=== 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: 5db0f86a89f1e2e120bbf3166966efaf01984442 ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, # Cray-FFTW 3.3.8.10, COSMA 2.6.6, ELPA 2023.05.001, # HDF5 1.14.2, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.2.2, LIBVORI 220621, LIBXSMM 1.17, # PLUMED 2.9.0, SIRIUS 7.4.3, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Last update: 04.10.2023 # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2023.05.001 USE_HDF5 := 1.14.2 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.9.0 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a endif ifneq ($(USE_HDF5),) USE_HDF5 := $(strip $(USE_HDF5)) HDF5_INC := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/include HDF5_LIB := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/lib CFLAGS += -I$(HDF5_INC) DFLAGS += -D__HDF5 LIBS += $(HDF5_LIB)/libhdf5_fortran.a LIBS += $(HDF5_LIB)/libhdf5_hl.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5db0f86a89f1e2e120bbf3166966efaf01984442_performance_tests/01 job id: 49400935 --- 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/5db0f86a89f1e2e120bbf3166966efaf01984442_performance_tests/02 job id: 49400938 --- Point --- name: 20 plot: h2o_32_ri_rpa_mp2 regex: Total MP2 Time= label: RI-MP2 (8n/6r/2t) --- Point --- name: 21 plot: h2o_32_ri_rpa_mp2_mem regex: Estimated peak process memory label: RI-MP2 (8n/6r/2t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5db0f86a89f1e2e120bbf3166966efaf01984442_performance_tests/03 job id: 49400939 --- Point --- name: 100 plot: h2o_64_md regex: CP2K label: (8n/12r/1t) --- Point --- name: 101 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/12r/1t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 6 # threads/rank = 2 nrepeat = 1 time[min] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5db0f86a89f1e2e120bbf3166966efaf01984442_performance_tests/04 job id: 49400941 --- Point --- name: 102 plot: h2o_64_md regex: CP2K label: (8n/6r/2t) --- Point --- name: 103 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/6r/2t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 4 # threads/rank = 3 nrepeat = 1 time[min] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5db0f86a89f1e2e120bbf3166966efaf01984442_performance_tests/05 job id: 49400943 --- Point --- name: 104 plot: h2o_64_md regex: CP2K label: (8n/4r/3t) --- Point --- name: 105 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/4r/3t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 3 # threads/rank = 4 nrepeat = 1 time[min] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5db0f86a89f1e2e120bbf3166966efaf01984442_performance_tests/06 job id: 49400945 --- Point --- name: 106 plot: h2o_64_md regex: CP2K label: (8n/3r/4t) --- Point --- name: 107 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/3r/4t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5db0f86a89f1e2e120bbf3166966efaf01984442_performance_tests/07 job id: 49400946 --- Point --- name: 108 plot: h2o_64_md regex: CP2K label: (8n/2r/6t) --- Point --- name: 109 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/2r/6t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 1 # threads/rank = 12 nrepeat = 1 time[min] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5db0f86a89f1e2e120bbf3166966efaf01984442_performance_tests/08 job id: 49400947 --- 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/5db0f86a89f1e2e120bbf3166966efaf01984442_performance_tests/09 job id: 49400948 --- 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/5db0f86a89f1e2e120bbf3166966efaf01984442_performance_tests/10 job id: 49400949 --- 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/5db0f86a89f1e2e120bbf3166966efaf01984442_performance_tests/11 job id: 49400950 --- 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/5db0f86a89f1e2e120bbf3166966efaf01984442_performance_tests/12 job id: 49400951 --- 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/5db0f86a89f1e2e120bbf3166966efaf01984442_performance_tests/13 job id: 49400952 --- 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/5db0f86a89f1e2e120bbf3166966efaf01984442_performance_tests/14 job id: 49400953 --- 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/5db0f86a89f1e2e120bbf3166966efaf01984442_performance_tests/15 job id: 49400954 --- 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/5db0f86a89f1e2e120bbf3166966efaf01984442_performance_tests/16 job id: 49400955 --- 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/5db0f86a89f1e2e120bbf3166966efaf01984442_performance_tests/17 job id: 49400956 --- 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/5db0f86a89f1e2e120bbf3166966efaf01984442_performance_tests/18 job id: 49400957 --- 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/5db0f86a89f1e2e120bbf3166966efaf01984442_performance_tests/19 job id: 49400958 --- 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/5db0f86a89f1e2e120bbf3166966efaf01984442_performance_tests/20 job id: 49400959 --- Point --- name: 410 plot: h2o_256_md regex: CP2K label: (8n/1r/12t) --- Point --- name: 411 plot: h2o_256_md_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5db0f86a89f1e2e120bbf3166966efaf01984442_performance_tests/21 job id: 49400960 --- Point --- name: 500 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/12r/1t) --- Point --- name: 501 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/12r/1t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 6 # threads/rank = 2 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5db0f86a89f1e2e120bbf3166966efaf01984442_performance_tests/22 job id: 49400961 --- Point --- name: 502 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/6r/2t) --- Point --- name: 503 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/6r/2t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 4 # threads/rank = 3 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5db0f86a89f1e2e120bbf3166966efaf01984442_performance_tests/23 