=== 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: 908860eada1ff98329e2139c3f6fcecbf4b94d08 ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, # Cray-FFTW 3.3.8.10, COSMA 2.6.6, ELPA 2023.05.001, # HDF5 1.14.2, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.2.2, LIBVORI 220621, LIBXSMM 1.17, # PLUMED 2.9.0, SPGLIB 2.3.1, LIBGRPP 20231215, # SPFFT 1.0.6, SPLA 1.5.5 # # 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: 28.03.2024 # # \ 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; \ rm -rf build; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-hdf5 --with-libvdwxc --with-pexsi --with-plumed --with-sirius=no --with-spfft --with-spla; \ 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=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 #USE_DEEPMD := 2.2.7 USE_ELPA := 2023.05.001 USE_HDF5 := 1.14.2 USE_LIBGRPP := 20231215 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.9.0 #USE_QUIP := 0.9.10 #USE_SIRIUS := 7.5.2 USE_SPFFT := 1.0.6 USE_SPGLIB := 2.3.1 USE_SPLA := 1.5.5 # Only needed for SIRIUS #LIBVDWXC_VER := 0.4.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(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_DEEPMD),) USE_DEEPMD := $(strip $(USE_DEEPMD)) DEEPMD_INC := $(INSTALL_PATH)/libdeepmd_c-$(USE_DEEPMD)/include DEEPMD_LIB := $(INSTALL_PATH)/libdeepmd_c-$(USE_DEEPMD)/lib CFLAGS += -I$(DEEPMD_INC) DFLAGS += -D__DEEPMD LIBS += $(DEEPMD_LIB)/libdeepmd.so LIBS += $(DEEPMD_LIB)/libdeepmd_c.so LIBS += $(DEEPMD_LIB)/libdeepmd_cc.so LIBS += $(DEEPMD_LIB)/libdeepmd_dyn_cudart.so LIBS += $(DEEPMD_LIB)/libdeepmd_op.so LIBS += $(DEEPMD_LIB)/libdeepmd_op_cuda.so LIBS += $(DEEPMD_LIB)/libtensorflow_cc.so.2 LIBS += $(DEEPMD_LIB)/libtensorflow_framework.so.2 endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBGRPP),) USE_LIBGRPP := $(strip $(USE_LIBGRPP)) LIBGRPP_INC := $(INSTALL_PATH)/libgrpp-main-$(USE_LIBGRPP)/include LIBGRPP_LIB := $(INSTALL_PATH)/libgrpp-main-$(USE_LIBGRPP)/lib CFLAGS += -I$(LIBGRPP_INC) DFLAGS += -D__LIBGRPP LIBS += $(LIBGRPP_LIB)/liblibgrpp.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib ifeq ($(USE_ACC), yes) SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__LIBVDWXC DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a endif ifneq ($(USE_SPFFT),) USE_SPFFT := $(strip $(USE_SPFFT)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(USE_SPFFT)/include ifeq ($(USE_ACC), yes) SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(USE_SPFFT)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(USE_SPFFT)/lib endif CFLAGS += -I$(SPFFT_INC) DFLAGS += -D__SPFFT LIBS += $(SPFFT_LIB)/libspfft.a endif ifneq ($(USE_SPLA),) USE_SPLA := $(strip $(USE_SPLA)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(USE_SPLA)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPLA_LIB := $(INSTALL_PATH)/SpLA-$(USE_SPLA)/lib/cuda else SPLA_LIB := $(INSTALL_PATH)/SpLA-$(USE_SPLA)/lib endif CFLAGS += -I$(SPLA_INC) DFLAGS += -D__SPLA LIBS += $(SPLA_LIB)/libspla.a endif ifneq ($(USE_HDF5),) USE_HDF5 := $(strip $(USE_HDF5)) HDF5_INC := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/include HDF5_LIB := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/lib CFLAGS += -I$(HDF5_INC) DFLAGS += -D__HDF5 LIBS += $(HDF5_LIB)/libhdf5_fortran.a LIBS += $(HDF5_LIB)/libhdf5_f90cstub.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/908860eada1ff98329e2139c3f6fcecbf4b94d08_performance_tests/01 job id: 53222758 --- 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/908860eada1ff98329e2139c3f6fcecbf4b94d08_performance_tests/02 job id: 53222759 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/908860eada1ff98329e2139c3f6fcecbf4b94d08_performance_tests/03 job id: 53222760 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/908860eada1ff98329e2139c3f6fcecbf4b94d08_performance_tests/04 job id: 53222762 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/908860eada1ff98329e2139c3f6fcecbf4b94d08_performance_tests/05 job id: 53222764 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/908860eada1ff98329e2139c3f6fcecbf4b94d08_performance_tests/06 job id: 53222765 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/908860eada1ff98329e2139c3f6fcecbf4b94d08_performance_tests/07 job id: 53222766 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/908860eada1ff98329e2139c3f6fcecbf4b94d08_performance_tests/08 job id: 53222767 --- 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/908860eada1ff98329e2139c3f6fcecbf4b94d08_performance_tests/09 job id: 53222768 --- 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/908860eada1ff98329e2139c3f6fcecbf4b94d08_performance_tests/10 job id: 53222769 --- 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/908860eada1ff98329e2139c3f6fcecbf4b94d08_performance_tests/11 job id: 53222770 --- 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/908860eada1ff98329e2139c3f6fcecbf4b94d08_performance_tests/12 job id: 53222772 --- 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/908860eada1ff98329e2139c3f6fcecbf4b94d08_performance_tests/13 job id: 53222773 --- 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/908860eada1ff98329e2139c3f6fcecbf4b94d08_performance_tests/14 job id: 53222774 --- 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/908860eada1ff98329e2139c3f6fcecbf4b94d08_performance_tests/15 job id: 53222776 --- 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/908860eada1ff98329e2139c3f6fcecbf4b94d08_performance_tests/16 job id: 53222777 --- 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/908860eada1ff98329e2139c3f6fcecbf4b94d08_performance_tests/17 job id: 53222778 --- 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/908860eada1ff98329e2139c3f6fcecbf4b94d08_performance_tests/18 job id: 53222779 --- 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/908860eada1ff98329e2139c3f6fcecbf4b94d08_performance_tests/19 job id: 53222780 --- 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/908860eada1ff98329e2139c3f6fcecbf4b94d08_performance_tests/20 job id: 53222781 --- 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] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/908860eada1ff98329e2139c3f6fcecbf4b94d08_performance_tests/21 job id: 53222782 --- 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] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/908860eada1ff98329e2139c3f6fcecbf4b94d08_performance_tests/22 job id: 53222783 --- 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] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/908860eada1ff98329e2139c3f6fcecbf4b94d08_performance_tests/23 job id: 53222784 --- 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] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/908860eada1ff98329e2139c3f6fcecbf4b94d08_performance_tests/24 job id: 53222785 --- 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] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/908860eada1ff98329e2139c3f6fcecbf4b94d08_performance_tests/25 job id: 53222786 --- 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] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/908860eada1ff98329e2139c3f6fcecbf4b94d08_performance_tests/26 job id: 53222787 --- 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] = 20 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/908860eada1ff98329e2139c3f6fcecbf4b94d08_performance_tests/27 job id: 53222788 --- 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/908860eada1ff98329e2139c3f6fcecbf4b94d08_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 172669. MP_Allreduce 424 8. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.043 0.050 130.345 130.346 farming_run 1 2.0 128.568 128.569 130.293 130.307 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 4194304 0.0% 0.0% 100.0% flops 14 x 32 x 32 154140672 0.0% 0.0% 100.0% flops 29 x 32 x 32 159645696 0.0% 0.0% 100.0% flops 14 x 14 x 32 208732160 0.0% 0.0% 100.0% flops 29 x 14 x 32 212860928 0.0% 0.0% 100.0% flops 14 x 29 x 32 212860928 0.0% 0.0% 100.0% flops 29 x 29 x 32 227352576 0.0% 0.0% 100.0% flops 14 x 32 x 14 896801644032 0.0% 0.0% 100.0% flops 29 x 32 x 14 928925089792 0.0% 0.0% 100.0% flops 14 x 32 x 29 928925089792 0.0% 0.0% 100.0% flops 29 x 32 x 29 962100985856 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693169221632 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753639550976 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.164741E+12 0.0% 0.0% 100.0% flops max/rank 447.801317E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249492158 0.0% 0.0% 100.0% number of processed stacks 164328 0.0% 0.0% 100.0% average stack size 0.0 0.0 1518.3 marketing flops 7.165779E+12 ------------------------------------------------------------------------------- # multiplications 1160 max memory usage/rank 1.529401E+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 822089. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 230 1131389. MP_Allreduce 571 1938539. MP_Sync 25 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 10 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.006 0.023 109.922 109.922 qs_energies 1 2.0 0.009 0.011 109.738 109.741 mp2_main 1 3.0 0.000 0.000 107.925 107.936 mp2_gpw_main 1 4.0 0.044 0.077 107.022 107.024 mp2_ri_gpw_compute_in 1 5.0 0.175 0.187 88.317 88.419 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 57.158 57.247 mp2_eri_3c_integrate_gpw 272 7.0 0.149 0.157 43.606 48.419 get_2c_integrals 1 6.0 0.018 0.044 30.772 30.978 integrate_v_rspace 273 8.0 0.442 0.458 25.489 29.718 fft_wrap_pw1pw2 5465 10.4 0.066 0.069 28.034 28.674 fft_wrap_pw1pw2_100 2178 11.4 1.320 1.451 25.487 26.155 grid_integrate_task_list 273 9.0 20.733 25.871 20.733 25.871 compute_2c_integrals 1 7.0 0.002 0.002 21.468 21.488 compute_2c_integrals_loop_lm 1 8.0 0.002 0.003 20.760 21.161 mp2_eri_2c_integrate_gpw 1 9.0 2.390 2.425 20.758 21.160 rpa_ri_compute_en 1 5.0 0.019 0.026 18.568 18.662 fft3d_s 5443 12.4 16.832 17.192 16.853 17.213 ao_to_mo_and_store_B_mult_1 272 7.0 10.766 15.313 10.766 15.313 calculate_wavefunction 272 8.0 5.299 5.378 12.791 13.716 calc_potential_gpw 544 9.5 0.004 0.004 10.916 11.656 rpa_num_int 1 6.0 0.000 0.002 10.597 10.606 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.474 10.543 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 9.766 10.537 potential_pw2rs 545 10.0 0.106 0.108 8.386 10.324 cp_fm_cholesky_decompose 12 8.2 9.448 9.656 9.448 9.656 calc_mat_Q 8 8.0 0.000 0.000 9.461 9.604 contract_S_to_Q 8 9.0 0.000 0.000 8.885 9.028 cholesky_decomp 1 7.0 0.000 0.000 8.484 8.691 collocate_single_gaussian 272 10.0 0.040 0.042 7.788 8.643 parallel_gemm_fm 14 9.1 0.000 0.000 8.514 8.613 parallel_gemm_fm_cosma 14 10.1 8.514 8.613 8.514 8.613 create_integ_mat 1 6.0 0.015 0.028 7.782 7.792 array2fm 1 7.0 0.000 0.000 6.740 7.184 pw_poisson_solve 545 10.5 0.010 0.011 4.639 4.847 pw_gather_s 2722 12.2 3.902 4.843 3.902 4.843 pw_scatter_s 2720 12.7 4.504 4.746 4.504 4.746 pw_poisson_set 548 11.5 0.018 0.019 3.335 3.547 array2fm_buffer_send 1 8.0 2.974 3.144 2.974 3.144 pw_copy 4911 11.6 2.582 2.709 2.582 2.709 ao_to_mo_and_store_B_E_Ex_1 272 7.0 1.082 1.228 2.023 2.208 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=107.021716, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2787.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/908860eada1ff98329e2139c3f6fcecbf4b94d08_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 200775. MP_Allreduce 424 9. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.034 0.048 431.945 431.946 farming_run 1 2.0 430.489 430.503 431.897 431.909 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827128576 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788821 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.272537E+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 713945. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 757 480173. MP_Allreduce 2021 21391. MP_Sync 37 MP_Alltoall 77 2393255. MP_SendRecv 2876 2171443. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.033 223.609 223.609 qs_energies 1 2.0 0.000 0.000 223.383 223.390 scf_env_do_scf 1 3.0 0.000 0.000 115.391 115.391 qs_ks_update_qs_env 5 5.0 0.000 0.000 114.483 114.490 rebuild_ks_matrix 4 6.0 0.000 0.000 114.481 114.489 qs_ks_build_kohn_sham_matrix 4 7.0 0.055 0.063 114.481 114.489 hfx_ks_matrix 4 8.0 0.001 0.001 114.128 114.131 integrate_four_center 4 9.0 0.153 0.469 114.127 114.130 mp2_main 1 3.0 0.000 0.000 107.707 107.714 mp2_gpw_main 1 4.0 0.048 0.073 106.742 106.754 integrate_four_center_main 4 10.0 0.126 0.532 102.047 105.556 integrate_four_center_bin 265 11.0 101.921 105.024 101.921 105.024 init_scf_loop 1 4.0 0.000 0.000 97.149 97.149 mp2_ri_gpw_compute_in 1 5.0 0.066 0.123 80.027 81.141 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 56.801 57.911 mp2_eri_3c_integrate_gpw 91 7.0 0.146 0.160 44.617 49.589 integrate_v_rspace 95 8.0 0.398 0.565 28.288 33.108 fft_wrap_pw1pw2 1868 10.4 0.029 0.037 30.520 31.186 fft_wrap_pw1pw2_100 730 11.4 0.649 0.776 28.211 28.888 grid_integrate_task_list 95 9.0 23.575 28.604 23.575 28.604 mp2_ri_gpw_compute_en 1 5.0 0.055 0.063 26.564 28.534 ao_to_mo_and_store_B_mult_1 91 7.0 10.513 28.482 10.513 28.482 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.831 1.888 24.796 24.807 get_2c_integrals 1 6.0 0.001 0.005 23.110 23.129 compute_2c_integrals 1 7.0 0.002 0.003 22.279 22.285 compute_2c_integrals_loop_lm 1 8.0 0.001 0.002 21.854 22.144 mp2_eri_2c_integrate_gpw 1 9.0 1.726 1.830 21.853 22.143 fft3d_s 1823 12.4 19.207 19.799 19.220 19.812 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 18.240 18.240 calc_potential_gpw 182 9.5 0.002 0.002 12.434 13.336 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 10.831 11.488 calculate_wavefunction 91 8.0 1.983 2.033 10.004 10.757 potential_pw2rs 186 10.0 0.033 0.035 9.111 10.602 mp2_ri_gpw_compute_en_expansio 172 7.0 0.555 0.578 8.783 9.346 collocate_single_gaussian 91 10.0 0.018 0.025 8.309 8.791 local_gemm 172 8.0 8.228 8.772 8.228 8.772 mp2_ri_gpw_compute_en_comm 22 7.0 0.503 0.532 7.800 8.345 mp_sync 37 10.5 4.069 6.822 4.069 6.822 integrate_four_center_load 4 10.0 0.000 0.000 6.768 6.772 hfx_load_balance 1 11.0 0.000 0.000 6.768 6.772 mp2_ri_gpw_compute_en_ener 172 7.0 6.337 6.399 6.337 6.399 mp_sendrecv_dm3 2068 8.0 5.827 6.371 5.827 6.371 pw_poisson_solve 186 10.4 0.005 0.006 5.639 6.262 pw_gather_s 912 12.2 4.803 5.534 4.803 5.534 pw_poisson_set 189 11.4 0.009 0.011 4.166 4.855 pw_scatter_s 910 12.7 4.090 4.822 4.090 4.822 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=106.740082, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1526.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/908860eada1ff98329e2139c3f6fcecbf4b94d08_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 445.284352E+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 3690 72628. 