4 0.080 0.102 11.932 12.196 calculate_dm_sparse 110 9.5 0.000 0.001 11.871 12.079 fft_wrap_pw1pw2 1425 13.5 0.012 0.015 11.793 12.060 dbcsr_mm_accdrv_process 154766 15.8 6.223 6.387 10.048 10.879 fft_wrap_pw1pw2_240 915 15.0 1.142 1.237 10.362 10.585 qs_vxc_create 110 10.4 0.002 0.005 10.515 10.565 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 10.439 10.509 cp_fm_cholesky_invert 7 10.6 10.204 10.211 10.204 10.211 check_diag 80 13.5 8.595 8.888 9.612 9.763 transfer_rs2pw 445 10.6 0.007 0.008 8.300 8.927 fft3d_pb 915 16.0 2.393 2.679 8.569 8.876 sum_up_and_integrate 60 10.3 0.028 0.033 8.696 8.711 integrate_v_rspace 60 11.3 0.002 0.002 8.668 8.684 acc_transpose_blocks 69360 15.2 0.368 0.387 7.560 8.042 xc_rho_set_and_dset_create 110 12.4 0.077 0.098 7.394 7.658 calculate_first_density_matrix 1 7.0 0.001 0.011 7.620 7.638 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.001 7.576 7.592 make_m2s 2890 13.2 0.079 0.087 6.502 7.120 xc_vxc_pw_create 60 11.3 0.039 0.050 7.036 7.086 make_full_single_inverse 7 9.6 0.001 0.004 7.026 7.062 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 6.990 7.039 make_images 2890 14.2 0.239 0.259 6.396 7.014 multiply_cannon_metrocomm1 69360 15.2 0.097 0.103 4.658 7.006 xc_pw_derive 510 13.4 0.006 0.007 6.263 6.349 mp_alltoall_z22v 2340 17.7 5.610 5.872 5.610 5.872 mp_waitany 7680 13.5 4.576 5.464 4.576 5.464 acc_transpose_blocks_kernels 69360 16.2 0.849 0.888 4.769 5.114 potential_pw2rs 60 12.3 0.003 0.003 4.902 4.937 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=231.232000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=561.000000, yerr=2.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 5080151e8b1ac516ab82332683adfb2cbf2d479f Summary: empty Status: OK