job id: 49400962 --- Point --- name: 504 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/4r/3t) --- Point --- name: 505 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/4r/3t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 3 # threads/rank = 4 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5db0f86a89f1e2e120bbf3166966efaf01984442_performance_tests/24 job id: 49400963 --- Point --- name: 506 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/3r/4t) --- Point --- name: 507 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/3r/4t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5db0f86a89f1e2e120bbf3166966efaf01984442_performance_tests/25 job id: 49400964 --- Point --- name: 508 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/2r/6t) --- Point --- name: 509 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/2r/6t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 1 # threads/rank = 12 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5db0f86a89f1e2e120bbf3166966efaf01984442_performance_tests/26 job id: 49400966 --- Point --- name: 510 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/1r/12t) --- Point --- name: 511 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: 512 H2O (4 NVE MD steps on 64 nodes) input file: benchmarks/QS/00512_H2O/H2O-512_md.inp required files: [] output file: result.log # nodes = 64 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5db0f86a89f1e2e120bbf3166966efaf01984442_performance_tests/27 job id: 49400967 --- 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/5db0f86a89f1e2e120bbf3166966efaf01984442_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 424 8. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.032 0.053 133.661 133.661 farming_run 1 2.0 132.930 132.932 133.593 133.595 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.490006E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 230 1134128. MP_Allreduce 571 1938539. MP_Sync 25 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 10 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.036 114.811 114.811 qs_energies 1 2.0 0.000 0.000 114.593 114.595 mp2_main 1 3.0 0.000 0.000 112.502 112.504 mp2_gpw_main 1 4.0 0.031 0.036 111.480 111.482 mp2_ri_gpw_compute_in 1 5.0 0.172 0.173 92.594 92.940 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 54.662 55.010 mp2_eri_3c_integrate_gpw 272 7.0 0.152 0.164 41.059 46.149 get_2c_integrals 1 6.0 0.008 0.009 37.050 37.759 integrate_v_rspace 273 8.0 0.437 0.453 24.650 29.526 pw_transfer 6555 10.6 0.376 0.394 26.840 27.575 fft_wrap_pw1pw2 5465 11.4 0.045 0.049 25.454 26.171 grid_integrate_task_list 273 9.0 20.559 25.904 20.559 25.904 fft_wrap_pw1pw2_100 2178 12.4 1.145 1.421 23.013 23.685 compute_2c_integrals 1 7.0 0.002 0.003 19.352 19.352 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.640 19.083 mp2_eri_2c_integrate_gpw 1 9.0 2.365 2.430 18.638 19.082 rpa_ri_compute_en 1 5.0 0.019 0.020 18.766 19.036 cp_fm_cholesky_decompose 12 8.2 17.676 18.401 17.676 18.401 cholesky_decomp 1 7.0 0.000 0.000 16.540 17.268 fft3d_s 5443 13.4 16.126 16.378 16.149 16.399 ao_to_mo_and_store_B_mult_1 272 7.0 10.771 15.331 10.771 15.331 calculate_wavefunction 272 8.0 5.401 5.550 12.322 12.983 rpa_num_int 1 6.0 0.000 0.002 10.696 10.697 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.678 10.694 calc_mat_Q 8 8.0 0.000 0.000 9.496 9.586 contract_S_to_Q 8 9.0 0.000 0.000 8.917 9.009 calc_potential_gpw 544 9.5 0.005 0.006 8.253 8.719 parallel_gemm_fm 14 9.1 0.000 0.000 8.500 8.608 parallel_gemm_fm_cosma 14 10.1 8.500 8.608 8.500 8.608 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.110 8.411 potential_pw2rs 545 10.0 0.107 0.109 7.493 8.231 create_integ_mat 1 6.0 0.022 0.027 7.631 7.632 collocate_single_gaussian 272 10.0 0.039 0.042 7.355 7.591 array2fm 1 7.0 0.000 0.000 6.743 7.078 pw_scatter_s 2720 13.7 4.341 4.481 4.341 4.481 pw_gather_s 2722 13.2 3.443 3.818 3.443 3.818 array2fm_buffer_send 1 8.0 3.027 3.182 3.027 3.182 pw_poisson_solve 545 10.5 1.116 1.167 2.209 2.455 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.480269, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2807.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5db0f86a89f1e2e120bbf3166966efaf01984442_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 205321. MP_Allreduce 424 9. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.028 0.044 421.484 421.485 farming_run 1 2.0 420.493 420.502 421.431 421.443 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.234412E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 757 478553. MP_Allreduce 2021 21391. MP_Sync 37 MP_Alltoall 77 27544160. MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.033 217.141 217.141 qs_energies 1 2.0 0.000 0.000 216.883 216.890 scf_env_do_scf 1 3.0 0.000 0.000 115.160 115.160 qs_ks_update_qs_env 5 5.0 0.000 0.000 114.233 114.240 rebuild_ks_matrix 4 6.0 0.000 0.000 114.232 114.238 qs_ks_build_kohn_sham_matrix 4 7.0 0.055 0.061 114.232 114.238 hfx_ks_matrix 4 8.0 0.001 0.001 113.869 113.874 integrate_four_center 4 9.0 0.152 0.461 113.869 113.873 integrate_four_center_main 4 10.0 0.132 0.489 102.072 105.527 integrate_four_center_bin 267 11.0 101.940 105.465 101.940 105.465 mp2_main 1 3.0 0.000 0.000 101.422 101.429 mp2_gpw_main 1 4.0 0.048 0.072 100.536 100.545 init_scf_loop 1 4.0 0.000 0.000 97.150 97.150 mp2_ri_gpw_compute_in 1 5.0 0.064 0.065 73.811 74.848 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 53.743 54.780 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.164 41.428 46.539 integrate_v_rspace 95 8.0 0.394 0.561 27.917 32.893 pw_transfer 2240 10.6 0.144 0.164 29.484 29.814 fft_wrap_pw1pw2 1868 11.4 0.018 0.021 28.500 28.811 mp2_ri_gpw_compute_en 1 5.0 0.055 0.063 26.565 28.631 grid_integrate_task_list 95 9.0 23.324 28.494 23.324 28.494 ao_to_mo_and_store_B_mult_1 91 7.0 10.641 28.094 10.641 28.094 fft_wrap_pw1pw2_100 730 12.4 1.265 1.401 