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 0.093 0.551 57.295 57.297 qs_mol_dyn_low 1 2.0 0.015 0.099 56.039 56.047 qs_forces 11 3.9 0.015 0.035 55.794 55.806 qs_energies 11 4.9 0.006 0.019 54.274 54.330 scf_env_do_scf 11 5.9 0.000 0.001 47.408 47.408 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 45.391 45.392 dbcsr_multiply_generic 2286 12.5 0.112 0.158 35.672 36.104 qs_scf_new_mos 108 7.5 0.000 0.001 35.487 35.804 qs_scf_loop_do_ot 108 8.5 0.000 0.001 35.487 35.803 ot_scf_mini 108 9.5 0.002 0.002 33.780 33.980 multiply_cannon 2286 13.5 0.180 0.199 27.720 29.182 multiply_cannon_loop 2286 14.5 1.837 1.936 27.054 28.536 velocity_verlet 10 3.0 0.018 0.126 27.234 27.240 ot_mini 108 10.5 0.001 0.001 20.923 21.193 qs_ot_get_derivative 108 11.5 0.001 0.001 17.903 18.081 mp_waitall_1 245248 16.5 9.658 16.161 9.658 16.161 multiply_cannon_metrocomm3 54864 15.5 0.073 0.079 6.677 14.274 multiply_cannon_multrec 54864 15.5 3.607 5.719 7.957 11.902 qs_ot_get_p 119 10.4 0.013 0.035 8.211 8.464 rebuild_ks_matrix 119 8.3 0.000 0.000 7.885 8.054 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.010 7.885 8.054 mp_sum_l 7287 12.8 5.818 7.658 5.818 7.658 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.951 7.110 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 6.450 6.869 dbcsr_mm_accdrv_process 76910 16.1 1.852 2.895 4.264 6.301 multiply_cannon_sync_h2d 54864 15.5 5.112 6.096 5.112 6.096 init_scf_run 11 5.9 0.001 0.042 5.481 5.481 scf_env_initial_rho_setup 11 6.9 0.036 0.057 5.480 5.481 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.299 5.419 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.201 5.239 cp_dbcsr_syevd 50 12.0 0.002 0.003 4.450 4.450 sum_up_and_integrate 119 10.3 0.001 0.002 4.365 4.370 integrate_v_rspace 119 11.3 0.002 0.003 4.354 4.361 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.245 4.246 cp_fm_redistribute_end 50 14.0 2.165 4.216 2.171 4.220 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.019 4.155 calculate_rho_elec 119 8.7 0.011 0.016 4.019 4.154 cp_fm_diag_elpa_base 50 14.0 2.042 4.107 2.046 4.116 multiply_cannon_metrocomm1 54864 15.5 0.057 0.063 2.297 3.727 jit_kernel_multiply 13 15.8 2.348 3.627 2.348 3.627 calculate_first_density_matrix 1 7.0 0.000 0.000 3.512 3.551 calculate_dm_sparse 119 9.5 0.000 0.000 2.952 3.132 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.918 3.128 apply_single 119 13.6 0.000 0.000 2.918 3.128 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.950 2.966 acc_transpose_blocks 54864 15.5 0.221 0.241 2.242 2.810 ot_diis_step 108 11.5 0.006 0.006 2.740 2.741 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.529 2.577 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.494 2.541 mp_sum_d 4139 12.0 1.549 2.217 1.549 2.217 grid_integrate_task_list 119 12.3 2.009 2.134 2.009 2.134 density_rs2pw 119 9.7 0.003 0.004 2.038 2.106 init_scf_loop 11 6.9 0.000 0.000 2.000 2.000 make_m2s 4572 13.5 0.054 0.057 1.811 1.876 wfi_extrapolate 11 7.9 0.001 0.001 1.864 1.865 make_images 4572 14.5 0.132 0.138 1.728 1.793 potential_pw2rs 119 12.3 0.004 0.004 1.746 1.757 fft_wrap_pw1pw2 1201 11.6 0.011 0.012 1.481 1.549 transfer_rs2pw 487 10.6 0.006 0.007 1.380 1.509 mp_waitany 12084 13.8 1.278 1.475 1.278 1.475 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.423 1.456 mp_alltoall_d11v 2130 13.8 1.324 1.456 1.324 1.456 acc_transpose_blocks_sync 164592 16.5 1.200 1.437 1.200 1.437 grid_collocate_task_list 119 9.7 1.346 1.426 1.346 1.426 fft3d_ps 1201 13.6 0.365 0.465 1.257 1.321 transfer_pw2rs 487 13.2 0.005 0.006 1.313 1.320 fft_wrap_pw1pw2_140 487 12.2 0.048 0.054 1.146 1.213 dbcsr_dot_sd 1205 11.9 0.047 0.058 0.798 1.200 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=57.297000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=424.727273, yerr=0.862439 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/908860eada1ff98329e2139c3f6fcecbf4b94d08_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 487.034880E+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 3679 72937. MP_Allreduce 10306 303. MP_Sync 54 MP_Alltoall 2060 991560. 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 0.215 0.288 41.864 42.036 qs_mol_dyn_low 1 2.0 0.026 0.069 40.484 40.507 qs_forces 11 3.9 0.002 0.003 40.359 40.382 qs_energies 11 4.9 0.006 0.037 38.593 38.618 scf_env_do_scf 11 5.9 0.001 0.002 32.885 32.887 scf_env_do_scf_inner_loop 108 6.5 0.007 0.044 30.420 30.425 dbcsr_multiply_generic 2286 12.5 0.102 0.105 23.427 23.785 qs_scf_new_mos 108 7.5 0.001 0.001 21.797 22.032 qs_scf_loop_do_ot 108 8.5 0.001 0.001 21.797 22.031 ot_scf_mini 108 9.5 0.002 0.003 20.882 21.050 multiply_cannon 2286 13.5 0.207 0.215 18.306 19.870 multiply_cannon_loop 2286 14.5 1.191 1.256 17.152 18.679 velocity_verlet 10 3.0 0.004 0.017 18.093 18.100 ot_mini 108 10.5 0.001 0.001 13.254 13.493 mp_waitall_1 200699 16.5 6.304 11.932 6.304 11.932 qs_ot_get_derivative 108 11.5 0.001 0.001 10.911 11.082 multiply_cannon_metrocomm3 27432 15.5 0.072 0.075 4.818 10.610 multiply_cannon_multrec 27432 15.5 1.905 4.717 6.683 9.975 dbcsr_mm_accdrv_process 47894 16.0 3.572 5.981 4.699 7.529 rebuild_ks_matrix 119 8.3 0.000 0.000 6.924 7.067 qs_ks_build_kohn_sham_matrix 119 9.3 0.015 0.040 6.923 7.067 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.028 6.158 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 4.246 5.084 qs_ot_get_p 119 10.4 0.002 0.011 4.824 5.046 mp_sum_l 7287 12.8 2.348 4.677 2.348 4.677 init_scf_run 11 5.9 0.000 0.001 4.308 4.308 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.308 4.308 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.973 4.107 apply_single 119 13.6 0.000 0.000 2.972 4.106 sum_up_and_integrate 119 10.3 0.003 0.019 3.747 3.760 integrate_v_rspace 119 11.3 0.002 0.003 3.731 3.747 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.602 3.640 calculate_rho_elec 119 8.7 0.021 0.024 3.602 3.640 qs_ot_p2m_diag 50 11.0 0.008 0.012 3.016 3.034 make_m2s 4572 13.5 0.052 0.054 2.617 2.961 make_images 4572 14.5 0.205 0.245 2.526 2.873 calculate_first_density_matrix 1 7.0 0.000 0.003 2.790 2.792 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.623 2.623 jit_kernel_multiply 10 16.3 1.069 2.617 1.069 2.617 acc_transpose_blocks 27432 15.5 0.114 0.121 1.733 2.547 init_scf_loop 11 6.9 0.005 0.038 2.445 2.462 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.326 2.413 ot_diis_step 108 11.5 0.011 0.011 2.296 2.296 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.291 2.291 cp_fm_redistribute_end 50 14.0 1.161 2.264 1.165 2.267 cp_fm_diag_elpa_base 50 14.0 1.069 2.168 1.097 2.214 calculate_dm_sparse 119 9.5 0.000 0.001 2.127 2.207 multiply_cannon_sync_h2d 27432 15.5 1.714 2.175 1.714 2.175 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.098 2.100 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.908 1.945 grid_integrate_task_list 119 12.3 1.837 1.935 1.837 1.935 density_rs2pw 119 9.7 0.004 0.004 1.844 1.925 fft_wrap_pw1pw2 1201 11.6 0.012 0.014 1.723 1.761 make_images_data 4572 15.5 0.047 0.053 1.226 1.535 potential_pw2rs 119 12.3 0.006 0.006 1.481 1.497 fft3d_ps 1201 13.6 0.512 0.567 1.443 1.477 acc_transpose_blocks_kernels 27432 16.5 0.188 0.276 0.776 1.459 hybrid_alltoall_any 4725 16.4 0.053 0.115 1.090 1.448 mp_alltoall_d11v 2130 13.8 1.268 1.448 1.268 1.448 wfi_extrapolate 11 7.9 0.002 0.010 1.437 1.438 fft_wrap_pw1pw2_140 487 12.2 0.047 0.050 1.376 1.416 prepare_preconditioner 11 7.9 0.000 0.000 1.347 1.374 make_preconditioner 11 8.9 0.002 0.018 1.347 1.374 grid_collocate_task_list 119 9.7 1.296 1.356 1.296 1.356 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.318 1.326 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.246 1.294 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.210 1.257 transfer_rs2pw 487 10.6 0.005 0.006 1.150 1.242 jit_kernel_transpose 5 15.5 0.588 1.186 0.588 1.186 mp_sum_d 4139 12.0 0.680 1.175 0.680 1.175 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.099 1.122 mp_allgather_i34 2286 14.5 0.579 0.983 0.579 0.983 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.953 0.966 acc_transpose_blocks_sync 82296 16.5 0.816 0.938 0.816 0.938 transfer_pw2rs 487 13.2 0.004 0.005 0.923 0.927 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=42.036000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=462.727273, yerr=1.911254 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/908860eada1ff98329e2139c3f6fcecbf4b94d08_performance_tests/05/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 59.051995E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3143552 0.0% 0.0% 100.0% average stack size 0.0 0.0 46.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 520.843264E+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 3629 73909. 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 0.044 0.062 34.575 34.575 qs_mol_dyn_low 1 2.0 0.003 0.004 32.910 32.927 qs_forces 11 3.9 0.002 0.003 32.806 32.807 qs_energies 11 4.9 0.027 0.036 31.274 31.277 scf_env_do_scf 11 5.9 0.000 0.001 25.815 25.815 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 23.380 23.381 dbcsr_multiply_generic 2286 12.5 0.096 0.098 17.630 17.713 qs_scf_new_mos 108 7.5 0.001 0.001 15.808 15.822 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.808 15.821 ot_scf_mini 108 9.5 0.002 0.002 15.045 15.056 velocity_verlet 10 3.0 0.004 0.007 15.034 15.037 multiply_cannon 2286 13.5 0.191 0.194 14.187 14.863 multiply_cannon_loop 2286 14.5 0.865 0.904 13.426 14.224 ot_mini 108 10.5 0.001 0.001 9.322 9.337 qs_ot_get_derivative 108 11.5 0.001 0.001 7.811 7.822 multiply_cannon_multrec 18288 15.5 1.873 2.824 7.262 7.610 dbcsr_mm_accdrv_process 38222 16.0 5.215 6.116 5.298 6.183 rebuild_ks_matrix 119 8.3 0.000 0.000 6.106 6.121 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.105 6.121 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.401 5.415 init_scf_run 11 5.9 0.002 0.054 4.177 4.177 scf_env_initial_rho_setup 11 6.9 0.038 0.053 4.175 4.177 mp_waitall_1 158411 16.6 2.800 4.092 2.800 4.092 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.004 3.617 qs_ot_get_p 119 10.4 0.001 0.002 3.455 3.480 sum_up_and_integrate 119 10.3 0.001 0.001 3.470 3.476 integrate_v_rspace 119 11.3 0.003 0.003 3.458 3.466 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.190 3.194 calculate_rho_elec 119 8.7 0.031 0.031 3.190 3.194 calculate_first_density_matrix 1 7.0 0.000 0.000 2.930 2.970 multiply_cannon_metrocomm3 18288 15.5 0.048 0.049 1.601 2.778 init_scf_loop 11 6.9 0.000 0.000 2.417 2.418 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.031 2.343 apply_single 119 13.6 0.000 0.000 2.030 2.343 qs_ot_p2m_diag 50 11.0 0.012 0.012 2.290 2.296 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.213 2.215 make_m2s 4572 13.5 0.045 0.046 1.944 2.082 calculate_dm_sparse 119 9.5 0.000 0.001 2.027 2.038 make_images 4572 14.5 0.191 0.203 1.857 1.995 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.993 1.994 grid_integrate_task_list 119 12.3 1.811 1.913 1.811 1.913 density_rs2pw 119 9.7 0.003 0.004 1.771 1.895 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.821 1.830 acc_transpose_blocks 18288 15.5 0.079 0.081 1.726 1.802 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.747 1.756 cp_fm_diag_elpa_base 50 14.0 1.723 1.736 1.745 1.755 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.734 1.742 mp_sum_l 7287 12.8 1.171 1.639 1.171 1.639 fft_wrap_pw1pw2 1201 11.6 0.012 0.013 1.624 1.634 ot_diis_step 108 11.5 0.011 0.012 1.493 1.493 prepare_preconditioner 11 7.9 0.000 0.000 1.481 1.485 make_preconditioner 11 8.9 0.000 0.000 1.481 1.485 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.361 1.418 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.362 1.366 fft3d_ps 1201 13.6 0.535 0.553 1.309 1.319 grid_collocate_task_list 119 9.7 1.231 1.314 1.231 1.314 fft_wrap_pw1pw2_140 487 12.2 0.061 0.063 1.294 1.306 potential_pw2rs 119 12.3 0.007 0.008 1.275 1.281 qs_init_subsys 1 2.0 0.161 0.564 1.261 1.262 qs_env_setup 1 3.0 0.000 0.000 1.054 1.230 qs_env_rebuild_pw_env 23 5.3 0.020 0.037 1.030 1.181 pw_env_rebuild 1 5.0 0.155 0.219 1.005 1.181 wfi_extrapolate 11 7.9 0.001 0.001 1.155 1.155 multiply_cannon_sync_h2d 18288 15.5 1.021 1.143 1.021 1.143 transfer_rs2pw 487 10.6 0.005 0.005 1.008 1.135 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.962 0.980 qs_energies_init_hamiltonians 11 5.9 0.016 0.037 0.979 0.980 make_images_data 4572 15.5 0.047 0.051 0.809 0.964 acc_transpose_blocks_kernels 18288 16.5 0.218 0.223 0.889 0.940 hybrid_alltoall_any 4725 16.4 0.058 0.119 0.705 0.877 transfer_pw2rs 487 13.2 0.004 0.004 0.812 0.814 acc_transpose_blocks_sync 54864 16.5 0.738 0.782 0.738 0.782 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.774 0.775 mp_alltoall_d11v 2130 13.8 0.665 0.755 0.665 0.755 dbcsr_complete_redistribute 329 12.2 0.284 0.330 0.676 0.746 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.671 0.732 jit_kernel_transpose 5 15.6 0.671 0.717 0.671 0.717 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=34.575000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=494.090909, yerr=3.553604 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/908860eada1ff98329e2139c3f6fcecbf4b94d08_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 555.368448E+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 3679 72932. MP_Allreduce 10304 342. MP_Sync 54 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_comm_split 50 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 0.081 0.130 38.941 38.942 qs_mol_dyn_low 1 2.0 0.004 0.019 37.422 37.447 qs_forces 11 3.9 0.012 0.036 36.958 36.960 qs_energies 11 4.9 0.002 0.005 35.237 35.250 scf_env_do_scf 11 5.9 0.000 0.001 29.358 29.358 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 26.401 26.401 dbcsr_multiply_generic 2286 12.5 0.102 0.104 20.618 20.746 qs_scf_new_mos 108 7.5 0.001 0.001 18.380 18.438 qs_scf_loop_do_ot 108 8.5 0.001 0.001 18.379 18.438 velocity_verlet 10 3.0 0.357 0.409 18.355 18.357 multiply_cannon 2286 13.5 0.215 0.223 16.824 17.734 ot_scf_mini 108 9.5 0.002 0.003 17.341 17.388 multiply_cannon_loop 2286 14.5 1.503 1.622 15.866 16.576 ot_mini 108 10.5 0.001 0.001 10.654 10.719 multiply_cannon_multrec 27432 15.5 2.595 3.331 9.576 9.876 qs_ot_get_derivative 108 11.5 0.001 0.001 8.778 8.827 dbcsr_mm_accdrv_process 47916 15.9 6.419 8.272 6.882 8.414 rebuild_ks_matrix 119 8.3 0.000 0.000 6.437 6.488 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.436 6.487 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.712 5.759 init_scf_run 11 5.9 0.002 0.038 4.452 4.452 scf_env_initial_rho_setup 11 6.9 0.031 0.084 4.450 4.452 qs_ot_get_p 119 10.4 0.001 0.002 3.825 3.900 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.271 3.662 sum_up_and_integrate 119 10.3 0.001 0.001 3.466 3.476 integrate_v_rspace 119 11.3 0.003 0.003 3.455 3.465 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.315 3.346 calculate_rho_elec 119 8.7 0.040 0.046 3.314 3.346 calculate_first_density_matrix 1 7.0 0.000 0.000 3.007 3.040 init_scf_loop 11 6.9 0.000 0.000 2.935 2.936 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.113 2.649 apply_single 119 13.6 0.000 0.000 2.113 2.649 acc_transpose_blocks 27432 15.5 0.117 0.120 2.445 2.577 mp_waitall_1 137007 16.6 1.969 2.528 1.969 2.528 make_m2s 4572 13.5 