26.250 26.514 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.844 1.891 24.881 24.891 get_2c_integrals 1 6.0 0.000 0.000 19.948 20.005 compute_2c_integrals 1 7.0 0.003 0.005 18.935 18.938 fft3d_s 1823 13.4 18.491 18.858 18.505 18.873 compute_2c_integrals_loop_lm 1 8.0 0.001 0.002 18.597 18.804 mp2_eri_2c_integrate_gpw 1 9.0 1.744 1.834 18.596 18.804 scf_env_do_scf_inner_loop 4 4.0 0.000 0.001 18.008 18.008 calculate_wavefunction 91 8.0 2.017 2.049 9.622 9.847 mp2_ri_gpw_compute_en_expansio 172 7.0 0.557 0.585 8.763 9.245 potential_pw2rs 186 10.0 0.033 0.034 8.410 9.078 local_gemm 172 8.0 8.206 8.671 8.206 8.671 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.127 8.442 mp2_ri_gpw_compute_en_comm 22 7.0 0.498 0.519 7.888 8.346 calc_potential_gpw 182 9.5 0.002 0.002 7.897 8.047 collocate_single_gaussian 91 10.0 0.017 0.020 7.737 7.937 integrate_four_center_load 4 10.0 0.000 0.000 6.782 6.786 hfx_load_balance 1 11.0 0.000 0.000 6.782 6.786 mp_sync 37 10.5 3.778 6.672 3.778 6.672 mp_sendrecv_dm3 2068 8.0 5.926 6.402 5.926 6.402 mp2_ri_gpw_compute_en_ener 172 7.0 6.341 6.393 6.341 6.393 pw_gather_s 912 13.2 4.486 4.980 4.486 4.980 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=100.533414, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1492.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5db0f86a89f1e2e120bbf3166966efaf01984442_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 453.066752E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62379. MP_Allreduce 10329 270. MP_Sync 530 MP_Alltoall 2083 592243. MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.033 55.986 55.987 qs_mol_dyn_low 1 2.0 0.003 0.004 55.524 55.532 qs_forces 11 3.9 0.003 0.007 55.451 55.452 qs_energies 11 4.9 0.014 0.087 53.896 53.909 scf_env_do_scf 11 5.9 0.000 0.001 47.505 47.506 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 45.445 45.446 qs_scf_new_mos 108 7.5 0.000 0.000 35.312 35.603 qs_scf_loop_do_ot 108 8.5 0.001 0.001 35.311 35.603 dbcsr_multiply_generic 2286 12.5 0.093 0.096 34.966 35.554 ot_scf_mini 108 9.5 0.002 0.003 33.593 33.796 multiply_cannon 2286 13.5 0.195 0.208 27.135 29.282 multiply_cannon_loop 2286 14.5 1.810 1.923 26.480 28.659 velocity_verlet 10 3.0 0.001 0.001 27.616 27.616 ot_mini 108 10.5 0.001 0.001 20.582 20.834 qs_ot_get_derivative 108 11.5 0.001 0.001 17.407 17.614 mp_waitall_1 245248 16.5 9.185 15.992 9.185 15.992 multiply_cannon_metrocomm3 54864 15.5 0.073 0.080 6.565 14.307 multiply_cannon_multrec 54864 15.5 3.640 5.665 7.816 11.271 qs_ot_get_p 119 10.4 0.001 0.001 8.200 8.475 rebuild_ks_matrix 119 8.3 0.000 0.000 7.790 7.953 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 7.790 7.953 mp_sum_l 7287 12.8 5.922 7.420 5.922 7.420 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.864 7.010 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 6.401 6.885 multiply_cannon_sync_h2d 54864 15.5 5.113 6.097 5.113 6.097 dbcsr_mm_accdrv_process 76910 16.1 1.812 2.891 4.090 5.856 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.253 5.368 qs_ot_p2m_diag 50 11.0 0.004 0.007 5.213 5.248 init_scf_run 11 5.9 0.000 0.001 4.871 4.871 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.870 4.871 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.255 4.455 calculate_rho_elec 119 8.7 0.011 0.016 4.255 4.455 cp_dbcsr_syevd 50 12.0 0.002 0.003 4.416 4.417 sum_up_and_integrate 119 10.3 0.001 0.002 4.400 4.407 integrate_v_rspace 119 11.3 0.002 0.004 4.389 4.398 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.220 4.221 cp_fm_redistribute_end 50 14.0 2.154 4.197 2.160 4.200 cp_fm_diag_elpa_base 50 14.0 2.035 4.092 2.039 4.101 multiply_cannon_metrocomm1 54864 15.5 0.057 0.062 1.989 3.555 calculate_dm_sparse 119 9.5 0.000 0.001 3.053 3.185 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.953 3.152 apply_single 119 13.6 0.000 0.000 2.953 3.151 jit_kernel_multiply 13 15.8 2.215 2.876 2.215 2.876 acc_transpose_blocks 54864 15.5 0.231 0.260 2.255 2.821 ot_diis_step 108 11.5 0.006 0.006 2.809 2.810 calculate_first_density_matrix 1 7.0 0.000 0.002 2.700 2.704 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.586 2.647 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.353 2.355 mp_sum_d 4135 12.0 1.620 2.208 1.620 2.208 density_rs2pw 119 9.7 0.004 0.004 2.067 2.165 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.079 2.132 grid_integrate_task_list 119 12.3 2.010 2.112 2.010 2.112 wfi_extrapolate 11 7.9 0.001 0.001 2.094 2.094 init_scf_loop 11 6.9 0.001 0.005 2.044 2.045 potential_pw2rs 119 12.3 0.004 0.004 1.782 1.795 mp_alltoall_d11v 2130 13.8 1.507 1.742 1.507 1.742 pw_transfer 1439 11.6 0.051 0.055 1.606 1.674 make_m2s 4572 13.5 0.053 0.055 1.596 1.646 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.531 1.600 make_images 4572 14.5 0.134 0.139 1.514 1.564 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.493 1.514 acc_transpose_blocks_sync 164592 16.5 1.198 1.447 1.198 1.447 mp_waitany 12084 13.8 1.270 1.437 1.270 1.437 transfer_rs2pw 487 10.6 0.005 0.006 1.341 1.432 grid_collocate_task_list 119 9.7 1.346 1.408 1.346 1.408 dbcsr_dot_sd 1205 11.9 0.048 0.060 0.917 1.325 transfer_pw2rs 487 13.2 0.005 0.006 1.307 1.314 fft3d_ps 1201 14.6 0.367 0.473 1.246 1.309 fft_wrap_pw1pw2_140 487 13.2 0.141 0.154 1.189 1.260 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=55.987000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=431.909091, yerr=1.504813 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5db0f86a89f1e2e120bbf3166966efaf01984442_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 487.428096E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2194560 MPI messages size (bytes): total size 310.646604E+09 min size 0.000000E+00 max size 1.145520E+06 average size 141.553031E+03 MPI breakdown and total messages size (bytes): size <= 128 724648 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 281952 4619501568 32768 < size <= 131072 494448 39143342080 131072 < size <= 4194304 440000 264807943488 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62658. MP_Allreduce 10306 303. MP_Sync 54 MP_Alltoall 2060 1472857. 