0.054 0.056 2.368 2.476 calculate_dm_sparse 119 9.5 0.000 0.000 2.351 2.411 qs_ot_p2m_diag 50 11.0 0.015 0.022 2.375 2.386 make_images 4572 14.5 0.271 0.330 2.261 2.367 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.353 2.354 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.323 2.347 mp_sum_l 7287 12.8 1.203 2.084 1.203 2.084 prepare_preconditioner 11 7.9 0.000 0.000 1.981 1.990 make_preconditioner 11 8.9 0.000 0.000 1.981 1.990 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.970 1.970 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.924 1.937 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.719 1.920 grid_integrate_task_list 119 12.3 1.828 1.905 1.828 1.905 density_rs2pw 119 9.7 0.003 0.004 1.778 1.873 ot_diis_step 108 11.5 0.012 0.012 1.834 1.835 jit_kernel_multiply 8 16.1 0.396 1.769 0.396 1.769 fft_wrap_pw1pw2 1201 11.6 0.012 0.014 1.728 1.759 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.696 1.697 cp_fm_redistribute_end 50 14.0 0.849 1.667 0.856 1.670 cp_fm_diag_elpa_base 50 14.0 0.786 1.590 0.811 1.635 multiply_cannon_metrocomm3 27432 15.5 0.040 0.041 0.878 1.578 acc_transpose_blocks_sync 82296 16.5 1.394 1.504 1.394 1.504 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.444 1.458 fft_wrap_pw1pw2_140 487 12.2 0.071 0.072 1.416 1.453 fft3d_ps 1201 13.6 0.577 0.639 1.390 1.417 wfi_extrapolate 11 7.9 0.001 0.001 1.344 1.344 grid_collocate_task_list 119 9.7 1.248 1.328 1.248 1.328 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.306 1.327 potential_pw2rs 119 12.3 0.008 0.009 1.274 1.280 qs_energies_init_hamiltonians 11 5.9 0.021 0.042 1.133 1.159 qs_init_subsys 1 2.0 0.079 0.124 1.104 1.118 dbcsr_complete_redistribute 329 12.2 0.296 0.344 0.942 1.108 transfer_rs2pw 487 10.6 0.005 0.005 0.990 1.095 qs_env_setup 1 3.0 0.000 0.000 0.985 1.082 mp_alltoall_d11v 2130 13.8 0.781 1.033 0.781 1.033 make_images_data 4572 15.5 0.047 0.051 0.887 1.013 qs_env_rebuild_pw_env 23 5.3 0.001 0.001 0.909 0.985 hybrid_alltoall_any 4725 16.4 0.066 0.156 0.764 0.935 acc_transpose_blocks_kernels 27432 16.5 0.270 0.276 0.905 0.933 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.802 0.878 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.704 0.866 pw_env_rebuild 1 5.0 0.017 0.045 0.759 0.856 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.811 0.816 transfer_pw2rs 487 13.2 0.004 0.004 0.780 0.783 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=38.942000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=527.000000, yerr=3.618136 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/908860eada1ff98329e2139c3f6fcecbf4b94d08_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 621.785088E+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 62672. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3629 73907. 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 0.061 0.142 30.957 30.958 qs_mol_dyn_low 1 2.0 0.003 0.004 29.287 29.359 qs_forces 11 3.9 0.008 0.014 29.192 29.193 qs_energies 11 4.9 0.011 0.023 27.485 27.493 scf_env_do_scf 11 5.9 0.000 0.001 21.886 21.887 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 19.407 19.407 velocity_verlet 10 3.0 0.002 0.002 14.437 14.441 dbcsr_multiply_generic 2286 12.5 0.097 0.101 13.701 13.771 qs_scf_new_mos 108 7.5 0.001 0.001 11.947 11.972 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.946 11.971 multiply_cannon 2286 13.5 0.225 0.235 11.085 11.520 ot_scf_mini 108 9.5 0.002 0.002 11.244 11.270 multiply_cannon_loop 2286 14.5 0.648 0.669 10.161 10.348 multiply_cannon_multrec 9144 15.5 1.735 1.961 6.483 6.741 ot_mini 108 10.5 0.001 0.001 6.539 6.570 rebuild_ks_matrix 119 8.3 0.000 0.000 5.889 5.913 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.888 5.913 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.262 5.285 qs_ot_get_derivative 108 11.5 0.001 0.001 5.230 5.258 dbcsr_mm_accdrv_process 12550 15.8 3.904 4.682 4.640 4.750 init_scf_run 11 5.9 0.005 0.075 4.137 4.137 scf_env_initial_rho_setup 11 6.9 0.069 0.075 4.132 4.137 sum_up_and_integrate 119 10.3 0.001 0.001 3.342 3.346 integrate_v_rspace 119 11.3 0.003 0.003 3.332 3.336 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.204 3.210 calculate_rho_elec 119 8.7 0.060 0.061 3.203 3.209 calculate_first_density_matrix 1 7.0 0.000 0.000 2.912 2.916 qs_ot_get_p 119 10.4 0.001 0.002 2.801 2.840 init_scf_loop 11 6.9 0.000 0.000 2.457 2.458 mp_waitall_1 115863 16.7 1.418 2.078 1.418 2.078 calculate_dm_sparse 119 9.5 0.000 0.000 2.039 2.058 make_m2s 4572 13.5 0.034 0.035 1.826 1.981 grid_integrate_task_list 119 12.3 1.852 1.934 1.852 1.934 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.883 1.903 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.893 1.894 make_images 4572 14.5 0.268 0.298 1.736 1.890 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.796 1.798 fft_wrap_pw1pw2 1201 11.6 0.012 0.013 1.747 1.756 density_rs2pw 119 9.7 0.003 0.004 1.614 1.689 acc_transpose_blocks 9144 15.5 0.042 0.044 1.534 1.579 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.555 1.556 prepare_preconditioner 11 7.9 0.000 0.000 1.547 1.552 make_preconditioner 11 8.9 0.000 0.000 1.547 1.552 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.520 1.528 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.442 1.467 fft_wrap_pw1pw2_140 487 12.2 0.094 0.096 1.444 1.452 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.435 1.441 fft3d_ps 1201 13.6 0.638 0.653 1.360 1.369 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.353 1.365 grid_collocate_task_list 119 9.7 1.295 1.362 1.295 1.362 ot_diis_step 108 11.5 0.013 0.013 1.295 1.295 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.276 1.285 cp_fm_diag_elpa_base 50 14.0 1.249 1.265 1.275 1.283 jit_kernel_multiply 6 15.7 0.696 1.268 0.696 1.268 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.238 1.266 apply_single 119 13.6 0.000 0.000 1.238 1.266 qs_energies_init_hamiltonians 11 5.9 0.012 0.019 1.232 1.239 qs_init_subsys 1 2.0 0.267 0.576 1.199 1.199 potential_pw2rs 119 12.3 0.010 0.010 1.157 1.161 wfi_extrapolate 11 7.9 0.001 0.001 1.102 1.102 qs_env_setup 1 3.0 0.000 0.000 0.881 1.029 qs_env_rebuild_pw_env 23 5.3 0.025 0.089 0.807 1.022 hybrid_alltoall_any 4725 16.4 0.066 0.179 0.767 0.990 make_images_data 4572 15.5 0.042 0.046 0.798 0.964 mp_alltoall_d11v 2130 13.8 0.780 0.951 0.780 0.951 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.879 0.932 pw_env_rebuild 1 5.0 0.075 0.150 0.760 0.908 transfer_rs2pw 487 10.6 0.004 0.005 0.789 0.849 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.824 0.829 acc_transpose_blocks_kernels 9144 16.5 0.118 0.121 0.753 0.778 dbcsr_complete_redistribute 329 12.2 0.330 0.353 0.732 0.756 cp_fm_cholesky_invert 11 10.9 0.744 0.749 0.744 0.749 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.747 0.749 acc_transpose_blocks_sync 27432 16.5 0.723 0.748 0.723 0.748 qs_env_update_s_mstruct 11 6.9 0.002 0.003 0.671 0.721 mp_allgather_i34 2286 14.5 0.239 0.673 0.239 0.673 transfer_pw2rs 487 13.2 0.003 0.004 0.659 0.661 jit_kernel_transpose 5 15.6 0.635 0.659 0.635 0.659 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 0.351 0.659 multiply_cannon_metrocomm1 9144 15.5 0.023 0.024 0.288 0.649 mp_alltoall_z22v 1201 15.6 0.596 0.629 0.596 0.629 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=30.958000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=589.454545, yerr=5.382862 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/908860eada1ff98329e2139c3f6fcecbf4b94d08_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 773.095424E+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 3629 74142. 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 0.053 0.061 43.531 43.532 qs_mol_dyn_low 1 2.0 0.003 0.006 42.961 42.968 qs_forces 11 3.9 0.007 0.009 42.564 42.565 qs_energies 11 4.9 0.001 0.001 40.561 40.565 scf_env_do_scf 11 5.9 0.000 0.001 32.824 32.824 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 26.824 26.825 velocity_verlet 10 3.0 0.128 0.131 22.568 22.600 dbcsr_multiply_generic 2286 12.5 0.109 0.110 20.471 20.741 qs_scf_new_mos 108 7.5 0.001 0.001 17.600 17.702 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.600 17.701 multiply_cannon 2286 13.5 0.293 0.302 16.350 17.180 ot_scf_mini 108 9.5 0.002 0.002 16.438 16.541 multiply_cannon_loop 2286 14.5 0.869 0.905 15.091 15.843 ot_mini 108 10.5 0.001 0.001 10.105 10.225 multiply_cannon_multrec 9144 15.5 3.511 4.893 9.466 9.672 qs_ot_get_derivative 108 11.5 0.001 0.001 8.035 8.137 rebuild_ks_matrix 119 8.3 0.000 0.000 7.006 7.151 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.006 7.150 dbcsr_mm_accdrv_process 12550 15.8 4.618 6.089 5.824 7.012 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.322 6.453 init_scf_loop 11 6.9 0.000 0.000 5.972 5.973 init_scf_run 11 5.9 0.000 0.001 5.036 5.036 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.036 5.036 prepare_preconditioner 11 7.9 0.000 0.000 4.978 4.992 make_preconditioner 11 8.9 0.000 0.000 4.978 4.992 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.997 4.879 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.927 3.932 calculate_rho_elec 119 8.7 0.118 0.122 3.927 3.931 sum_up_and_integrate 119 10.3 0.001 0.001 3.684 3.691 integrate_v_rspace 119 11.3 0.004 0.004 3.674 3.681 calculate_first_density_matrix 1 7.0 0.000 0.000 3.576 3.578 qs_ot_get_p 119 10.4 0.023 0.023 3.407 3.550 mp_waitall_1 94719 16.7 2.605 3.492 2.605 3.492 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.872 3.276 cp_fm_upper_to_full 72 14.2 2.101 2.990 2.101 2.990 calculate_dm_sparse 119 9.5 0.000 0.000 2.704 2.720 make_m2s 4572 13.5 0.038 0.038 2.465 2.620 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.578 2.579 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.231 2.579 apply_single 119 13.6 0.000 0.000 2.231 2.579 make_images 4572 14.5 0.350 0.381 2.343 2.497 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.425 2.427 fft_wrap_pw1pw2 1201 11.6 0.014 0.014 2.379 2.384 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.276 2.340 multiply_cannon_metrocomm3 9144 15.5 0.020 0.021 1.470 2.264 acc_transpose_blocks 9144 15.5 0.044 0.045 2.214 2.261 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.074 2.126 grid_integrate_task_list 119 12.3 2.073 2.109 2.073 2.109 dbcsr_complete_redistribute 329 12.2 0.271 0.277 1.559 2.104 ot_diis_step 108 11.5 0.014 0.014 2.042 2.042 fft_wrap_pw1pw2_140 487 12.2 0.179 0.180 2.022 2.030 density_rs2pw 119 9.7 0.003 0.003 1.957 1.968 qs_ot_p2m_diag 50 11.0 0.042 0.043 1.962 1.964 mp_sum_l 7287 12.8 1.103 1.887 1.103 1.887 fft3d_ps 1201 13.6 0.861 0.879 1.831 1.835 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.233 1.788 qs_env_update_s_mstruct 11 6.9 0.020 0.052 1.718 1.748 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.701 1.708 mp_alltoall_i22 627 13.8 0.980 1.618 0.980 1.618 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.608 1.608 jit_kernel_multiply 6 15.5 1.177 1.561 1.177 1.561 grid_collocate_task_list 119 9.7 1.516 1.533 1.516 1.533 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 0.975 1.529 wfi_extrapolate 11 7.9 0.001 0.001 1.401 1.401 hybrid_alltoall_any 4725 16.4 0.090 0.151 1.151 1.391 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.321 1.340 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.308 1.308 cp_fm_diag_elpa_base 50 14.0 1.220 1.234 1.306 1.306 make_images_data 4572 15.5 0.046 0.049 1.094 1.298 mp_alltoall_d11v 2130 13.8 1.204 1.242 1.204 1.242 potential_pw2rs 119 12.3 0.013 0.014 1.198 1.200 acc_transpose_blocks_sync 27432 16.5 1.113 1.173 1.113 1.173 cp_fm_cholesky_invert 11 10.9 1.129 1.136 1.129 1.136 acc_transpose_blocks_kernels 9144 16.5 0.119 0.121 1.041 1.055 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.001 1.046 transfer_rs2pw 487 10.6 0.005 0.005 0.960 0.986 qs_create_task_list 11 7.9 0.001 0.001 0.927 0.937 generate_qs_task_list 11 8.9 0.369 0.387 0.927 0.937 jit_kernel_transpose 5 15.6 0.921 0.936 0.921 0.936 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.894 0.907 copy_dbcsr_to_fm 153 11.3 0.002 0.002 0.733 0.873 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=43.532000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=727.181818, yerr=12.626222 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/908860eada1ff98329e2139c3f6fcecbf4b94d08_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 497.680384E+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 3528 76079. MP_Allreduce 9920 483. MP_Sync 52 MP_Alltoall 1938 1071797. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_comm_split 48 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 0.060 0.178 86.962 86.992 qs_mol_dyn_low 1 2.0 0.025 0.176 85.992 86.002 qs_forces 11 3.9 0.018 0.021 85.854 85.874 qs_energies 11 4.9 0.001 0.002 82.999 83.038 scf_env_do_scf 11 5.9 0.000 0.001 71.936 71.937 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 66.688 66.689 dbcsr_multiply_generic 2055 12.4 0.108 0.111 54.328 54.623 qs_scf_new_mos 99 7.5 0.000 0.001 49.950 50.106 qs_scf_loop_do_ot 99 8.5 0.000 0.001 49.949 50.105 ot_scf_mini 99 9.5 0.002 0.002 47.472 47.601 multiply_cannon 2055 13.4 0.174 0.180 44.465 45.609 multiply_cannon_loop 2055 14.4 1.831 1.910 43.552 44.739 velocity_verlet 10 3.0 0.007 0.037 43.147 43.149 ot_mini 99 10.5 0.001 0.001 28.574 28.694 qs_ot_get_derivative 99 11.5 0.001 0.001 21.606 21.729 multiply_cannon_multrec 49320 15.4 11.059 11.854 18.108 19.137 rebuild_ks_matrix 110 8.3 0.000 0.000 14.343 14.482 qs_ks_build_kohn_sham_matrix 110 9.3 0.010 0.011 14.342 14.481 mp_waitall_1 220248 16.4 12.019 13.223 12.019 13.223 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.593 12.717 qs_ot_get_p 110 10.4 0.001 0.001 10.176 10.336 multiply_cannon_sync_h2d 49320 15.4 9.294 9.993 9.294 9.993 init_scf_run 11 5.9 0.000 0.001 8.872 8.872 scf_env_initial_rho_setup 11 6.9 0.005 0.012 8.872 8.872 multiply_cannon_metrocomm3 49320 15.4 0.081 0.086 7.322 8.601 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.827 8.369 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.223 7.782 apply_single 110 13.6 0.000 0.001 7.223 7.781 dbcsr_mm_accdrv_process 87628 16.1 3.219 3.444 6.923 7.529 sum_up_and_integrate 110 10.3 0.002 0.003 6.776 6.789 integrate_v_rspace 110 11.3 0.003 0.003 6.751 6.770 qs_ot_p2m_diag 48 11.0 0.012 0.018 6.719 6.766 mp_sum_l 6594 12.7 5.826 6.717 5.826 6.717 ot_diis_step 99 11.5 0.005 0.006 6.632 6.632 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.532 6.625 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.088 6.270 calculate_rho_elec 110 8.6 0.020 0.024 6.088 6.270 cp_dbcsr_syevd 48 12.0 0.003 0.003 5.739 5.739 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.275 5.276 cp_fm_redistribute_end 48 14.0 2.678 5.250 2.687 5.254 init_scf_loop 11 6.9 0.000 0.000 5.218 5.218 cp_fm_diag_elpa_base 48 14.0 2.556 5.142 2.564 5.156 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.121 5.124 calculate_first_density_matrix 1 7.0 0.000 0.000 4.810 4.819 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.250 4.306 calculate_dm_sparse 110 9.5 0.001 0.001 3.924 4.051 make_m2s 4110 13.4 0.061 0.065 3.895 4.013 multiply_cannon_metrocomm1 49320 15.4 0.065 0.068 2.918 3.964 wfi_extrapolate 11 7.9 0.001 0.001 3.956 3.956 make_images 4110 14.4 0.177 0.190 3.798 3.921 jit_kernel_multiply 13 15.9 3.420 3.646 3.420 3.646 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.452 3.524 grid_integrate_task_list 110 12.3 3.240 3.400 3.240 3.400 acc_transpose_blocks 49320 15.4 0.215 0.230 2.782 3.322 density_rs2pw 110 9.6 0.004 0.004 3.077 3.265 prepare_preconditioner 11 7.9 0.000 0.000 3.014 3.035 make_preconditioner 11 8.9 0.000 0.000 3.014 3.035 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.760 2.802 mp_alltoall_d11v 2046 13.8 2.224 2.744 2.224 2.744 fft_wrap_pw1pw2 1111 11.6 0.014 0.018 2.572 2.631 potential_pw2rs 110 12.3 0.006 0.006 2.391 2.404 grid_collocate_task_list 110 9.6 2.157 2.263 2.157 2.263 fft_wrap_pw1pw2_140 451 12.1 0.092 0.099 2.194 2.257 mp_sum_d 3893 11.9 1.647 2.216 1.647 2.216 transfer_rs2pw 451 10.6 0.006 0.007 1.923 2.174 fft3d_ps 1111 13.6 0.733 0.819 2.112 2.162 mp_waitany 14300 13.8 1.903 2.078 1.903 2.078 make_images_data 4110 15.4 0.043 0.047 1.767 1.929 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.887 1.914 hybrid_alltoall_any 4261 16.3 0.084 0.482 1.526 1.813 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.735 1.762 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=86.992000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=472.909091, yerr=3.058655 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/908860eada1ff98329e2139c3f6fcecbf4b94d08_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 589.955072E+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 3528 76294. MP_Allreduce 9919 558. MP_Sync 52 MP_Alltoall 1717 2272043. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 MP_comm_split 48 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 0.110 0.200 72.660 72.667 qs_mol_dyn_low 1 2.0 0.014 0.037 70.995 71.019 qs_forces 11 3.9 0.003 0.004 70.802 70.818 qs_energies 11 4.9 0.010 0.030 67.473 67.491 scf_env_do_scf 11 5.9 0.000 0.001 56.394 56.395 scf_env_do_scf_inner_loop 99 6.5 0.004 0.015 49.359 49.359 dbcsr_multiply_generic 2055 12.4 0.419 0.430 39.945 40.122 velocity_verlet 10 3.0 0.027 0.077 35.342 35.358 multiply_cannon 2055 13.4 0.223 0.244 32.957 34.117 qs_scf_new_mos 99 7.5 0.001 0.001 33.632 33.749 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.632 33.748 multiply_cannon_loop 2055 14.4 1.161 1.185 31.811 32.736 ot_scf_mini 99 9.5 0.003 0.003 31.957 32.079 ot_mini 99 10.5 0.001 0.001 18.763 18.899 multiply_cannon_multrec 24660 15.4 6.939 8.632 15.235 17.182 rebuild_ks_matrix 110 8.3 0.000 0.001 13.565 13.700 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 13.565 13.700 qs_ot_get_derivative 99 11.5 0.001 0.001 13.008 13.131 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.938 12.061 mp_waitall_1 176588 16.5 7.813 10.059 7.813 10.059 dbcsr_mm_accdrv_process 52282 16.1 6.325 7.507 8.127 8.459 multiply_cannon_metrocomm3 24660 15.4 0.073 0.074 5.336 8.353 init_scf_run 11 5.9 0.001 0.024 8.258 8.259 scf_env_initial_rho_setup 11 6.9 0.022 0.091 8.257 8.258 multiply_cannon_sync_h2d 24660 15.4 6.387 7.438 6.387 7.438 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.445 7.045 apply_single 110 13.6 0.000 0.001 6.445 7.045 init_scf_loop 11 6.9 0.000 0.000 6.993 6.994 qs_ot_get_p 110 10.4 0.030 0.054 6.628 6.781 sum_up_and_integrate 110 10.3 0.001 0.003 6.258 6.271 integrate_v_rspace 110 11.3 0.003 0.003 6.231 6.246 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.886 5.902 calculate_rho_elec 110 8.6 0.078 0.202 5.885 5.901 ot_diis_step 99 11.5 0.010 0.010 5.708 5.708 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.815 5.544 prepare_preconditioner 11 7.9 0.000 0.000 4.861 4.888 make_preconditioner 11 8.9 0.000 0.000 4.861 4.888 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 4.781 4.782 qs_ot_p2m_diag 48 11.0 0.028 0.043 4.663 4.684 make_full_inverse_cholesky 11 9.9 0.000 0.000 4.495 4.597 calculate_first_density_matrix 1 7.0 0.000 0.001 4.560 4.584 make_m2s 4110 13.4 0.057 0.060 4.253 4.569 make_images 4110 14.4 0.403 0.458 4.140 4.451 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.278 4.317 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.202 4.203 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.653 3.653 cp_fm_redistribute_end 48 14.0 1.841 3.625 1.847 3.628 cp_fm_diag_elpa_base 48 14.0 1.719 3.488 1.775 3.571 wfi_extrapolate 11 7.9 0.001 0.001 3.428 3.428 calculate_dm_sparse 110 9.5 0.001 0.001 3.324 3.354 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.287 3.351 grid_integrate_task_list 110 12.3 3.168 3.340 3.168 3.340 fft_wrap_pw1pw2 1111 11.6 0.014 0.016 3.101 3.254 density_rs2pw 110 9.6 0.004 0.004 2.939 3.151 jit_kernel_multiply 9 16.3 1.433 2.912 1.433 2.912 fft_wrap_pw1pw2_140 451 12.1 0.112 0.132 2.631 2.781 make_images_data 4110 15.4 0.048 0.053 2.280 2.622 fft3d_ps 1111 13.6 1.070 1.278 2.476 2.606 hybrid_alltoall_any 4261 16.3 0.105 0.447 1.986 2.567 mp_sum_l 6594 12.7 1.856 2.447 1.856 2.447 cp_fm_cholesky_invert 11 10.9 2.371 2.386 2.371 2.386 grid_collocate_task_list 110 9.6 2.168 2.341 2.168 2.341 qs_energies_init_hamiltonians 11 5.9 0.017 0.031 2.282 2.300 mp_alltoall_d11v 2046 13.8 1.995 2.278 1.995 2.278 acc_transpose_blocks 24660 15.4 0.115 0.119 2.225 2.269 potential_pw2rs 110 12.3 0.008 0.009 2.139 2.151 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.984 1.997 multiply_cannon_metrocomm4 22605 15.4 0.079 0.083 0.782 1.904 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.885 1.899 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.788 1.800 transfer_rs2pw 451 10.6 0.007 0.007 1.539 1.755 mp_irecv_dv 57340 16.2 0.649 1.676 0.649 1.676 mp_sum_d 3893 11.9 1.157 1.597 1.157 1.597 dbcsr_complete_redistribute 325 12.2 0.292 0.423 1.311 1.498 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.006 1.392 1.498 mp_allgather_i34 2055 14.4 0.443 1.464 0.443 1.464 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=72.667000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=557.363636, yerr=7.010028 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/908860eada1ff98329e2139c3f6fcecbf4b94d08_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 664.715264E+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 3480 77281. 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 0.044 0.088 61.427 61.428 qs_mol_dyn_low 1 2.0 0.009 0.052 61.024 61.060 qs_forces 11 3.9 0.003 0.003 60.867 60.868 qs_energies 11 4.9 0.017 0.028 57.702 57.706 scf_env_do_scf 11 5.9 0.001 0.001 47.808 47.808 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 40.315 40.315 velocity_verlet 10 3.0 0.006 0.036 31.324 31.332 dbcsr_multiply_generic 2055 12.4 0.111 0.116 30.904 31.131 qs_scf_new_mos 99 7.5 0.001 0.001 25.914 26.016 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.913 26.016 multiply_cannon 2055 13.4 0.210 0.219 24.233 25.627 ot_scf_mini 99 9.5 0.002 0.003 24.643 24.757 multiply_cannon_loop 2055 14.4 0.820 0.844 23.067 24.271 ot_mini 99 10.5 0.001 0.001 14.427 14.542 multiply_cannon_multrec 16440 15.4 3.690 5.098 10.944 12.351 rebuild_ks_matrix 110 8.3 0.000 0.000 12.158 12.290 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.157 12.289 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.722 10.841 mp_waitall_1 139946 16.5 7.278 10.212 7.278 10.212 qs_ot_get_derivative 99 11.5 0.001 0.001 10.007 10.124 init_scf_loop 11 6.9 0.000 0.000 7.457 7.458 multiply_cannon_metrocomm3 16440 15.4 0.045 0.046 4.584 7.445 dbcsr_mm_accdrv_process 34862 16.1 6.409 6.856 7.098 7.381 init_scf_run 11 5.9 0.000 0.001 7.048 7.048 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.048 7.048 sum_up_and_integrate 110 10.3 0.001 0.002 6.174 6.187 integrate_v_rspace 110 11.3 0.003 0.003 6.148 6.162 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.128 5.773 apply_single 110 13.6 0.000 0.000 5.127 5.773 prepare_preconditioner 11 7.9 0.000 0.000 5.650 5.667 make_preconditioner 11 8.9 0.000 0.000 5.650 5.667 qs_ot_get_p 110 10.4 0.001 0.004 5.389 5.528 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.510 5.519 calculate_rho_elec 110 8.6 0.058 0.058 5.509 5.518 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.100 5.358 make_m2s 4110 13.4 0.050 0.052 4.358 4.704 make_images 4110 14.4 0.397 0.518 4.242 4.588 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.530 4.379 ot_diis_step 99 11.5 0.011 0.011 4.377 4.377 calculate_first_density_matrix 1 7.0 0.000 0.000 4.065 4.066 multiply_cannon_sync_h2d 16440 15.4 3.262 3.834 3.262 3.834 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.736 3.741 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.472 3.473 calculate_dm_sparse 110 9.5 0.001 0.001 3.367 3.403 grid_integrate_task_list 110 12.3 3.197 3.386 3.197 3.386 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.365 3.366 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.106 3.154 fft_wrap_pw1pw2 1111 11.6 0.013 0.017 3.066 3.073 density_rs2pw 110 9.6 0.004 0.004 2.732 2.931 wfi_extrapolate 11 7.9 0.001 0.001 2.892 2.892 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.811 2.822 cp_fm_diag_elpa_base 48 14.0 2.770 2.785 2.810 2.821 make_images_data 4110 15.4 0.045 0.050 2.342 2.792 hybrid_alltoall_any 4261 16.3 0.109 0.383 2.092 2.679 fft_wrap_pw1pw2_140 451 12.1 0.127 0.131 2.629 2.640 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.500 2.569 mp_sum_l 6594 12.7 1.712 2.495 1.712 2.495 acc_transpose_blocks 16440 15.4 0.077 0.079 1.841 2.446 multiply_cannon_metrocomm4 14385 15.4 0.048 0.054 0.872 2.433 fft3d_ps 1111 13.6 1.088 1.108 2.379 2.393 cp_fm_cholesky_invert 11 10.9 2.358 2.370 2.358 2.370 grid_collocate_task_list 110 9.6 2.219 2.343 2.219 2.343 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 2.286 2.300 mp_irecv_dv 48980 15.7 0.796 2.295 0.796 2.295 mp_alltoall_d11v 2046 13.8 1.862 2.165 1.862 2.165 potential_pw2rs 110 12.3 0.010 0.010 2.013 2.017 qs_energies_init_hamiltonians 11 5.9 0.003 0.010 1.991 1.992 dbcsr_complete_redistribute 325 12.2 0.300 0.366 1.380 1.688 acc_transpose_blocks_kernels 16440 16.4 0.218 0.222 1.048 1.653 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.634 1.647 transfer_rs2pw 451 10.6 0.005 0.006 1.276 1.487 build_core_hamiltonian_matrix_ 11 4.9 0.002 0.012 1.382 1.485 mp_allgather_i34 2055 14.4 0.481 1.456 0.481 1.456 jit_kernel_transpose 5 15.6 0.831 1.434 0.831 1.434 mp_waitany 17072 13.8 1.143 1.390 1.143 1.390 cp_fm_cholesky_decompose 22 10.9 1.324 1.358 1.324 1.358 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.283 1.293 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=61.428000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=630.181818, yerr=7.493248 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/908860eada1ff98329e2139c3f6fcecbf4b94d08_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 743.194624E+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 3480 77279. 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 0.031 0.058 65.255 65.255 qs_mol_dyn_low 1 2.0 0.003 0.004 64.686 64.698 qs_forces 11 3.9 0.016 0.017 64.009 64.009 qs_energies 11 4.9 0.001 0.001 60.622 60.630 scf_env_do_scf 11 5.9 0.000 0.001 51.429 51.435 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 42.138 42.139 velocity_verlet 10 3.0 0.002 0.002 35.323 35.351 dbcsr_multiply_generic 2055 12.4 0.119 0.123 31.673 31.952 qs_scf_new_mos 99 7.5 0.001 0.001 27.812 27.941 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.812 27.941 ot_scf_mini 99 9.5 0.002 0.003 26.116 26.235 multiply_cannon 2055 13.4 0.237 0.258 24.523 25.627 multiply_cannon_loop 2055 14.4 1.395 1.490 23.164 23.942 ot_mini 99 10.5 0.001 0.001 15.271 15.424 multiply_cannon_multrec 24660 15.4 4.168 7.340 13.707 14.819 rebuild_ks_matrix 110 8.3 0.000 0.000 11.879 11.991 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.878 11.990 qs_ot_get_derivative 99 11.5 0.001 0.001 11.074 11.196 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.505 10.613 dbcsr_mm_accdrv_process 52304 16.0 8.148 9.851 9.384 10.424 init_scf_loop 11 6.9 0.000 0.000 9.252 9.252 prepare_preconditioner 11 7.9 0.000 0.000 7.482 7.498 make_preconditioner 11 8.9 0.000 0.000 7.482 7.498 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.324 7.132 init_scf_run 11 5.9 0.000 0.001 6.477 6.477 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.476 6.477 mp_waitall_1 121746 16.5 4.209 6.162 4.209 6.162 sum_up_and_integrate 110 10.3 0.001 0.002 6.066 6.078 integrate_v_rspace 110 11.3 0.003 0.003 6.040 6.052 qs_ot_get_p 110 10.4 0.013 0.031 5.744 5.910 make_m2s 4110 13.4 0.059 0.061 5.444 5.739 make_images 4110 14.4 0.574 0.695 5.305 5.595 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.586 5.595 calculate_rho_elec 110 8.6 0.077 0.081 5.585 5.595 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.059 4.158 apply_single 110 13.6 0.000 0.001 4.059 4.158 ot_diis_step 99 11.5 0.011 0.011 4.156 4.156 qs_ot_p2m_diag 48 11.0 0.054 0.063 3.854 3.870 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.536 3.594 calculate_first_density_matrix 1 7.0 0.000 0.000 3.465 3.468 grid_integrate_task_list 110 12.3 3.280 3.450 3.280 3.450 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.359 3.424 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.317 3.318 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.293 3.293 fft_wrap_pw1pw2 1111 11.6 0.013 0.017 3.169 3.195 multiply_cannon_metrocomm3 24660 15.4 0.038 0.039 1.390 3.133 calculate_dm_sparse 110 9.5 0.001 0.001 3.057 3.086 make_images_data 4110 15.4 0.048 0.053 2.686 3.010 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.925 2.965 density_rs2pw 110 9.6 0.003 0.004 2.713 2.951 hybrid_alltoall_any 4261 16.3 0.121 0.458 2.241 2.933 cp_fm_upper_to_full 70 14.2 2.063 2.917 2.063 2.917 wfi_extrapolate 11 7.9 0.001 0.001 2.913 2.913 acc_transpose_blocks 24660 15.4 0.110 0.113 2.725 2.886 dbcsr_complete_redistribute 325 12.2 0.392 0.455 2.084 2.782 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.745 2.756 cp_fm_diag_elpa_base 48 14.0 2.681 2.706 2.742 2.752 fft_wrap_pw1pw2_140 451 12.1 0.146 0.149 2.705 2.732 multiply_cannon_sync_h2d 24660 15.4 2.366 2.530 2.366 2.530 fft3d_ps 1111 13.6 1.129 1.172 2.437 2.454 cp_fm_cholesky_invert 11 10.9 2.432 2.442 2.432 2.442 grid_collocate_task_list 110 9.6 2.262 2.380 2.262 2.380 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.565 2.248 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.236 2.240 jit_kernel_multiply 8 15.6 0.887 2.072 0.887 2.072 mp_alltoall_d11v 2046 13.8 1.815 1.977 1.815 1.977 potential_pw2rs 110 12.3 0.012 0.013 1.883 1.886 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.811 1.841 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.149 1.813 mp_alltoall_i22 605 13.7 1.034 1.765 1.034 1.765 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.617 1.713 mp_sum_l 6594 12.7 1.009 1.709 1.009 1.709 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.620 1.631 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.609 1.630 mp_allgather_i34 2055 14.4 0.453 1.549 0.453 1.549 acc_transpose_blocks_sync 73980 16.4 1.384 1.517 1.384 1.517 multiply_cannon_metrocomm4 20550 15.4 0.059 0.063 0.880 1.461 transfer_rs2pw 451 10.6 0.005 0.006 1.180 1.437 cp_fm_cholesky_decompose 22 10.9 1.364 1.412 1.364 1.412 mp_irecv_dv 62702 16.1 0.775 1.383 0.775 1.383 mp_waitany 13376 13.8 1.028 1.344 1.028 1.344 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=65.255000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=701.454545, yerr=11.896246 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/908860eada1ff98329e2139c3f6fcecbf4b94d08_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 851.120128E+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 258767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3480 77277. 