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 0.011 0.026 39.676 39.678 qs_mol_dyn_low 1 2.0 0.003 0.006 39.434 39.441 qs_forces 11 3.9 0.002 0.003 39.373 39.374 qs_energies 11 4.9 0.002 0.003 37.643 37.648 scf_env_do_scf 11 5.9 0.000 0.001 32.526 32.527 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 30.023 30.023 dbcsr_multiply_generic 2286 12.5 0.100 0.102 22.362 22.724 qs_scf_new_mos 108 7.5 0.001 0.001 21.132 21.360 qs_scf_loop_do_ot 108 8.5 0.001 0.001 21.131 21.360 ot_scf_mini 108 9.5 0.002 0.003 20.214 20.379 velocity_verlet 10 3.0 0.001 0.002 18.921 18.922 multiply_cannon 2286 13.5 0.211 0.218 17.112 18.608 multiply_cannon_loop 2286 14.5 1.189 1.251 15.912 17.398 ot_mini 108 10.5 0.001 0.001 12.519 12.750 mp_waitall_1 200699 16.5 5.855 11.070 5.855 11.070 qs_ot_get_derivative 108 11.5 0.001 0.001 10.141 10.308 multiply_cannon_metrocomm3 27432 15.5 0.071 0.074 4.169 9.702 multiply_cannon_multrec 27432 15.5 1.848 4.204 6.171 8.910 rebuild_ks_matrix 119 8.3 0.000 0.000 6.932 7.072 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.931 7.071 dbcsr_mm_accdrv_process 47894 16.0 3.498 5.686 4.240 6.342 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.082 6.209 qs_ot_get_p 119 10.4 0.001 0.001 4.752 4.974 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.833 4.677 mp_sum_l 7287 12.8 2.118 4.195 2.118 4.195 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.007 4.075 apply_single 119 13.6 0.000 0.000 3.007 4.074 init_scf_run 11 5.9 0.000 0.001 3.874 3.875 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.874 3.874 sum_up_and_integrate 119 10.3 0.001 0.001 3.841 3.847 integrate_v_rspace 119 11.3 0.002 0.002 3.827 3.833 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.772 3.799 calculate_rho_elec 119 8.7 0.021 0.024 3.772 3.798 qs_ot_p2m_diag 50 11.0 0.009 0.013 3.064 3.084 make_m2s 4572 13.5 0.052 0.053 2.697 3.048 make_images 4572 14.5 0.207 0.246 2.609 2.962 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.660 2.660 init_scf_loop 11 6.9 0.000 0.000 2.475 2.475 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.351 2.438 ot_diis_step 108 11.5 0.011 0.011 2.328 2.329 calculate_first_density_matrix 1 7.0 0.000 0.001 2.325 2.326 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.312 2.312 cp_fm_redistribute_end 50 14.0 1.173 2.285 1.177 2.287 cp_fm_diag_elpa_base 50 14.0 1.078 2.189 1.106 2.224 multiply_cannon_sync_h2d 27432 15.5 1.690 2.215 1.690 2.215 density_rs2pw 119 9.7 0.004 0.004 2.065 2.168 calculate_dm_sparse 119 9.5 0.000 0.001 2.060 2.141 pw_transfer 1439 11.6 0.065 0.068 1.929 1.971 acc_transpose_blocks 27432 15.5 0.115 0.121 1.613 1.949 grid_integrate_task_list 119 12.3 1.833 1.945 1.833 1.945 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.887 1.889 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.838 1.882 jit_kernel_multiply 9 16.3 0.684 1.833 0.684 1.833 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.656 1.695 prepare_preconditioner 11 7.9 0.000 0.000 1.574 1.601 make_preconditioner 11 8.9 0.000 0.000 1.574 1.601 potential_pw2rs 119 12.3 0.006 0.007 1.584 1.600 make_images_data 4572 15.5 0.047 0.053 1.257 1.596 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.474 1.530 fft3d_ps 1201 14.6 0.520 0.576 1.471 1.509 hybrid_alltoall_any 4725 16.4 0.053 0.115 1.126 1.497 wfi_extrapolate 11 7.9 0.001 0.001 1.489 1.489 mp_alltoall_d11v 2130 13.8 1.281 1.476 1.281 1.476 fft_wrap_pw1pw2_140 487 13.2 0.162 0.175 1.423 1.465 grid_collocate_task_list 119 9.7 1.288 1.343 1.288 1.343 transfer_rs2pw 487 10.6 0.005 0.005 1.194 1.292 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.266 1.274 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.219 1.266 mp_allgather_i34 2286 14.5 0.618 1.032 0.618 1.032 mp_sum_d 4135 12.0 0.576 1.030 0.576 1.030 transfer_pw2rs 487 13.2 0.004 0.005 1.017 1.030 qs_energies_init_hamiltonians 11 5.9 0.000 0.002 0.943 0.944 acc_transpose_blocks_sync 82296 16.5 0.814 0.941 0.814 0.941 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.907 0.921 acc_transpose_blocks_kernels 27432 16.5 0.189 0.278 0.658 0.880 make_images_sizes 4572 15.5 0.005 0.005 0.571 0.809 mp_alltoall_i44 4572 16.5 0.566 0.804 0.566 0.804 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=39.678000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=464.454545, yerr=0.987525 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5db0f86a89f1e2e120bbf3166966efaf01984442_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 520.065024E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63490. MP_Allreduce 10155 305. MP_Sync 54 MP_Alltoall 1821 1607811. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.009 0.027 31.986 31.988 qs_mol_dyn_low 1 2.0 0.003 0.003 31.773 31.781 qs_forces 11 3.9 0.002 0.003 31.717 31.717 qs_energies 11 4.9 0.002 0.002 30.177 30.179 scf_env_do_scf 11 5.9 0.000 0.001 25.574 25.574 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 23.143 23.144 dbcsr_multiply_generic 2286 12.5 0.094 0.098 16.874 16.964 qs_scf_new_mos 108 7.5 0.001 0.001 15.483 15.502 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.483 15.501 velocity_verlet 10 3.0 0.004 0.007 15.168 15.169 ot_scf_mini 108 9.5 0.002 0.002 14.721 14.740 multiply_cannon 2286 13.5 0.193 0.196 13.564 14.278 multiply_cannon_loop 2286 14.5 0.858 0.918 12.818 13.560 ot_mini 108 10.5 0.001 0.001 9.087 9.112 qs_ot_get_derivative 108 11.5 0.001 0.001 7.593 7.611 multiply_cannon_multrec 18288 15.5 1.902 2.967 6.961 7.324 rebuild_ks_matrix 119 8.3 0.000 0.000 5.953 5.971 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.953 5.971 dbcsr_mm_accdrv_process 38222 16.0 4.902 5.679 4.966 5.744 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.246 5.262 sum_up_and_integrate 119 10.3 0.001 0.001 3.553 3.557 integrate_v_rspace 119 