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 0.035 0.082 55.666 55.667 qs_mol_dyn_low 1 2.0 0.003 0.004 55.323 55.342 qs_forces 11 3.9 0.003 0.003 54.964 54.965 qs_energies 11 4.9 0.001 0.002 51.338 51.344 scf_env_do_scf 11 5.9 0.000 0.001 41.948 41.948 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 34.881 34.882 velocity_verlet 10 3.0 0.047 0.057 30.380 30.401 dbcsr_multiply_generic 2055 12.4 0.129 0.183 23.836 24.020 qs_scf_new_mos 99 7.5 0.001 0.001 20.667 20.713 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.667 20.712 multiply_cannon 2055 13.4 0.239 0.245 18.179 19.546 ot_scf_mini 99 9.5 0.002 0.002 19.414 19.439 multiply_cannon_loop 2055 14.4 0.607 0.629 16.904 17.289 rebuild_ks_matrix 110 8.3 0.000 0.000 11.484 11.521 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.012 11.483 11.521 ot_mini 99 10.5 0.001 0.001 10.733 10.749 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.217 10.252 multiply_cannon_multrec 8220 15.4 3.181 4.289 8.029 8.925 mp_waitall_1 103326 16.6 5.850 7.408 5.850 7.408 qs_ot_get_derivative 99 11.5 0.001 0.001 7.050 7.076 init_scf_loop 11 6.9 0.000 0.000 7.020 7.021 init_scf_run 11 5.9 0.000 0.001 6.286 6.286 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.286 6.286 sum_up_and_integrate 110 10.3 0.001 0.002 6.020 6.033 integrate_v_rspace 110 11.3 0.003 0.003 5.993 6.006 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.719 5.732 calculate_rho_elec 110 8.6 0.115 0.116 5.718 5.732 dbcsr_mm_accdrv_process 17442 15.9 3.120 4.111 4.708 5.603 prepare_preconditioner 11 7.9 0.000 0.000 5.393 5.397 make_preconditioner 11 8.9 0.000 0.000 5.393 5.397 make_full_inverse_cholesky 11 9.9 0.000 0.000 4.998 5.073 qs_ot_get_p 110 10.4 0.002 0.002 4.726 4.746 multiply_cannon_metrocomm3 8220 15.4 0.018 0.019 2.991 4.607 make_m2s 4110 13.4 0.038 0.039 4.258 4.547 make_images 4110 14.4 0.643 0.707 4.128 4.418 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.723 3.781 apply_single 110 13.6 0.000 0.000 3.722 3.781 calculate_first_density_matrix 1 7.0 0.000 0.000 3.683 3.685 ot_diis_step 99 11.5 0.012 0.014 3.661 3.662 grid_integrate_task_list 110 12.3 3.373 3.518 3.373 3.518 qs_ot_p2m_diag 48 11.0 0.081 0.085 3.352 3.357 fft_wrap_pw1pw2 1111 11.6 0.014 0.016 3.319 3.325 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.075 3.076 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.015 3.016 fft_wrap_pw1pw2_140 451 12.1 0.192 0.196 2.863 2.871 density_rs2pw 110 9.6 0.004 0.004 2.679 2.822 hybrid_alltoall_any 4261 16.3 0.200 0.849 2.282 2.745 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.699 2.711 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.671 2.672 make_images_data 4110 15.4 0.041 0.047 2.264 2.636 calculate_dm_sparse 110 9.5 0.001 0.001 2.541 2.585 cp_fm_cholesky_invert 11 10.9 2.557 2.565 2.557 2.565 wfi_extrapolate 11 7.9 0.001 0.001 2.509 2.510 fft3d_ps 1111 13.6 1.270 1.277 2.478 2.489 multiply_cannon_sync_h2d 8220 15.4 2.378 2.486 2.378 2.486 grid_collocate_task_list 110 9.6 2.367 2.478 2.367 2.478 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.432 2.440 cp_fm_diag_elpa_base 48 14.0 2.377 2.400 2.430 2.438 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.015 2.047 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.788 1.991 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.917 1.937 mp_alltoall_d11v 2046 13.8 1.754 1.905 1.754 1.905 potential_pw2rs 110 12.3 0.015 0.015 1.823 1.826 jit_kernel_multiply 8 15.5 1.267 1.785 1.267 1.785 qs_env_update_s_mstruct 11 6.9 0.021 0.038 1.499 1.619 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.598 1.602 mp_allgather_i34 2055 14.4 0.443 1.544 0.443 1.544 dbcsr_complete_redistribute 325 12.2 0.528 0.562 1.428 1.515 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.491 1.502 acc_transpose_blocks 8220 15.4 0.039 0.041 1.441 1.478 cp_fm_cholesky_decompose 22 10.9 1.419 1.438 1.419 1.438 make_basis_sm 11 9.8 0.000 0.000 1.428 1.430 qs_create_task_list 11 7.9 0.000 0.001 1.204 1.304 generate_qs_task_list 11 8.9 0.375 0.443 1.204 1.303 transfer_rs2pw 451 10.6 0.005 0.006 1.112 1.269 parallel_gemm_fm 81 9.0 0.000 0.000 1.231 1.232 parallel_gemm_fm_cosma 81 10.0 1.230 1.232 1.230 1.232 mp_waitany 9240 13.8 1.040 1.213 1.040 1.213 copy_dbcsr_to_fm 151 11.3 0.002 0.003 1.188 1.210 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=55.667000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=805.272727, yerr=9.323391 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/908860eada1ff98329e2139c3f6fcecbf4b94d08_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 1.341882E+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 3469 77990. 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 0.032 0.060 81.222 81.222 qs_mol_dyn_low 1 2.0 0.003 0.003 80.610 80.622 qs_forces 11 3.9 0.003 0.003 80.541 80.542 qs_energies 11 4.9 0.001 0.001 76.413 76.415 scf_env_do_scf 11 5.9 0.000 0.001 65.656 65.656 velocity_verlet 10 3.0 0.002 0.002 48.456 48.462 scf_env_do_scf_inner_loop 99 6.5 0.015 0.029 44.633 44.635 dbcsr_multiply_generic 2055 12.4 0.122 0.127 31.036 31.119 qs_scf_new_mos 99 7.5 0.001 0.001 27.632 27.681 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.631 27.681 ot_scf_mini 99 9.5 0.002 0.002 25.814 25.854 multiply_cannon 2055 13.4 0.336 0.356 23.496 24.283 multiply_cannon_loop 2055 14.4 0.837 0.912 21.651 22.056 init_scf_loop 11 6.9 0.000 0.000 20.951 20.956 prepare_preconditioner 11 7.9 0.000 0.000 18.598 18.605 make_preconditioner 11 8.9 0.000 0.000 18.598 18.605 make_full_inverse_cholesky 11 9.9 0.000 0.000 14.592 18.101 ot_mini 99 10.5 0.001 0.001 14.858 14.890 rebuild_ks_matrix 110 8.3 0.000 0.001 13.476 13.520 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 13.476 13.520 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.235 12.276 cp_fm_upper_to_full 70 14.2 8.316 11.926 8.316 11.926 multiply_cannon_multrec 8220 15.4 4.171 4.302 10.335 10.581 qs_ot_get_derivative 99 11.5 0.001 0.001 10.141 10.180 mp_waitall_1 84994 16.7 8.067 8.876 8.067 8.876 dbcsr_complete_redistribute 325 12.2 0.929 0.959 5.545 7.686 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.843 6.882 calculate_rho_elec 110 8.6 0.227 0.227 6.842 6.881 copy_fm_to_dbcsr 174 11.2 0.001 0.001 4.542 6.675 sum_up_and_integrate 110 10.3 0.002 0.002 6.606 6.619 integrate_v_rspace 110 11.3 0.004 0.004 6.578 6.590 init_scf_run 11 5.9 0.000 0.001 6.440 6.440 scf_env_initial_rho_setup 11 6.9 0.001 0.002 6.440 6.440 dbcsr_mm_accdrv_process 11614 15.7 4.001 4.461 6.015 6.331 make_m2s 4110 13.4 0.044 0.044 5.660 6.168 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 3.990 6.102 make_images 4110 14.4 0.882 0.929 5.471 5.978 qs_ot_get_p 110 10.4 0.278 0.280 5.940 5.978 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 5.184 5.816 mp_alltoall_i22 605 13.7 3.614 5.755 3.614 5.755 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.892 5.604 apply_single 110 13.6 0.000 0.000 4.892 5.604 ot_diis_step 99 11.5 0.015 0.015 4.694 4.694 fft_wrap_pw1pw2 1111 11.6 0.016 0.016 4.486 4.491 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.499 3.989 fft_wrap_pw1pw2_140 451 12.1 0.363 0.364 3.913 3.923 qs_ot_p2m_diag 48 11.0 0.151 0.155 3.795 3.803 cp_fm_cholesky_invert 11 10.9 3.791 3.800 3.791 3.800 hybrid_alltoall_any 4261 16.3 0.265 0.576 3.010 3.769 grid_integrate_task_list 110 12.3 3.683 3.757 3.683 3.757 qs_energies_init_hamiltonians 11 5.9 0.033 0.042 3.726 3.726 make_images_data 4110 15.4 0.044 0.048 3.023 3.665 calculate_dm_sparse 110 9.5 0.001 0.001 3.340 3.362 fft3d_ps 1111 13.6 1.829 1.844 3.345 3.353 density_rs2pw 110 9.6 0.004 0.004 3.301 3.313 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.276 3.276 wfi_extrapolate 11 7.9 0.001 0.001 3.202 3.202 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.152 3.154 multiply_cannon_sync_h2d 8220 15.4 3.122 3.149 3.122 3.149 calculate_first_density_matrix 1 7.0 0.000 0.000 3.127 3.128 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.686 2.705 grid_collocate_task_list 110 9.6 2.672 2.683 2.672 2.683 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.676 2.677 cp_fm_diag_elpa_base 48 14.0 2.378 2.454 2.674 2.675 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.648 2.657 qs_env_update_s_mstruct 11 6.9 0.012 0.022 2.209 2.256 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.123 2.215 jit_kernel_multiply 10 15.3 1.817 2.152 1.817 2.152 potential_pw2rs 110 12.3 0.021 0.021 2.136 2.143 mp_alltoall_d11v 2046 13.8 2.067 2.121 2.067 2.121 qs_create_task_list 11 7.9 0.000 0.001 1.886 1.932 generate_qs_task_list 11 8.9 0.732 0.788 1.886 1.931 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.845 1.849 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.820 1.835 acc_transpose_blocks 8220 15.4 0.040 0.041 1.797 1.827 cp_fm_cholesky_decompose 22 10.9 1.754 1.777 1.754 1.777 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.732 1.743 qs_ot_get_orbitals 99 10.5 0.000 0.000 1.645 1.667 mp_sum_l 6594 12.7 0.980 1.648 0.980 1.648 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=81.222000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1229.909091, yerr=37.780357 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/908860eada1ff98329e2139c3f6fcecbf4b94d08_performance_tests/15/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 1.094965E+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 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 631.058432E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175955383376 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4009 67203. MP_Allreduce 11084 796. MP_Sync 87 MP_Alltoall 2226 2529087. MP_SendRecv 24320 18752. MP_ISendRecv 24320 18752. MP_Wait 42476 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.040 0.064 206.988 206.989 qs_mol_dyn_low 1 2.0 0.007 0.031 206.417 206.444 qs_forces 11 3.9 0.005 0.008 206.221 206.225 qs_energies 11 4.9 0.001 0.002 200.585 200.608 scf_env_do_scf 11 5.9 0.001 0.001 183.594 183.610 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 164.600 164.602 dbcsr_multiply_generic 2507 12.6 0.178 0.182 127.442 128.366 qs_scf_new_mos 117 7.6 0.001 0.001 125.359 125.721 qs_scf_loop_do_ot 117 8.6 0.001 0.001 125.359 125.720 velocity_verlet 10 3.0 0.015 0.095 122.284 122.287 ot_scf_mini 117 9.6 0.003 0.003 118.613 118.919 multiply_cannon 2507 13.6 0.237 0.248 103.053 105.016 multiply_cannon_loop 2507 14.6 2.392 2.446 100.845 103.014 ot_mini 117 10.6 0.001 0.001 67.223 67.578 multiply_cannon_multrec 60168 15.6 31.818 34.265 41.975 44.016 qs_ot_get_derivative 117 11.6 0.001 0.001 42.285 42.617 rebuild_ks_matrix 128 8.3 0.001 0.001 33.745 34.127 qs_ks_build_kohn_sham_matrix 128 9.3 0.014 0.017 33.745 34.127 mp_waitall_1 267128 16.5 29.869 33.827 29.869 33.827 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.316 30.696 qs_ot_get_p 128 10.4 0.002 0.007 28.956 29.447 multiply_cannon_sync_h2d 60168 15.6 26.382 29.020 26.382 29.020 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.448 25.456 apply_single 128 13.6 0.001 0.001 24.448 25.455 ot_diis_step 117 11.6 0.007 0.008 24.695 24.697 qs_ot_p2m_diag 83 11.4 0.079 0.091 22.267 22.374 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 20.333 20.575 multiply_cannon_metrocomm3 60168 15.6 0.119 0.127 16.628 19.482 cp_dbcsr_syevd 83 12.4 0.005 0.005 19.393 19.394 init_scf_loop 11 6.9 0.000 0.000 18.911 18.913 cp_fm_diag_elpa 83 13.4 0.000 0.000 16.291 16.333 cp_fm_diag_elpa_base 83 14.4 16.242 16.278 16.285 16.325 prepare_preconditioner 11 7.9 0.000 0.000 14.229 14.277 make_preconditioner 11 8.9 0.000 0.000 14.229 14.277 make_m2s 5014 13.6 0.105 0.115 13.788 14.157 make_images 5014 14.6 0.399 0.419 13.602 13.981 sum_up_and_integrate 128 10.3 0.002 0.004 13.786 13.799 integrate_v_rspace 128 11.3 0.004 0.004 13.727 13.742 make_full_inverse_cholesky 11 9.9 0.000 0.000 13.492 13.659 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.811 12.995 calculate_rho_elec 128 8.7 0.045 0.064 12.810 12.995 init_scf_run 11 5.9 0.000 0.002 12.903 12.904 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.903 12.904 mp_sum_l 7950 12.9 9.351 10.842 9.351 10.842 dbcsr_mm_accdrv_process 124484 16.2 4.744 4.880 9.725 10.320 multiply_cannon_metrocomm1 60168 15.6 0.096 0.100 6.768 8.994 wfi_extrapolate 11 7.9 0.001 0.001 8.957 8.957 calculate_dm_sparse 128 9.5 0.001 0.001 8.682 8.800 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.572 8.727 qs_ot_get_orbitals 117 10.6 0.001 0.001 8.292 8.398 cp_fm_cholesky_invert 11 10.9 8.158 8.176 8.158 8.176 make_images_data 5014 15.6 0.066 0.071 6.726 7.577 grid_integrate_task_list 128 12.3 7.023 7.460 7.023 7.460 hybrid_alltoall_any 5200 16.5 0.294 2.276 5.884 7.175 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 7.158 7.170 density_rs2pw 128 9.7 0.005 0.006 6.404 6.870 fft_wrap_pw1pw2 1291 11.7 0.018 0.025 6.527 6.823 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 6.110 6.232 mp_alltoall_d11v 2415 14.1 4.537 5.936 4.537 5.936 fft_wrap_pw1pw2_140 523 12.2 0.234 0.244 5.624 5.797 fft3d_ps 1291 13.7 2.097 2.607 5.346 5.586 grid_collocate_task_list 128 9.7 4.836 5.259 4.836 5.259 mp_sum_d 4474 12.1 3.500 4.451 3.500 4.451 potential_pw2rs 128 12.3 0.009 0.010 4.291 4.305 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=206.989000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=596.727273, yerr=6.579457 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/908860eada1ff98329e2139c3f6fcecbf4b94d08_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 860.344320E+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 4027 67357. MP_Allreduce 11128 958. MP_Sync 87 MP_Alltoall 1969 5847593. 