11.3 0.002 0.003 3.540 3.545 init_scf_run 11 5.9 0.000 0.001 3.449 3.449 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.449 3.449 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.822 3.440 qs_ot_get_p 119 10.4 0.001 0.001 3.393 3.420 mp_waitall_1 158411 16.6 2.577 3.347 2.577 3.347 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.247 3.249 calculate_rho_elec 119 8.7 0.031 0.031 3.246 3.248 init_scf_loop 11 6.9 0.000 0.000 2.415 2.416 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.026 2.343 apply_single 119 13.6 0.000 0.000 2.025 2.343 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.265 2.271 calculate_first_density_matrix 1 7.0 0.000 0.000 2.222 2.223 multiply_cannon_metrocomm3 18288 15.5 0.047 0.049 1.418 2.080 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.962 1.963 make_m2s 4572 13.5 0.044 0.045 1.811 1.937 density_rs2pw 119 9.7 0.004 0.004 1.822 1.918 grid_integrate_task_list 119 12.3 1.805 1.900 1.805 1.900 calculate_dm_sparse 119 9.5 0.000 0.001 1.842 1.857 make_images 4572 14.5 0.193 0.204 1.726 1.851 pw_transfer 1439 11.6 0.064 0.068 1.827 1.835 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.735 1.745 acc_transpose_blocks 18288 15.5 0.081 0.082 1.628 1.741 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.724 1.733 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.710 1.722 cp_fm_diag_elpa_base 50 14.0 1.686 1.697 1.708 1.720 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.671 1.672 prepare_preconditioner 11 7.9 0.000 0.000 1.666 1.669 make_preconditioner 11 8.9 0.000 0.000 1.666 1.669 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.530 1.610 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.502 1.509 ot_diis_step 108 11.5 0.011 0.011 1.470 1.471 mp_sum_l 7287 12.8 1.112 1.440 1.112 1.440 fft_wrap_pw1pw2_140 487 13.2 0.212 0.219 1.386 1.397 potential_pw2rs 119 12.3 0.007 0.009 1.356 1.361 grid_collocate_task_list 119 9.7 1.233 1.335 1.233 1.335 fft3d_ps 1201 14.6 0.527 0.545 1.304 1.314 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.177 1.183 wfi_extrapolate 11 7.9 0.001 0.001 1.173 1.173 multiply_cannon_sync_h2d 18288 15.5 1.028 1.164 1.028 1.164 transfer_rs2pw 487 10.6 0.005 0.005 0.980 1.075 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.959 0.975 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.937 0.937 make_images_data 4572 15.5 0.047 0.052 0.774 0.925 acc_transpose_blocks_sync 54864 16.5 0.744 0.854 0.744 0.854 hybrid_alltoall_any 4725 16.4 0.058 0.115 0.668 0.845 transfer_pw2rs 487 13.2 0.004 0.004 0.823 0.827 acc_transpose_blocks_kernels 18288 16.5 0.218 0.228 0.784 0.801 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.777 0.779 mp_alltoall_d11v 2130 13.8 0.654 0.751 0.654 0.751 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.679 0.731 cp_fm_cholesky_invert 11 10.9 0.723 0.727 0.723 0.727 mp_alltoall_z22v 1201 16.6 0.624 0.694 0.624 0.694 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=31.988000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=495.000000, yerr=1.128152 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5db0f86a89f1e2e120bbf3166966efaf01984442_performance_tests/06/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 114.044384E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3805952 0.0% 0.0% 100.0% average stack size 0.0 0.0 38.6 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 558.608384E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63489. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.009 0.043 36.485 36.486 qs_mol_dyn_low 1 2.0 0.003 0.003 36.278 36.309 qs_forces 11 3.9 0.002 0.002 36.218 36.220 qs_energies 11 4.9 0.002 0.002 34.506 34.512 scf_env_do_scf 11 5.9 0.000 0.001 29.438 29.440 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 26.050 26.051 dbcsr_multiply_generic 2286 12.5 0.106 0.118 19.909 20.007 velocity_verlet 10 3.0 0.002 0.002 18.532 18.534 qs_scf_new_mos 108 7.5 0.001 0.001 17.914 17.970 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.913 17.969 ot_scf_mini 108 9.5 0.002 0.003 16.878 16.936 multiply_cannon 2286 13.5 0.219 0.226 16.170 16.470 multiply_cannon_loop 2286 14.5 1.550 1.614 15.194 15.592 ot_mini 108 10.5 0.001 0.001 10.358 10.423 multiply_cannon_multrec 27432 15.5 2.448 3.134 8.999 9.302 qs_ot_get_derivative 108 11.5 0.001 0.001 8.486 8.545 dbcsr_mm_accdrv_process 47916 15.9 6.007 7.572 6.447 7.859 rebuild_ks_matrix 119 8.3 0.000 0.000 6.341 6.386 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.341 6.386 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.613 5.654 init_scf_run 11 5.9 0.000 0.001 3.737 3.737 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.736 3.737 qs_ot_get_p 119 10.4 0.001 0.001 3.654 3.727 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.159 3.549 sum_up_and_integrate 119 10.3 0.001 0.001 3.510 3.519 integrate_v_rspace 119 11.3 0.002 0.003 3.499 3.507 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.407 3.447 calculate_rho_elec 119 8.7 0.040 0.046 3.407 3.447 init_scf_loop 11 6.9 0.000 0.000 3.369 3.369 prepare_preconditioner 11 7.9 0.000 0.000 2.540 2.549 make_preconditioner 11 8.9 0.000 0.000 2.540 2.549 acc_transpose_blocks 27432 15.5 0.121 0.125 2.433 2.526 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.144 2.477 make_m2s 4572 13.5 0.054 0.056 2.339 2.458 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.352 2.380 mp_waitall_1 137007 16.6 1.805 2.361 1.805 2.361 make_images 4572 14.5 0.274 0.337 2.231 2.348 calculate_first_density_matrix 1 7.0 0.000 0.000 2.342 2.345 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.225 2.234 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.103 2.224 apply_single 119 13.6 0.000 0.000 2.103 2.223 calculate_dm_sparse 119 9.5 0.000 0.000 2.140 2.198 pw_transfer 1439 11.6 0.064 0.068 1.939 1.975 density_rs2pw 119 9.7 0.004 0.004 1.823 1.917 grid_integrate_task_list 119 12.3 1.832 1.910 1.832 1.910 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.847 