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 0.026 0.050 182.885 182.885 qs_mol_dyn_low 1 2.0 0.003 0.004 182.484 182.498 qs_forces 11 3.9 0.004 0.005 182.395 182.395 qs_energies 11 4.9 0.001 0.002 175.636 175.647 scf_env_do_scf 11 5.9 0.001 0.001 159.280 159.297 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 133.141 133.144 velocity_verlet 10 3.0 0.005 0.009 112.679 112.680 dbcsr_multiply_generic 2507 12.6 0.188 0.195 98.036 99.322 qs_scf_new_mos 117 7.6 0.001 0.001 95.012 95.578 qs_scf_loop_do_ot 117 8.6 0.001 0.001 95.011 95.577 ot_scf_mini 117 9.6 0.004 0.004 90.185 90.813 multiply_cannon 2507 13.6 0.504 0.570 77.809 82.563 multiply_cannon_loop 2507 14.6 1.567 1.631 74.366 77.226 ot_mini 117 10.6 0.001 0.001 50.774 51.347 mp_waitall_1 214728 16.6 24.576 40.703 24.576 40.703 multiply_cannon_multrec 30084 15.6 20.654 25.338 31.529 36.646 rebuild_ks_matrix 128 8.3 0.001 0.001 32.367 33.177 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.019 32.367 33.177 multiply_cannon_metrocomm3 30084 15.6 0.097 0.103 16.177 30.447 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.128 29.855 qs_ot_get_derivative 117 11.6 0.001 0.002 28.549 29.168 init_scf_loop 11 6.9 0.000 0.000 26.038 26.039 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.396 23.571 apply_single 128 13.6 0.001 0.001 22.396 23.571 qs_ot_get_p 128 10.4 0.001 0.002 21.430 22.092 ot_diis_step 117 11.6 0.014 0.015 22.045 22.047 prepare_preconditioner 11 7.9 0.000 0.000 21.714 21.779 make_preconditioner 11 8.9 0.000 0.000 21.714 21.779 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.569 20.977 multiply_cannon_sync_h2d 30084 15.6 17.843 20.832 17.843 20.832 qs_ot_p2m_diag 83 11.4 0.188 0.217 16.568 16.602 make_m2s 5014 13.6 0.088 0.095 14.203 15.941 make_images 5014 14.6 1.163 1.360 13.984 15.720 cp_dbcsr_syevd 83 12.4 0.006 0.006 15.347 15.348 sum_up_and_integrate 128 10.3 0.002 0.004 13.693 13.721 integrate_v_rspace 128 11.3 0.003 0.004 13.632 13.662 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.816 12.847 calculate_rho_elec 128 8.7 0.088 0.106 12.816 12.847 cp_fm_diag_elpa 83 13.4 0.000 0.000 11.944 11.982 cp_fm_diag_elpa_base 83 14.4 11.767 11.807 11.933 11.968 cp_fm_cholesky_invert 11 10.9 11.918 11.956 11.918 11.956 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 11.247 11.681 init_scf_run 11 5.9 0.000 0.001 11.592 11.594 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.592 11.593 multiply_cannon_metrocomm4 27577 15.6 0.106 0.120 3.913 10.932 dbcsr_mm_accdrv_process 62242 16.2 5.610 6.244 10.318 10.928 mp_irecv_dv 69486 16.3 3.701 10.523 3.701 10.523 make_images_data 5014 15.6 0.066 0.077 8.379 10.462 hybrid_alltoall_any 5200 16.5 0.350 1.542 7.189 9.874 wfi_extrapolate 11 7.9 0.001 0.001 8.224 8.224 grid_integrate_task_list 128 12.3 7.158 7.497 7.158 7.497 fft_wrap_pw1pw2 1291 11.7 0.019 0.026 7.391 7.476 density_rs2pw 128 9.7 0.006 0.006 6.532 7.107 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.335 7.093 calculate_dm_sparse 128 9.5 0.001 0.001 6.603 6.726 fft_wrap_pw1pw2_140 523 12.2 0.248 0.258 6.473 6.559 mp_sum_l 7950 12.9 4.225 6.341 4.225 6.341 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.243 6.255 fft3d_ps 1291 13.7 2.770 2.961 5.839 5.913 cp_fm_cholesky_decompose 22 10.9 5.601 5.698 5.601 5.698 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.409 5.482 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.256 5.481 grid_collocate_task_list 128 9.7 5.004 5.316 5.004 5.316 mp_alltoall_d11v 2415 14.1 4.458 5.295 4.458 5.295 mp_allgather_i34 2507 14.6 1.965 5.093 1.965 5.093 potential_pw2rs 128 12.3 0.015 0.017 4.207 4.223 mp_sum_d 4483 12.1 2.603 4.048 2.603 4.048 dbcsr_complete_redistribute 395 12.7 0.719 0.876 3.328 3.991 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.653 3.723 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=182.885000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=819.909091, yerr=3.449757 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/908860eada1ff98329e2139c3f6fcecbf4b94d08_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 1.031717E+09 # 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 3999 67815. 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 0.023 0.032 170.032 170.033 qs_mol_dyn_low 1 2.0 0.003 0.003 169.518 169.533 qs_forces 11 3.9 0.004 0.004 169.429 169.431 qs_energies 11 4.9 0.001 0.002 162.897 162.907 scf_env_do_scf 11 5.9 0.001 0.001 146.554 146.555 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 117.729 117.729 velocity_verlet 10 3.0 0.005 0.010 105.697 105.699 dbcsr_multiply_generic 2507 12.6 0.186 0.189 83.226 84.425 qs_scf_new_mos 117 7.6 0.001 0.001 81.190 81.568 qs_scf_loop_do_ot 117 8.6 0.001 0.001 81.189 81.567 ot_scf_mini 117 9.6 0.003 0.004 77.008 77.426 multiply_cannon 2507 13.6 0.504 0.521 63.426 68.285 multiply_cannon_loop 2507 14.6 1.137 1.203 60.744 63.834 ot_mini 117 10.6 0.001 0.001 43.001 43.401 mp_waitall_1 170520 16.6 25.023 35.230 25.023 35.230 rebuild_ks_matrix 128 8.3 0.001 0.001 30.264 30.890 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.016 30.264 30.890 init_scf_loop 11 6.9 0.000 0.000 28.723 28.724 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.313 27.882 multiply_cannon_multrec 20056 15.6 13.015 16.248 23.438 26.577 multiply_cannon_metrocomm3 20056 15.6 0.060 0.063 15.602 25.915 prepare_preconditioner 11 7.9 0.000 0.000 24.740 24.795 make_preconditioner 11 8.9 0.000 0.000 24.740 24.795 qs_ot_get_derivative 117 11.6 0.001 0.002 23.344 23.764 make_full_inverse_cholesky 11 9.9 0.000 0.000 22.767 23.728 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.726 20.700 apply_single 128 13.6 0.001 0.001 19.726 20.700 qs_ot_get_p 128 10.4 0.001 0.002 19.379 19.865 ot_diis_step 117 11.6 0.018 0.019 19.547 19.547 make_m2s 5014 13.6 0.079 0.084 14.732 15.460 make_images 5014 14.6 1.140 1.224 14.494 15.215 qs_ot_p2m_diag 83 11.4 0.265 0.272 15.110 15.119 multiply_cannon_sync_h2d 20056 15.6 13.589 15.104 13.589 15.104 cp_dbcsr_syevd 83 12.4 0.005 0.006 14.044 14.045 sum_up_and_integrate 128 10.3 0.002 0.003 13.660 13.688 integrate_v_rspace 128 11.3 0.003 0.004 13.600 13.627 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.878 12.908 calculate_rho_elec 128 8.7 0.132 0.146 12.877 12.907 cp_fm_cholesky_invert 11 10.9 11.859 11.880 11.859 11.880 init_scf_run 11 5.9 0.000 0.001 11.416 11.416 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.415 11.416 cp_fm_diag_elpa 83 13.4 0.000 0.000 10.594 10.617 cp_fm_diag_elpa_base 83 14.4 10.346 10.420 10.590 10.614 dbcsr_mm_accdrv_process 41502 16.2 6.494 7.172 9.877 10.151 make_images_data 5014 15.6 0.062 0.072 8.842 9.997 multiply_cannon_metrocomm4 17549 15.6 0.066 0.077 3.510 9.689 hybrid_alltoall_any 5200 16.5 0.451 2.055 7.603 9.474 mp_irecv_dv 50230 16.2 3.377 9.417 3.377 9.417 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 9.086 9.392 grid_integrate_task_list 128 12.3 7.314 7.797 7.314 7.797 fft_wrap_pw1pw2 1291 11.7 0.019 0.024 7.487 7.584 wfi_extrapolate 11 7.9 0.001 0.001 7.333 7.333 density_rs2pw 128 9.7 0.005 0.005 6.387 6.847 fft_wrap_pw1pw2_140 523 12.2 0.283 0.290 6.584 6.698 calculate_dm_sparse 128 9.5 0.001 0.001 6.199 6.282 cp_fm_cholesky_decompose 22 10.9 5.932 6.017 5.932 6.017 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.007 6.013 fft3d_ps 1291 13.7 2.771 2.990 5.793 5.871 dbcsr_complete_redistribute 395 12.7 1.031 1.092 4.463 5.626 grid_collocate_task_list 128 9.7 5.158 5.480 5.158 5.480 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.663 5.305 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.106 5.275 mp_alltoall_d11v 2415 14.1 4.574 5.260 4.574 5.260 cp_fm_upper_to_full 105 14.8 3.918 4.902 3.918 4.902 mp_sum_l 7950 12.9 3.230 4.817 3.230 4.817 mp_allgather_i34 2507 14.6 1.185 4.400 1.185 4.400 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.236 4.387 potential_pw2rs 128 12.3 0.019 0.022 4.117 4.126 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.982 4.018 calculate_first_density_matrix 1 7.0 0.000 0.000 3.897 3.898 qs_energies_init_hamiltonians 11 5.9 0.001 0.007 3.696 3.697 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.464 3.519 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=170.033000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=979.636364, yerr=3.723945 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/908860eada1ff98329e2139c3f6fcecbf4b94d08_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420243808256 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528908111872 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514772E+12 0.0% 0.0% 100.0% flops max/rank 4.353792E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755942624 0.0% 0.0% 100.0% number of processed stacks 5977344 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.105777E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1143192 MPI messages size (bytes): total size 2.023815E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.770320E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 319024 36042702848 131072 < size <= 4194304 715736 785529176064 4194304 < size <= 16777216 70320 665379572320 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4027 67348. MP_Allreduce 11127 1081. MP_Sync 87 MP_Alltoall 1712 12503107. MP_SendRecv 5888 75008. MP_ISendRecv 5888 75008. MP_Wait 22442 MP_ISend 14952 244818. MP_IRecv 14952 244818. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.029 177.827 177.828 qs_mol_dyn_low 1 2.0 0.003 0.003 177.423 177.437 qs_forces 11 3.9 0.004 0.004 177.309 177.320 qs_energies 11 4.9 0.001 0.002 170.211 170.219 scf_env_do_scf 11 5.9 0.001 0.001 152.212 152.227 velocity_verlet 10 3.0 0.004 0.007 115.391 115.394 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 115.256 115.258 dbcsr_multiply_generic 2507 12.6 0.191 0.197 81.024 81.775 qs_scf_new_mos 117 7.6 0.001 0.001 79.561 79.882 qs_scf_loop_do_ot 117 8.6 0.001 0.001 79.561 79.881 ot_scf_mini 117 9.6 0.003 0.003 74.991 75.323 multiply_cannon 2507 13.6 0.551 0.583 56.123 59.212 multiply_cannon_loop 2507 14.6 1.800 1.956 52.465 54.240 ot_mini 117 10.6 0.001 0.001 43.153 43.481 init_scf_loop 11 6.9 0.000 0.000 36.833 36.834 prepare_preconditioner 11 7.9 0.000 0.000 32.750 32.778 make_preconditioner 11 8.9 0.000 0.000 32.750 32.777 multiply_cannon_multrec 30084 15.6 13.672 18.858 26.786 31.670 make_full_inverse_cholesky 11 9.9 0.018 0.024 27.990 31.570 rebuild_ks_matrix 128 8.3 0.001 0.001 29.235 29.545 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.018 29.234 29.545 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.374 26.653 mp_waitall_1 147882 16.7 17.101 26.500 17.101 26.500 qs_ot_get_derivative 117 11.6 0.001 0.002 23.468 23.801 make_m2s 5014 13.6 0.093 0.096 20.393 21.358 make_images 5014 14.6 1.998 2.275 20.094 21.061 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 19.153 19.709 apply_single 128 13.6 0.001 0.001 19.153 19.709 ot_diis_step 117 11.6 0.017 0.020 19.560 19.561 qs_ot_get_p 128 10.4 0.001 0.002 17.961 18.284 multiply_cannon_metrocomm3 30084 15.6 0.049 0.052 6.325 14.593 qs_ot_p2m_diag 83 11.4 0.343 0.390 13.776 13.839 sum_up_and_integrate 128 10.3 0.002 0.003 13.636 13.663 integrate_v_rspace 128 11.3 0.003 0.004 13.576 13.605 dbcsr_mm_accdrv_process 62264 16.2 8.442 9.602 12.685 13.202 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.155 13.177 calculate_rho_elec 128 8.7 0.174 0.190 13.155 13.177 cp_fm_upper_to_full 105 14.8 8.718 12.409 8.718 12.409 cp_dbcsr_syevd 83 12.4 0.005 0.006 12.381 12.382 make_images_data 5014 15.6 0.065 0.070 10.868 12.257 cp_fm_cholesky_invert 11 10.9 12.016 12.037 12.016 12.037 init_scf_run 11 5.9 0.000 0.001 11.656 11.657 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.656 11.657 multiply_cannon_sync_h2d 30084 15.6 10.428 11.227 10.428 11.227 hybrid_alltoall_any 5200 16.5 0.530 2.224 9.892 11.010 dbcsr_complete_redistribute 395 12.7 1.413 1.505 7.513 10.314 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.852 10.084 cp_fm_diag_elpa 83 13.4 0.000 0.000 9.190 9.210 cp_fm_diag_elpa_base 83 14.4 8.714 8.851 9.183 9.197 copy_fm_to_dbcsr 209 11.7 0.001 0.002 6.167 8.955 grid_integrate_task_list 128 12.3 7.533 7.921 7.533 7.921 fft_wrap_pw1pw2 1291 11.7 0.019 0.024 7.733 7.784 wfi_extrapolate 11 7.9 0.001 0.001 7.508 7.508 transfer_fm_to_dbcsr 11 9.9 0.002 0.010 4.738 7.472 multiply_cannon_metrocomm4 25070 15.6 0.079 0.089 2.799 6.992 fft_wrap_pw1pw2_140 523 12.2 0.328 0.335 6.850 6.921 mp_alltoall_i22 716 14.1 4.099 6.883 4.099 6.883 mp_irecv_dv 76098 16.2 2.644 6.721 2.644 6.721 density_rs2pw 128 9.7 0.005 0.005 6.195 6.442 calculate_dm_sparse 128 9.5 0.001 0.001 6.340 6.437 cp_fm_cholesky_decompose 22 10.9 6.013 6.072 6.013 6.072 fft3d_ps 1291 13.7 2.936 3.009 5.924 5.970 mp_alltoall_d11v 2415 14.1 4.906 5.674 4.906 5.674 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.622 5.672 grid_collocate_task_list 128 9.7 5.288 5.625 5.288 5.625 qs_energies_init_hamiltonians 11 5.9 0.004 0.016 4.859 4.861 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.550 4.669 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.510 4.592 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.233 4.298 calculate_first_density_matrix 1 7.0 0.000 0.000 3.950 3.957 potential_pw2rs 128 12.3 0.022 0.022 3.900 3.912 dbcsr_special_finalize 12535 15.6 0.067 0.069 3.529 3.755 mp_sum_l 7950 12.9 2.584 3.698 2.584 3.698 dbcsr_merge_single_wm 12535 16.6 0.327 0.346 3.437 3.661 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=177.828000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1044.181818, yerr=14.402135 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/908860eada1ff98329e2139c3f6fcecbf4b94d08_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 1.446253E+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 750341. MP_ISend 40420 2100460. MP_IRecv 40420 2099564. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4052 66938. MP_Allreduce 11189 1162. 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 0.026 0.046 166.684 166.684 qs_mol_dyn_low 1 2.0 0.003 0.004 166.093 166.106 qs_forces 11 3.9 0.004 0.004 165.985 165.989 qs_energies 11 4.9 0.002 0.002 158.612 158.618 scf_env_do_scf 11 5.9 0.001 0.001 139.606 139.616 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 108.822 108.823 velocity_verlet 10 3.0 0.013 0.015 107.244 107.247 dbcsr_multiply_generic 2529 12.6 0.210 0.263 74.201 74.742 qs_scf_new_mos 118 7.6 0.001 0.001 72.379 72.498 qs_scf_loop_do_ot 118 8.6 0.001 0.001 72.378 72.498 ot_scf_mini 118 9.6 0.003 0.004 67.927 68.053 multiply_cannon 2529 13.6 0.564 0.601 55.378 58.149 multiply_cannon_loop 2529 14.6 0.819 0.851 52.277 53.460 ot_mini 118 10.6 0.001 0.001 38.030 38.143 mp_waitall_1 126876 16.7 25.099 31.818 25.099 31.818 init_scf_loop 11 6.9 0.000 0.000 30.626 30.628 rebuild_ks_matrix 129 8.3 0.001 0.001 29.208 29.375 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.018 29.207 29.374 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.665 26.813 prepare_preconditioner 11 7.9 0.000 0.000 26.733 26.746 make_preconditioner 11 8.9 0.000 0.000 26.733 26.746 make_full_inverse_cholesky 11 9.9 0.022 0.029 24.842 25.102 multiply_cannon_multrec 10116 15.6 10.377 16.618 18.865 23.567 qs_ot_get_derivative 118 11.6 0.002 0.002 20.652 20.774 multiply_cannon_metrocomm3 10116 15.6 0.025 0.026 13.098 20.396 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 17.448 17.683 apply_single 129 13.6 0.001 0.001 17.448 17.683 ot_diis_step 118 11.6 0.020 0.021 17.305 17.306 qs_ot_get_p 129 10.4 0.002 0.002 17.123 17.274 make_m2s 5058 13.6 0.064 0.071 14.807 15.732 make_images 5058 14.6 2.167 2.608 14.497 15.423 cp_fm_cholesky_invert 11 10.9 14.720 14.731 14.720 14.731 sum_up_and_integrate 129 10.3 0.002 0.002 13.849 13.900 integrate_v_rspace 129 11.3 0.004 0.004 13.789 13.838 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.519 13.562 calculate_rho_elec 129 8.7 0.259 0.270 13.519 13.562 qs_ot_p2m_diag 84 11.4 0.501 0.507 13.338 13.354 init_scf_run 11 5.9 0.001 0.020 12.171 12.171 scf_env_initial_rho_setup 11 6.9 0.018 0.020 12.169 12.171 cp_dbcsr_syevd 84 12.4 0.006 0.006 12.167 12.167 multiply_cannon_sync_h2d 10116 15.6 10.788 11.293 10.788 11.293 make_images_data 5058 15.6 0.056 0.067 8.327 9.761 hybrid_alltoall_any 5245 16.5 0.849 3.800 8.219 9.309 cp_fm_diag_elpa 84 13.4 0.000 0.000 9.220 9.230 cp_fm_diag_elpa_base 84 14.4 8.984 9.061 9.216 9.227 dbcsr_mm_accdrv_process 20934 16.1 3.340 5.076 8.138 9.025 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 8.107 8.193 grid_integrate_task_list 129 12.3 7.855 8.148 7.855 8.148 fft_wrap_pw1pw2 1301 11.7 0.018 0.020 8.108 8.133 multiply_cannon_metrocomm1 10116 15.6 0.030 0.031 4.459 7.737 wfi_extrapolate 11 7.9 0.001 0.001 7.322 7.322 fft_wrap_pw1pw2_140 527 12.2 0.438 0.443 7.138 7.178 calculate_dm_sparse 129 9.5 0.001 0.001 6.913 6.992 cp_fm_cholesky_decompose 22 10.9 6.499 6.545 6.499 6.545 density_rs2pw 129 9.7 0.005 0.005 6.161 6.392 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.308 6.360 multiply_cannon_metrocomm4 7587 15.6 0.025 0.028 1.874 6.199 mp_irecv_dv 29102 15.9 1.832 6.097 1.832 6.097 fft3d_ps 1301 13.7 3.132 3.250 6.073 6.090 grid_collocate_task_list 129 9.7 5.634 5.766 5.634 5.766 dbcsr_complete_redistribute 397 12.7 2.255 2.333 5.449 5.752 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.260 5.261 mp_alltoall_d11v 2429 14.1 4.684 4.959 4.684 4.959 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.850 4.910 calculate_first_density_matrix 1 7.0 0.000 0.000 4.579 4.581 mp_allgather_i34 2529 14.6 1.281 4.532 1.281 4.532 potential_pw2rs 129 12.3 0.026 0.026 3.947 3.955 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.592 3.874 copy_fm_to_dbcsr 210 11.7 0.002 0.002 3.555 3.862 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.604 3.695 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.638 3.679 copy_dbcsr_to_fm 187 11.8 0.004 0.004 3.510 3.630 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.451 3.470 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=166.684000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1344.000000, yerr=26.710570 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/908860eada1ff98329e2139c3f6fcecbf4b94d08_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019187724288 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019187724288 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796577E+12 0.0% 0.0% 100.0% flops max/rank 11.606413E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500544 0.0% 0.0% 100.0% number of processed stacks 1947808 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.6 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 2.980979E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 99400 MPI messages size (bytes): total size 1.127422E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.342272E+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 44768 34745614336 4194304 < size <= 16777216 43984 376564613120 16777216 < size 10032 716108613552 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3962 69364. MP_Allreduce 10965 1506. MP_Sync 86 MP_Alltoall 1700 36954383. MP_SendRecv 1778 218624. MP_ISendRecv 1778 218624. MP_Wait 9728 MP_ISend 6360 1080477. MP_IRecv 6360 1080477. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.095 0.145 247.682 247.683 qs_mol_dyn_low 1 2.0 0.006 0.030 246.904 246.922 qs_forces 11 3.9 0.015 0.025 246.777 246.785 qs_energies 11 4.9 0.002 0.002 237.941 237.957 scf_env_do_scf 11 5.9 0.001 0.001 213.422 213.445 velocity_verlet 10 3.0 0.013 0.014 170.056 170.065 scf_env_do_scf_inner_loop 116 6.6 0.038 0.049 130.633 130.635 qs_scf_new_mos 116 7.6 0.001 0.001 88.168 88.318 qs_scf_loop_do_ot 116 8.6 0.001 0.001 88.168 88.317 dbcsr_multiply_generic 2485 12.5 0.294 0.305 84.318 85.086 ot_scf_mini 116 9.6 0.003 0.004 83.219 83.330 init_scf_loop 11 6.9 0.000 0.000 82.463 82.465 prepare_preconditioner 11 7.9 0.000 0.000 77.565 77.613 make_preconditioner 11 8.9 0.000 0.000 77.565 77.613 make_full_inverse_cholesky 11 9.9 0.038 0.039 61.580 74.721 multiply_cannon 2485 13.5 0.669 0.716 60.650 61.583 multiply_cannon_loop 2485 14.5 1.038 1.070 56.468 58.600 cp_fm_upper_to_full 104 14.8 33.928 48.228 33.928 48.228 ot_mini 116 10.6 0.001 0.001 44.779 44.904 mp_waitall_1 102768 16.8 28.827 34.113 28.827 34.113 rebuild_ks_matrix 127 8.3 0.001 0.001 33.719 33.912 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.017 33.718 33.912 qs_ks_update_qs_env 127 7.6 0.001 0.001 31.319 31.483 dbcsr_complete_redistribute 393 12.7 4.134 4.307 22.471 30.346 copy_fm_to_dbcsr 208 11.6 0.001 0.001 18.796 26.643 qs_ot_get_derivative 116 11.6 0.002 0.002 24.682 24.791 transfer_fm_to_dbcsr 11 9.9 0.030 0.030 15.942 23.695 qs_ot_get_p 127 10.4 0.006 0.006 23.344 23.465 mp_alltoall_i22 712 14.1 13.708 21.729 13.708 21.729 multiply_cannon_metrocomm3 9940 15.5 0.024 0.025 19.441 21.467 make_m2s 4970 13.5 0.074 0.077 18.784 20.422 ot_diis_step 116 11.6 0.022 0.022 20.051 20.052 make_images 4970 14.5 3.054 3.270 18.305 19.946 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.247 19.556 apply_single 127 13.6 0.001 0.001 19.246 19.556 multiply_cannon_multrec 9940 15.5 10.243 12.075 19.188 19.414 qs_ot_p2m_diag 82 11.4 0.868 0.874 19.209 19.239 cp_fm_cholesky_invert 11 10.9 18.370 18.382 18.370 18.382 cp_dbcsr_syevd 82 12.4 0.006 0.006 17.435 17.437 qs_rho_update_rho_low 127 7.7 0.001 0.001 15.673 15.689 calculate_rho_elec 127 8.7 0.479 0.480 15.672 15.689 sum_up_and_integrate 127 10.3 0.002 0.002 15.567 15.657 integrate_v_rspace 127 11.3 0.005 0.007 15.504 15.594 cp_fm_diag_elpa 82 13.4 0.000 0.000 14.304 14.305 cp_fm_diag_elpa_base 82 14.4 11.533 12.535 14.300 14.300 multiply_cannon_sync_h2d 9940 15.5 14.213 14.231 14.213 14.231 init_scf_run 11 5.9 0.000 0.001 13.961 13.961 scf_env_initial_rho_setup 11 6.9 0.001 0.001 13.961 13.961 hybrid_alltoall_any 5155 16.4 1.306 3.035 10.908 13.381 make_images_data 4970 15.5 0.062 0.067 10.631 12.980 dbcsr_mm_accdrv_process 20590 16.0 4.366 6.362 8.696 10.570 fft_wrap_pw1pw2 1281 11.7 0.020 0.020 9.947 9.961 qs_ot_get_derivative_diag 76 12.4 0.002 0.003 9.666 9.734 wfi_extrapolate 11 7.9 0.001 0.001 9.065 9.065 fft_wrap_pw1pw2_140 519 12.2 0.818 0.819 8.811 8.822 grid_integrate_task_list 127 12.3 8.499 8.699 8.499 8.699 qs_energies_init_hamiltonians 11 5.9 0.005 0.009 8.245 8.263 calculate_dm_sparse 127 9.5 0.001 0.001 7.602 7.663 cp_fm_cholesky_decompose 22 10.9 7.409 7.446 7.409 7.446 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 7.349 7.422 mp_alltoall_d11v 2401 14.1 7.193 7.334 7.193 7.334 fft3d_ps 1281 13.7 3.942 3.962 7.244 7.283 density_rs2pw 127 9.7 0.005 0.005 6.833 6.873 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.379 6.552 grid_collocate_task_list 127 9.7 6.364 6.436 6.364 6.436 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.987 5.371 qs_env_update_s_mstruct 11 6.9 0.041 0.057 5.069 5.256 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.952 5.023 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=247.683000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2663.545455, yerr=159.100519 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/908860eada1ff98329e2139c3f6fcecbf4b94d08_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 1.259409E+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 1033 290947. MP_Allreduce 3139 6114. MP_Sync 4 MP_Alltoall 54 MP_SendRecv 285 19200. MP_ISendRecv 285 19200. MP_Wait 1017 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.070 0.092 86.078 86.079 qs_energies 1 2.0 0.000 0.000 85.444 85.454 ls_scf 1 3.0 0.000 0.000 84.100 84.110 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.658 72.857 multiply_cannon 111 7.7 0.017 0.020 56.037 56.950 multiply_cannon_loop 111 8.7 0.229 0.243 52.649 53.839 ls_scf_main 1 4.0 0.000 0.000 52.318 52.319 density_matrix_trs4 2 5.0 0.002 0.003 46.718 46.806 ls_scf_init_scf 1 4.0 0.000 0.000 28.745 28.746 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.573 27.624 mp_waitall_1 11031 10.9 22.276 25.636 22.276 25.636 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 25.329 25.350 multiply_cannon_multrec 2664 9.7 8.133 8.955 15.625 17.385 multiply_cannon_sync_h2d 2664 9.7 13.650 15.324 13.650 15.324 make_m2s 222 7.7 0.009 0.011 13.127 13.577 make_images 222 8.7 0.099 0.109 13.105 13.558 multiply_cannon_metrocomm1 2664 9.7 0.010 0.012 9.531 11.625 make_images_data 222 9.7 0.004 0.005 7.676 8.200 dbcsr_mm_accdrv_process 4760 10.4 0.600 0.706 7.109 8.097 hybrid_alltoall_any 227 10.6 0.216 1.837 6.683 7.862 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.416 7.403 dbcsr_mm_accdrv_process_sort 4760 11.4 6.307 7.236 6.307 7.236 calculate_norms 4752 9.8 5.538 6.204 5.538 6.204 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.043 5.201 mp_sum_l 887 5.1 2.951 4.560 2.951 4.560 multiply_cannon_metrocomm4 2442 9.7 0.012 0.016 2.069 3.963 mp_irecv_dv 6231 10.9 2.052 3.930 2.052 3.930 make_images_sizes 222 9.7 0.000 0.000 0.754 3.540 mp_alltoall_i44 222 10.7 0.754 3.540 0.754 3.540 arnoldi_extremal 4 6.8 0.000 0.000 3.259 3.294 arnoldi_normal_ev 4 7.8 0.044 0.053 3.259 3.294 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.185 3.245 build_subspace 16 8.4 0.009 0.012 3.116 3.121 ls_scf_post 1 4.0 0.000 0.000 3.037 3.047 ls_scf_store_result 1 5.0 0.000 0.000 2.850 2.895 dbcsr_special_finalize 555 9.7 0.005 0.006 2.337 2.788 dbcsr_merge_single_wm 555 10.7 0.457 0.595 2.329 2.779 make_images_pack 222 9.7 2.213 2.623 2.215 2.625 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.308 2.549 dbcsr_sort_data 658 11.4 2.129 2.531 2.129 2.531 dbcsr_matrix_vector_mult_local 304 10.0 2.062 2.456 2.064 2.458 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.333 2.440 buffer_matrices_ensure_size 222 8.7 1.744 2.034 1.744 2.034 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.858 1.862 rebuild_ks_matrix 3 7.3 0.000 0.000 1.849 1.853 qs_ks_build_kohn_sham_matrix 3 8.3 0.027 0.046 1.849 1.853 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=86.079000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1137.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/908860eada1ff98329e2139c3f6fcecbf4b94d08_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 2.178683E+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 1033 301899. 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 0.041 0.069 90.724 90.725 qs_energies 1 2.0 0.000 0.000 89.884 89.889 ls_scf 1 3.0 0.000 0.000 88.471 88.504 dbcsr_multiply_generic 111 6.7 0.015 0.015 74.521 74.898 multiply_cannon 111 7.7 0.028 0.038 53.129 57.359 ls_scf_main 1 4.0 0.000 0.000 54.366 54.375 multiply_cannon_loop 111 8.7 0.135 0.149 50.296 53.340 density_matrix_trs4 2 5.0 0.002 0.003 48.732 48.940 ls_scf_init_scf 1 4.0 0.000 0.000 30.417 30.451 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.242 29.352 mp_waitall_1 9105 10.9 20.554 29.259 20.554 29.259 multiply_cannon_multrec 1332 9.7 13.414 17.438 22.967 27.975 matrix_sqrt_Newton_Schulz 2 6.5 0.002 0.009 26.423 26.436 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.350 19.646 make_m2s 222 7.7 0.006 0.008 14.988 15.581 make_images 222 8.7 1.368 1.688 14.957 15.551 dbcsr_mm_accdrv_process 4041 10.4 0.385 0.955 9.139 10.695 dbcsr_mm_accdrv_process_sort 4041 11.4 8.506 10.084 8.506 10.084 make_images_data 222 9.7 0.004 0.005 8.643 9.518 hybrid_alltoall_any 227 10.6 0.541 2.536 8.019 9.352 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.217 7.929 mp_irecv_dv 3311 11.0 3.196 7.860 3.196 7.860 mp_sum_l 887 5.1 4.939 7.762 4.939 7.762 calculate_norms 2376 9.8 6.067 6.705 6.067 6.705 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.727 6.081 multiply_cannon_sync_h2d 1332 9.7 4.864 5.866 4.864 5.866 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.264 5.462 arnoldi_extremal 4 6.8 0.000 0.000 4.720 4.737 arnoldi_normal_ev 4 7.8 0.038 0.087 4.720 4.737 build_subspace 16 8.4 0.014 0.021 4.385 4.388 ls_scf_post 1 4.0 0.000 0.000 3.688 3.693 ls_scf_store_result 1 5.0 0.000 0.000 3.427 3.511 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.133 3.362 dbcsr_matrix_vector_mult_local 304 10.0 2.745 3.225 2.748 3.227 mp_allgather_i34 111 8.7 0.817 3.021 0.817 3.021 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.206 2.796 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.600 2.678 dbcsr_data_new 4174 10.1 2.121 2.412 2.121 2.412 make_images_pack 222 9.7 1.824 2.103 1.827 2.106 dbcsr_sort_data 436 11.2 1.822 2.102 1.822 2.102 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.890 1.893 rebuild_ks_matrix 3 7.3 0.000 0.000 1.878 1.880 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.878 1.880 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=90.725000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1795.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/908860eada1ff98329e2139c3f6fcecbf4b94d08_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 2.880836E+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 1033 300682. 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 0.028 0.055 94.902 94.904 qs_energies 1 2.0 0.000 0.000 93.988 93.994 ls_scf 1 3.0 0.004 0.031 91.851 91.883 dbcsr_multiply_generic 111 6.7 0.016 0.019 76.102 76.455 multiply_cannon 111 7.7 0.034 0.115 52.796 57.143 ls_scf_main 1 4.0 0.001 0.005 57.046 57.052 multiply_cannon_loop 111 8.7 0.117 0.128 49.980 53.871 density_matrix_trs4 2 5.0 0.002 0.003 51.159 51.291 mp_waitall_1 7281 11.0 24.050 34.223 24.050 34.223 ls_scf_init_scf 1 4.0 0.002 0.018 31.109 31.142 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.638 29.758 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.003 27.115 27.129 multiply_cannon_multrec 888 9.7 12.665 15.055 21.430 24.496 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.390 23.594 make_m2s 222 7.7 0.006 0.007 16.548 17.364 make_images 222 8.7 1.583 1.855 16.510 17.323 make_images_data 222 9.7 0.004 0.004 9.658 10.731 hybrid_alltoall_any 227 10.6 0.635 2.922 9.157 10.405 dbcsr_mm_accdrv_process 3754 10.4 0.396 0.798 8.277 9.766 dbcsr_mm_accdrv_process_sort 3754 11.4 7.759 8.968 7.759 8.968 mp_sum_l 887 5.1 5.070 8.386 5.070 8.386 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.639 7.610 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.478 7.154 multiply_cannon_sync_h2d 888 9.7 6.076 7.138 6.076 7.138 mp_irecv_dv 2335 11.1 2.462 7.093 2.462 7.093 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.889 6.539 arnoldi_extremal 4 6.8 0.000 0.000 5.075 5.093 arnoldi_normal_ev 4 7.8 0.001 0.005 5.074 5.093 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.812 5.062 calculate_norms 1584 9.8 4.382 4.799 4.382 4.799 build_subspace 16 8.4 0.014 0.020 4.776 4.780 mp_allgather_i34 111 8.7 0.926 3.888 0.926 3.888 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.424 3.753 ls_scf_post 1 4.0 0.000 0.001 3.692 3.698 dbcsr_matrix_vector_mult_local 304 10.0 3.025 3.595 3.028 3.597 ls_scf_store_result 1 5.0 0.000 0.000 3.420 3.512 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.841 2.949 dbcsr_data_new 4116 9.9 2.119 2.495 2.119 2.495 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.186 2.204 dbcsr_sort_data 325 11.1 1.911 2.188 1.911 2.188 rebuild_ks_matrix 3 7.3 0.000 0.000 2.168 2.186 qs_ks_build_kohn_sham_matrix 3 8.3 0.035 0.275 2.168 2.186 qs_energies_init_hamiltonians 1 3.0 0.038 0.072 2.129 2.168 make_images_sizes 222 9.7 0.000 0.000 0.987 2.116 mp_alltoall_i44 222 10.7 0.986 2.116 0.986 2.116 dbcsr_finalize 304 7.8 0.026 0.032 1.614 1.900 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=94.904000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2241.