1.886 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.835 1.836 ot_diis_step 108 11.5 0.012 0.013 1.830 1.830 jit_kernel_multiply 10 16.1 0.374 1.825 0.374 1.825 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.813 1.814 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.671 1.684 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.574 1.583 cp_fm_diag_elpa_base 50 14.0 1.540 1.557 1.572 1.581 fft_wrap_pw1pw2_140 487 13.2 0.245 0.258 1.518 1.559 acc_transpose_blocks_sync 82296 16.5 1.459 1.545 1.459 1.545 fft3d_ps 1201 14.6 0.553 0.602 1.367 1.391 wfi_extrapolate 11 7.9 0.001 0.001 1.341 1.341 multiply_cannon_metrocomm3 27432 15.5 0.040 0.042 0.779 1.335 grid_collocate_task_list 119 9.7 1.250 1.333 1.250 1.333 potential_pw2rs 119 12.3 0.009 0.010 1.311 1.314 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.296 1.312 mp_sum_l 7287 12.8 0.981 1.301 0.981 1.301 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.233 1.244 cp_fm_upper_to_full 72 14.2 0.818 1.167 0.818 1.167 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.069 1.070 dbcsr_complete_redistribute 329 12.2 0.118 0.146 0.755 1.029 transfer_rs2pw 487 10.6 0.004 0.005 0.926 1.021 make_images_data 4572 15.5 0.047 0.051 0.877 1.008 hybrid_alltoall_any 4725 16.4 0.065 0.154 0.733 0.928 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.800 0.877 mp_alltoall_d11v 2130 13.8 0.764 0.868 0.764 0.868 acc_transpose_blocks_kernels 27432 16.5 0.272 0.281 0.824 0.833 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.561 0.831 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.816 0.821 transfer_pw2rs 487 13.2 0.004 0.004 0.749 0.753 cp_fm_cholesky_invert 11 10.9 0.745 0.748 0.745 0.748 mp_alltoall_i22 627 13.8 0.428 0.730 0.428 0.730 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=36.486000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=530.000000, yerr=3.464102 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5db0f86a89f1e2e120bbf3166966efaf01984442_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 613.298176E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63488. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.038 29.248 29.250 qs_mol_dyn_low 1 2.0 0.003 0.003 29.058 29.066 qs_forces 11 3.9 0.002 0.002 28.963 28.964 qs_energies 11 4.9 0.002 0.002 27.255 27.258 scf_env_do_scf 11 5.9 0.000 0.001 22.388 22.388 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 19.870 19.871 velocity_verlet 10 3.0 0.014 0.016 14.955 14.960 dbcsr_multiply_generic 2286 12.5 0.090 0.094 13.309 13.433 qs_scf_new_mos 108 7.5 0.001 0.001 12.144 12.163 qs_scf_loop_do_ot 108 8.5 0.001 0.001 12.143 12.163 ot_scf_mini 108 9.5 0.002 0.002 11.432 11.455 multiply_cannon 2286 13.5 0.222 0.226 10.640 11.025 multiply_cannon_loop 2286 14.5 0.643 0.669 9.713 9.907 ot_mini 108 10.5 0.001 0.001 6.554 6.581 multiply_cannon_multrec 9144 15.5 1.683 2.022 6.035 6.407 rebuild_ks_matrix 119 8.3 0.000 0.000 5.839 5.861 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.839 5.861 qs_ot_get_derivative 108 11.5 0.001 0.001 5.254 5.277 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.197 5.217 dbcsr_mm_accdrv_process 12550 15.8 3.668 4.244 4.235 4.324 sum_up_and_integrate 119 10.3 0.001 0.001 3.509 3.513 integrate_v_rspace 119 11.3 0.003 0.003 3.499 3.503 init_scf_run 11 5.9 0.000 0.001 3.424 3.424 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.424 3.424 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.365 3.370 calculate_rho_elec 119 8.7 0.060 0.061 3.365 3.369 qs_ot_get_p 119 10.4 0.001 0.001 2.984 3.014 init_scf_loop 11 6.9 0.000 0.000 2.498 2.499 calculate_first_density_matrix 1 7.0 0.000 0.000 2.235 2.235 mp_waitall_1 115863 16.7 1.548 2.108 1.548 2.108 make_m2s 4572 13.5 0.034 0.035 1.858 2.009 pw_transfer 1439 11.6 0.065 0.067 1.994 2.001 grid_integrate_task_list 119 12.3 1.869 1.934 1.869 1.934 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.929 1.931 make_images 4572 14.5 0.271 0.302 1.769 1.918 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.901 1.911 density_rs2pw 119 9.7 0.003 0.004 1.757 1.834 calculate_dm_sparse 119 9.5 0.000 0.000 1.790 1.811 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.763 1.782 prepare_preconditioner 11 7.9 0.000 0.000 1.765 1.769 make_preconditioner 11 8.9 0.000 0.000 1.765 1.769 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.682 1.682 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.653 1.678 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.610 1.611 fft_wrap_pw1pw2_140 487 13.2 0.323 0.332 1.556 1.567 acc_transpose_blocks 9144 15.5 0.042 0.043 1.467 1.510 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.438 1.450 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.379 1.386 cp_fm_diag_elpa_base 50 14.0 1.351 1.366 1.377 1.384 grid_collocate_task_list 119 9.7 1.298 1.382 1.298 1.382 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.366 1.377 jit_kernel_multiply 7 15.4 0.529 1.366 0.529 1.366 fft3d_ps 1201 14.6 0.557 0.567 1.315 1.326 potential_pw2rs 119 12.3 0.010 0.011 1.301 1.305 ot_diis_step 108 11.5 0.013 0.013 1.286 1.287 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.218 1.254 apply_single 119 13.6 0.000 0.000 1.218 1.254 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.225 1.225 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.176 1.181 wfi_extrapolate 11 7.9 0.001 0.001 1.139 1.139 hybrid_alltoall_any 4725 16.4 0.065 0.176 0.791 1.007 make_images_data 4572 15.5 0.042 0.045 0.800 0.969 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.873 0.925 transfer_rs2pw 487 10.6 0.004 0.004 0.830 0.924 cp_fm_cholesky_invert 11 10.9 0.898 0.900 0.898 0.900 mp_alltoall_d11v 2130 13.8 0.807 0.874 0.807 0.874 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.826 0.833 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.765 0.767 multiply_cannon_metrocomm3 9144 15.5 0.020 0.021 0.390 0.766 acc_transpose_blocks_sync 27432 16.5 