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/908860eada1ff98329e2139c3f6fcecbf4b94d08_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 3.351007E+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 1033 301899. 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 0.049 0.077 99.212 99.213 qs_energies 1 2.0 0.010 0.051 98.408 98.448 ls_scf 1 3.0 0.007 0.028 96.029 96.039 dbcsr_multiply_generic 111 6.7 0.017 0.018 79.457 79.843 ls_scf_main 1 4.0 0.004 0.024 59.395 59.401 multiply_cannon 111 7.7 0.052 0.123 52.108 56.902 density_matrix_trs4 2 5.0 0.072 0.290 53.243 53.411 multiply_cannon_loop 111 8.7 0.152 0.170 47.064 50.649 ls_scf_init_scf 1 4.0 0.001 0.002 33.342 33.357 ls_scf_init_matrix_S 1 5.0 0.000 0.003 31.982 32.047 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 29.401 29.413 mp_waitall_1 6369 11.0 22.799 29.208 22.799 29.208 multiply_cannon_multrec 1332 9.7 14.185 17.396 22.280 25.740 make_m2s 222 7.7 0.006 0.008 21.106 22.530 make_images 222 8.7 3.145 3.591 21.056 22.483 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.131 17.626 make_images_data 222 9.7 0.004 0.004 11.747 13.263 hybrid_alltoall_any 227 10.6 0.800 3.797 11.178 12.825 dbcsr_mm_accdrv_process 3641 10.4 0.383 0.761 7.739 9.190 dbcsr_mm_accdrv_process_sort 3641 11.4 7.205 8.685 7.205 8.685 mp_sum_l 887 5.1 4.680 8.402 4.680 8.402 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.457 6.417 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.107 6.096 multiply_cannon_sync_h2d 1332 9.7 5.421 6.042 5.421 6.042 mp_irecv_dv 3229 10.9 2.082 6.015 2.082 6.015 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.749 5.532 arnoldi_extremal 4 6.8 0.000 0.000 5.197 5.213 arnoldi_normal_ev 4 7.8 0.001 0.004 5.197 5.212 build_subspace 16 8.4 0.014 0.021 4.857 4.866 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.619 4.819 calculate_norms 2376 9.8 4.167 4.514 4.167 4.514 mp_allgather_i34 111 8.7 2.150 4.493 2.150 4.493 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.584 3.874 dbcsr_matrix_vector_mult_local 304 10.0 3.195 3.689 3.197 3.691 dbcsr_sort_data 658 11.4 3.123 3.491 3.123 3.491 ls_scf_post 1 4.0 0.006 0.028 3.285 3.293 dbcsr_special_finalize 555 9.7 0.006 0.007 2.875 3.250 dbcsr_merge_single_wm 555 10.7 0.540 0.670 2.866 3.242 ls_scf_store_result 1 5.0 0.000 0.000 2.993 3.069 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.993 3.032 dbcsr_data_release 10477 10.7 1.580 2.415 1.580 2.415 qs_energies_init_hamiltonians 1 3.0 0.009 0.016 2.359 2.366 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.033 2.045 rebuild_ks_matrix 3 7.3 0.000 0.000 2.010 2.022 qs_ks_build_kohn_sham_matrix 3 8.3 0.011 0.046 2.010 2.022 dbcsr_finalize 304 7.8 0.049 0.061 1.803 1.987 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=99.213000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2753.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/908860eada1ff98329e2139c3f6fcecbf4b94d08_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 4.795662E+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 1033 300769. 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 0.071 0.083 92.867 92.868 qs_energies 1 2.0 0.000 0.000 91.854 91.857 ls_scf 1 3.0 0.000 0.000 89.547 89.553 dbcsr_multiply_generic 111 6.7 0.017 0.018 71.011 71.234 ls_scf_main 1 4.0 0.000 0.000 56.266 56.266 multiply_cannon 111 7.7 0.082 0.161 52.828 56.101 multiply_cannon_loop 111 8.7 0.088 0.096 50.259 51.482 density_matrix_trs4 2 5.0 0.002 0.003 49.328 49.392 ls_scf_init_scf 1 4.0 0.000 0.000 29.801 29.806 mp_waitall_1 5436 11.0 24.325 28.682 24.325 28.682 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.556 28.597 matrix_sqrt_Newton_Schulz 2 6.5 0.020 0.054 26.175 26.186 multiply_cannon_multrec 444 9.7 13.655 16.518 20.953 22.204 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.347 15.521 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.705 15.327 make_m2s 222 7.7 0.005 0.005 13.318 14.238 make_images 222 8.7 2.033 2.476 13.250 14.169 make_images_data 222 9.7 0.003 0.004 8.089 9.530 hybrid_alltoall_any 227 10.6 0.801 3.822 8.035 9.312 dbcsr_mm_accdrv_process 3003 10.4 0.470 0.696 6.987 8.243 multiply_cannon_sync_h2d 444 9.7 6.704 7.991 6.704 7.991 dbcsr_mm_accdrv_process_sort 3003 11.4 6.491 7.641 6.491 7.641 arnoldi_extremal 4 6.8 0.000 0.000 5.835 5.880 arnoldi_normal_ev 4 7.8 0.014 0.072 5.835 5.880 build_subspace 16 8.4 0.015 0.020 5.399 5.411 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.734 4.976 mp_sum_l 887 5.1 2.867 4.393 2.867 4.393 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.165 4.346 dbcsr_matrix_vector_mult_local 304 10.0 3.699 4.157 3.701 4.160 calculate_norms 792 9.8 3.627 3.757 3.627 3.757 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.546 3.667 mp_irecv_dv 1241 11.2 1.530 3.607 1.530 3.607 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.988 3.566 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.470 3.554 mp_allgather_i34 111 8.7 1.143 3.506 1.143 3.506 ls_scf_post 1 4.0 0.000 0.000 3.480 3.483 ls_scf_store_result 1 5.0 0.000 0.000 3.197 3.230 make_images_sizes 222 9.7 0.000 0.000 0.838 3.203 mp_alltoall_i44 222 10.7 0.837 3.203 0.837 3.203 qs_energies_init_hamiltonians 1 3.0 0.001 0.002 2.293 2.293 dbcsr_finalize 304 7.8 0.062 0.077 2.205 2.275 dbcsr_data_new 4608 9.7 1.815 2.238 1.815 2.238 dbcsr_merge_all 275 8.9 0.483 0.534 2.062 2.107 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.053 2.054 rebuild_ks_matrix 3 7.3 0.000 0.000 2.020 2.021 qs_ks_build_kohn_sham_matrix 3 8.3 0.016 0.024 2.020 2.021 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=92.868000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3771.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/908860eada1ff98329e2139c3f6fcecbf4b94d08_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 9.042395E+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 1033 319197. 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 0.132 0.165 106.000 106.001 qs_energies 1 2.0 0.000 0.000 104.173 104.175 ls_scf 1 3.0 0.000 0.000 101.148 101.151 dbcsr_multiply_generic 111 6.7 0.023 0.027 75.114 75.258 ls_scf_main 1 4.0 0.000 0.000 63.606 63.607 density_matrix_trs4 2 5.0 0.002 0.003 54.832 54.886 multiply_cannon 111 7.7 0.156 0.268 48.709 51.195 multiply_cannon_loop 111 8.7 0.097 0.099 45.565 46.114 ls_scf_init_scf 1 4.0 0.000 0.000 33.741 33.743 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.335 32.358 matrix_sqrt_Newton_Schulz 2 6.5 0.013 0.026 29.386 29.399 mp_waitall_1 4527 11.1 21.904 25.850 21.904 25.850 make_m2s 222 7.7 0.005 0.005 22.769 23.996 make_images 222 8.7 3.572 3.878 22.661 23.886 multiply_cannon_multrec 444 9.7 17.815 18.516 22.629 23.254 hybrid_alltoall_any 227 10.6 1.654 3.617 12.883 15.886 make_images_data 222 9.7 0.003 0.004 13.125 15.419 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.154 10.584 multiply_cannon_sync_h2d 444 9.7 8.788 8.833 8.788 8.833 arnoldi_extremal 4 6.8 0.000 0.000 7.380 7.405 arnoldi_normal_ev 4 7.8 0.003 0.009 7.379 7.405 build_subspace 16 8.4 0.026 0.036 6.813 6.825 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.451 5.613 dbcsr_matrix_vector_mult_local 304 10.0 5.042 5.351 5.045 5.354 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.083 5.339 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.916 5.009 dbcsr_mm_accdrv_process 1814 10.4 0.355 0.602 4.626 4.916 dbcsr_mm_accdrv_process_sort 1814 11.4 4.188 4.313 4.188 4.313 mp_allgather_i34 111 8.7 1.244 3.998 1.244 3.998 ls_scf_post 1 4.0 0.000 0.000 3.801 3.803 make_images_sizes 222 9.7 0.000 0.000 1.507 3.705 mp_alltoall_i44 222 10.7 1.507 3.705 1.507 3.705 ls_scf_store_result 1 5.0 0.000 0.000 3.497 3.534 calculate_norms 792 9.8 3.233 3.270 3.233 3.270 dbcsr_finalize 304 7.8 0.082 0.089 3.084 3.174 qs_energies_init_hamiltonians 1 3.0 0.011 0.015 2.995 3.000 dbcsr_merge_all 275 8.9 0.887 0.917 2.868 2.952 dbcsr_complete_redistribute 5 7.6 1.428 1.474 2.744 2.855 matrix_ls_to_qs 2 6.0 0.000 0.000 2.399 2.532 dbcsr_sort_data 325 11.1 2.442 2.506 2.442 2.506 dbcsr_data_new 6591 9.6 1.864 2.319 1.864 2.319 dbcsr_new_transposed 4 7.5 0.243 0.254 2.297 2.317 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.276 2.278 dbcsr_frobenius_norm 74 6.6 2.057 2.139 2.208 2.245 rebuild_ks_matrix 3 7.3 0.000 0.000 2.210 2.212 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.002 2.210 2.212 dbcsr_add_d 103 6.2 0.000 0.000 2.128 2.208 dbcsr_add_anytype 103 7.2 0.859 0.891 2.127 2.207 dbcsr_data_release 12724 10.6 1.979 2.177 1.979 2.177 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=106.001000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=7247.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/908860eada1ff98329e2139c3f6fcecbf4b94d08_performance_tests/27/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 7009386627072 0.0% 0.0% 100.0% flops 9 x 9 x 32 7335108845568 0.0% 0.0% 100.0% flops 9 x 22 x 32 9866241589248 0.0% 0.0% 100.0% flops 22 x 9 x 32 9884108906496 0.0% 0.0% 100.0% flops 22 x 22 x 32 13354440523776 0.0% 0.0% 100.0% flops 32 x 32 x 9 20607185977344 0.0% 0.0% 100.0% flops 32 x 32 x 22 25186560638976 0.0% 0.0% 100.0% flops 9 x 32 x 32 28458319085568 0.0% 0.0% 100.0% flops 22 x 32 x 32 34782389993472 0.0% 0.0% 100.0% flops 9 x 32 x 9 42881542373376 0.0% 0.0% 100.0% flops 22 x 32 x 9 55680402235392 0.0% 0.0% 100.0% flops 9 x 32 x 22 55680402235392 0.0% 0.0% 100.0% flops 22 x 32 x 22 72328573419520 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 383.054662E+12 0.0% 0.0% 100.0% flops max/rank 733.641090E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 26899403712 0.0% 0.0% 100.0% number of processed stacks 118860288 0.0% 0.0% 100.0% average stack size 0.0 0.0 226.3 marketing flops 780.439111E+12 ------------------------------------------------------------------------------- # multiplications 1445 max memory usage/rank 585.109504E+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 77937. MP_Allreduce 13232 2081. MP_Sync 1064 MP_Alltoall 2588 993455088. 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 0.038 0.108 233.336 233.339 qs_mol_dyn_low 1 2.0 0.008 0.057 231.591 231.976 qs_forces 5 3.8 0.005 0.033 230.819 230.829 qs_energies 5 4.8 0.003 0.126 227.512 227.573 scf_env_do_scf 5 5.8 0.000 0.009 213.060 213.062 scf_env_do_scf_inner_loop 105 6.6 0.005 0.028 187.421 187.422 qs_scf_new_mos 105 7.6 0.000 0.001 146.241 146.420 qs_scf_loop_do_ot 105 8.6 0.001 0.001 146.241 146.419 ot_scf_mini 105 9.6 0.003 0.013 136.325 136.500 dbcsr_multiply_generic 1445 12.2 0.129 0.137 134.980 135.385 multiply_cannon 1445 13.2 0.272 0.285 115.375 117.600 multiply_cannon_loop 1445 14.2 2.868 3.019 113.576 114.892 velocity_verlet 4 3.0 0.014 0.076 106.527 106.531 ot_mini 105 10.6 0.004 0.219 60.487 60.675 qs_ot_get_p 112 10.4 0.001 0.001 45.591 45.877 multiply_cannon_multrec 69360 15.2 29.810 34.780 40.707 45.804 mp_waitall_1 488190 16.1 35.210 42.116 35.210 42.116 qs_ot_get_derivative 55 11.6 0.001 0.001 38.641 38.784 qs_ot_p2m_diag 40 11.0 0.020 0.030 34.401 34.481 multiply_cannon_sync_h2d 69360 15.2 28.810 33.171 28.810 33.171 multiply_cannon_metrocomm3 69360 15.2 0.202 0.211 25.905 32.990 cp_dbcsr_syevd 40 12.0 0.002 0.002 30.914 30.915 rebuild_ks_matrix 110 8.4 0.000 0.001 29.997 30.183 qs_ks_build_kohn_sham_matrix 110 9.4 0.013 0.069 29.996 30.182 qs_ks_update_qs_env 112 7.6 0.001 0.001 27.433 27.607 cp_fm_syevd 40 13.0 0.000 0.000 25.534 25.681 init_scf_loop 7 6.6 0.001 0.047 25.576 25.579 apply_preconditioner_dbcsr 62 12.6 0.000 0.001 23.082 23.406 apply_single 62 13.6 0.000 0.000 23.082 23.406 ot_new_cg_direction 55 11.6 0.001 0.025 21.159 21.160 cp_fm_redistribute_end 40 14.0 10.394 20.705 10.409 20.711 cp_fm_syevd_base 40 14.0 10.290 20.602 10.290 20.602 prepare_preconditioner 7 7.6 0.000 0.000 20.485 20.517 make_preconditioner 7 8.6 0.001 0.078 20.485 20.517 qs_rho_update_rho_low 110 7.6 0.001 0.006 17.697 18.064 calculate_rho_elec 110 8.6 0.046 0.317 17.697 18.063 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.805 14.984 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 13.840 13.931 mp_sum_l 4764 12.2 12.373 13.091 12.373 13.091 make_full_inverse_cholesky 7 9.6 0.000 0.000 12.562 12.635 fft_wrap_pw1pw2 1425 12.5 0.019 0.025 12.211 12.485 calculate_dm_sparse 110 9.5 0.000 0.001 12.293 12.471 density_rs2pw 110 9.6 0.004 0.005 11.630 12.065 init_scf_run 5 5.8 0.000 0.002 11.998 11.999 scf_env_initial_rho_setup 5 6.8 0.006 0.277 11.998 11.999 dbcsr_mm_accdrv_process 154766 15.8 6.291 6.477 10.764 11.770 fft_wrap_pw1pw2_240 915 14.0 0.197 0.226 10.791 10.990 qs_vxc_create 110 10.4 0.002 0.005 10.806 10.837 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 10.478 10.568 fft3d_pb 915 15.0 2.236 2.433 9.778 9.953 check_diag 80 13.5 8.650 8.925 9.663 9.810 sum_up_and_integrate 60 10.3 0.001 0.003 8.808 8.833 integrate_v_rspace 60 11.3 0.002 0.002 8.791 8.817 transfer_rs2pw 445 10.6 0.008 0.009 8.253 8.738 cp_fm_cholesky_invert 7 10.6 8.342 8.361 8.342 8.361 acc_transpose_blocks 69360 15.2 0.348 0.362 7.621 7.960 xc_rho_set_and_dset_create 110 12.4 0.073 0.157 7.460 7.765 make_full_single_inverse 7 9.6 0.001 0.038 7.639 7.672 make_m2s 2890 13.2 0.078 0.089 6.836 7.397 xc_vxc_pw_create 60 11.3 0.036 0.047 7.357 7.391 make_images 2890 14.2 0.239 0.260 6.728 7.288 calculate_first_density_matrix 1 7.0 0.000 0.003 7.244 7.260 mp_alltoall_z22v 2340 16.7 7.008 7.213 7.008 7.213 multiply_cannon_metrocomm1 69360 15.2 0.098 0.104 4.563 7.026 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.002 6.942 6.958 xc_pw_derive 510 13.4 0.005 0.006 6.570 6.651 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 6.510 6.565 mp_waitany 7680 13.5 4.587 5.211 4.587 5.211 acc_transpose_blocks_kernels 69360 16.2 0.852 0.887 4.869 5.143 multiply_cannon_metrocomm4 67915 15.2 0.177 0.195 2.023 5.120 potential_pw2rs 60 12.3 0.002 0.003 4.917 4.958 mp_alltoall_d11v 1300 13.8 3.958 4.709 3.958 4.709 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=233.339000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=557.600000, yerr=1.356466 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 908860eada1ff98329e2139c3f6fcecbf4b94d08 Summary: empty Status: OK