0.722 0.758 0.722 0.758 qs_env_update_s_mstruct 11 6.9 0.001 0.001 0.675 0.723 transfer_pw2rs 487 13.2 0.003 0.004 0.720 0.723 acc_transpose_blocks_kernels 9144 16.5 0.118 0.120 0.688 0.692 mp_allgather_i34 2286 14.5 0.244 0.677 0.244 0.677 mp_alltoall_z22v 1201 16.6 0.631 0.664 0.631 0.664 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=29.250000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=580.090909, yerr=4.737751 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5db0f86a89f1e2e120bbf3166966efaf01984442_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 763.449344E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63723. MP_Allreduce 10154 429. MP_Sync 54 MP_Alltoall 1582 7383731. MP_SendRecv 2499 189067. MP_ISendRecv 2499 189067. MP_Wait 6399 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.036 41.876 41.877 qs_mol_dyn_low 1 2.0 0.003 0.003 41.664 41.673 qs_forces 11 3.9 0.003 0.004 41.599 41.600 qs_energies 11 4.9 0.041 0.042 39.599 39.602 scf_env_do_scf 11 5.9 0.001 0.001 33.985 33.985 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 26.263 26.264 velocity_verlet 10 3.0 0.002 0.003 23.666 23.672 dbcsr_multiply_generic 2286 12.5 0.098 0.100 18.656 18.900 qs_scf_new_mos 108 7.5 0.001 0.001 16.987 17.080 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.987 17.079 ot_scf_mini 108 9.5 0.002 0.002 15.835 15.933 multiply_cannon 2286 13.5 0.298 0.301 14.841 15.803 multiply_cannon_loop 2286 14.5 0.841 0.862 13.601 14.592 ot_mini 108 10.5 0.001 0.001 9.629 9.742 multiply_cannon_multrec 9144 15.5 3.382 4.614 8.745 8.877 qs_ot_get_derivative 108 11.5 0.001 0.001 7.603 7.700 init_scf_loop 11 6.9 0.000 0.000 7.695 7.698 rebuild_ks_matrix 119 8.3 0.000 0.000 7.008 7.148 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.008 7.148 prepare_preconditioner 11 7.9 0.000 0.000 6.756 6.769 make_preconditioner 11 8.9 0.000 0.000 6.756 6.769 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.333 6.646 dbcsr_mm_accdrv_process 12550 15.8 4.494 6.105 5.217 6.532 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.316 6.442 cp_fm_upper_to_full 72 14.2 3.207 4.571 3.207 4.571 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.967 3.971 calculate_rho_elec 119 8.7 0.118 0.121 3.967 3.970 sum_up_and_integrate 119 10.3 0.001 0.001 3.731 3.737 integrate_v_rspace 119 11.3 0.003 0.004 3.721 3.727 init_scf_run 11 5.9 0.000 0.001 3.509 3.509 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.509 3.509 qs_ot_get_p 119 10.4 0.001 0.001 3.318 3.453 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.786 3.181 mp_waitall_1 94719 16.7 2.150 3.110 2.150 3.110 dbcsr_complete_redistribute 329 12.2 0.297 0.303 1.963 2.781 pw_transfer 1439 11.6 0.068 0.069 2.551 2.557 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.175 2.477 apply_single 119 13.6 0.000 0.000 2.174 2.477 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.663 2.475 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.453 2.458 make_m2s 4572 13.5 0.037 0.038 2.297 2.437 make_images 4572 14.5 0.357 0.392 2.177 2.317 calculate_dm_sparse 119 9.5 0.000 0.000 2.230 2.246 mp_alltoall_i22 627 13.8 1.395 2.238 1.395 2.238 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.417 2.223 multiply_cannon_metrocomm3 9144 15.5 0.021 0.021 1.260 2.133 grid_integrate_task_list 119 12.3 2.076 2.104 2.076 2.104 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.034 2.082 calculate_first_density_matrix 1 7.0 0.000 0.000 2.078 2.081 fft_wrap_pw1pw2_140 487 13.2 0.575 0.579 2.073 2.080 density_rs2pw 119 9.7 0.003 0.004 2.005 2.029 ot_diis_step 108 11.5 0.014 0.014 2.001 2.001 qs_ot_p2m_diag 50 11.0 0.043 0.044 1.960 1.961 acc_transpose_blocks 9144 15.5 0.044 0.045 1.779 1.814 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.804 1.804 qs_energies_init_hamiltonians 11 5.9 0.011 0.015 1.771 1.771 mp_sum_l 7287 12.8 0.983 1.767 0.983 1.767 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.607 1.608 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.545 1.580 fft3d_ps 1201 14.6 0.591 0.603 1.550 1.553 grid_collocate_task_list 119 9.7 1.518 1.532 1.518 1.532 cp_fm_cholesky_invert 11 10.9 1.397 1.400 1.397 1.400 wfi_extrapolate 11 7.9 0.001 0.001 1.368 1.368 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.327 1.327 cp_fm_diag_elpa_base 50 14.0 1.180 1.232 1.326 1.326 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.292 1.311 potential_pw2rs 119 12.3 0.014 0.014 1.254 1.255 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.245 1.250 hybrid_alltoall_any 4725 16.4 0.090 0.152 1.025 1.231 mp_alltoall_d11v 2130 13.8 1.132 1.159 1.132 1.159 make_images_data 4572 15.5 0.046 0.049 0.957 1.134 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.089 1.104 acc_transpose_blocks_sync 27432 16.5 1.065 1.101 1.065 1.101 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 0.992 1.046 qs_create_task_list 11 7.9 0.000 0.000 0.929 0.939 generate_qs_task_list 11 8.9 0.366 0.386 0.929 0.939 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.901 0.915 mp_alltoall_z22v 1201 16.6 0.822 0.846 0.822 0.846 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=41.877000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=719.818182, yerr=12.096718 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5db0f86a89f1e2e120bbf3166966efaf01984442_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 499.957760E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66212. MP_Allreduce 9776 488. MP_Sync 52 MP_Alltoall 1938 1383689. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.008 0.029 80.973 80.973 qs_mol_dyn_low 1 2.0 0.003 0.004 80.586 80.595 qs_forces 11 3.9 0.003 0.004 80.373 80.381 qs_energies 11 4.9 0.008 0.010 77.528 77.555 scf_env_do_scf 11 5.9 0.000 0.001 68.773 68.776 scf_env_do_scf_inner_loop 99 6.5 0.004 0.016 63.406 63.407 dbcsr_multiply_generic 2055 12.4 0.105 0.109 50.464 50.642 qs_scf_new_mos 99 7.5 0.000 0.001 46.534 46.628 qs_scf_loop_do_ot 99 8.5 0.001 0.001 46.534 46.627 ot_scf_mini 99 9.5 0.002 0.002 44.184 44.275 multiply_cannon 2055 13.4 0.186 0.192 42.342 43.185 velocity_verlet 10 3.0 0.001 0.002 42.414 42.416 multiply_cannon_loop 2055 14.4 1.777 1.825 41.411 42.242 ot_mini 99 10.5 0.001 0.001 25.977 26.063 qs_ot_get_derivative 99 11.5 0.001 0.001 19.212 19.307 multiply_cannon_multrec 49320 15.4 11.466 12.124 17.445 18.182 rebuild_ks_matrix 110 8.3 0.000 0.001 14.399 14.508 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.399 14.508 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.624 12.728 mp_waitall_1 220248 16.4 10.366 11.246 10.366 11.246 multiply_cannon_sync_h2d 49320 15.4 9.730 10.198 9.730 10.198 qs_ot_get_p 110 10.4 0.001 0.001 9.873 9.972 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.260 7.834 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.209 7.692 apply_single 110 13.6 0.000 0.001 7.208 7.692 multiply_cannon_metrocomm3 49320 15.4 0.083 0.086 6.287 7.472 sum_up_and_integrate 110 10.3 0.002 0.003 6.919 6.932 integrate_v_rspace 110 11.3 0.003 0.003 6.894 6.912 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.715 6.730 init_scf_run 11 5.9 0.000 0.001 6.689 6.690 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.689 6.690 ot_diis_step 99 11.5 0.006 0.006 6.537 6.537 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.221 6.361 calculate_rho_elec 110 8.6 0.021 0.024 6.221 6.361 dbcsr_mm_accdrv_process 87628 16.1 2.997 3.120 5.848 6.169 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.637 5.637 init_scf_loop 11 6.9 0.000 0.000 5.341 5.341 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.246 5.278 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.073 5.105 cp_fm_diag_elpa_base 48 14.0 5.060 5.091 5.072 5.104 mp_sum_l 6594 12.7 3.535 4.294 3.535 4.294 make_m2s 4110 13.4 0.060 0.065 4.129 4.222 make_images 4110 14.4 0.178 0.189 4.033 4.132 wfi_extrapolate 11 7.9 0.001 0.001 4.015 4.015 calculate_dm_sparse 110 9.5 0.001 0.001 3.729 3.841 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.515 3.519 density_rs2pw 110 9.6 0.004 0.005 3.279 3.420 grid_integrate_task_list 110 12.3 3.260 3.379 3.260 3.379 multiply_cannon_metrocomm1 49320 15.4 0.065 0.069 2.292 3.345 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.176 3.235 prepare_preconditioner 11 7.9 0.000 0.000 3.190 3.206 make_preconditioner 11 8.9 0.000 0.000 3.190 3.206 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.060 3.105 pw_transfer 1331 11.6 0.055 0.065 3.019 3.099 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.981 3.028 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 2.931 3.011 jit_kernel_multiply 13 15.9 2.574 2.648 2.574 2.648 acc_transpose_blocks 49320 15.4 0.214 0.230 2.554 2.634 fft_wrap_pw1pw2_140 451 13.1 0.391 0.430 2.510 2.599 calculate_first_density_matrix 1 7.0 0.000 0.001 2.575 2.581 potential_pw2rs 110 12.3 0.005 0.006 2.527 2.543 mp_alltoall_d11v 2046 13.8 2.015 2.516 2.015 2.516 fft3d_ps 1111 14.6 0.792 0.882 2.224 2.283 grid_collocate_task_list 110 9.6 2.156 2.273 2.156 2.273 transfer_rs2pw 451 10.6 0.005 0.006 1.858 2.048 mp_waitany 14300 13.8 1.813 2.040 1.813 2.040 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.904 1.921 make_images_data 4110 15.4 0.043 0.047 1.757 1.870 mp_sum_d 3889 11.9 1.303 1.801 1.303 1.801 hybrid_alltoall_any 4261 16.3 0.085 0.484 1.541 1.788 cp_fm_cholesky_invert 11 10.9 1.733 1.737 1.733 1.737 transfer_pw2rs 451 13.1 0.006 0.007 1.709 1.720 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.667 1.706 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=80.973000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=475.545455, yerr=1.827250 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5db0f86a89f1e2e120bbf3166966efaf01984442_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 593.727488E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1972800 MPI messages size (bytes): total size 1.077520E+12 min size 0.000000E+00 max size 4.537280E+06 average size 546.188250E+03 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 222984 1826684928 8192 < size <= 32768 520356 13399818240 32768 < size <= 131072 372336 35386294272 131072 < size <= 4194304 787758 788321309808 4194304 < size <= 16777216 54450 238588003280 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66430. MP_Allreduce 9775 566. MP_Sync 52 MP_Alltoall 1717 3643667. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.067 69.632 69.634 qs_mol_dyn_low 1 2.0 0.003 0.005 69.274 69.291 qs_forces 11 3.9 0.006 0.007 69.204 69.205 qs_energies 11 4.9 0.003 0.012 65.870 65.874 scf_env_do_scf 11 5.9 0.001 0.001 57.163 57.167 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 49.548 49.548 dbcsr_multiply_generic 2055 12.4 0.116 0.120 38.046 38.201 velocity_verlet 10 3.0 0.001 0.003 36.284 36.285 qs_scf_new_mos 99 7.5 0.001 0.001 33.615 33.735 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.614 33.734 multiply_cannon 2055 13.4 0.235 0.255 31.237 32.364 ot_scf_mini 99 9.5 0.003 0.003 31.926 32.039 multiply_cannon_loop 2055 14.4 1.159 1.187 30.050 30.989 ot_mini 99 10.5 0.001 0.001 18.468 18.594 multiply_cannon_multrec 24660 15.4 6.961 8.763 13.988 15.570 rebuild_ks_matrix 110 8.3 0.000 0.000 13.559 13.642 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.015 13.559 13.642 qs_ot_get_derivative 99 11.5 0.001 0.001 12.699 12.819 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.931 12.003 mp_waitall_1 176588 16.5 7.659 9.918 7.659 9.918 multiply_cannon_metrocomm3 24660 15.4 0.072 0.074 5.169 7.744 init_scf_loop 11 6.9 0.001 0.005 7.579 7.580 multiply_cannon_sync_h2d 24660 15.4 6.383 7.511 6.383 7.511 dbcsr_mm_accdrv_process 52282 16.1 5.327 6.042 6.855 7.146 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.449 7.033 apply_single 110 13.6 0.000 0.001 6.448 7.033 qs_ot_get_p 110 10.4 0.001 0.001 6.806 6.949 sum_up_and_integrate 110 10.3 0.001 0.003 6.416 6.426 integrate_v_rspace 110 11.3 0.002 0.003 6.389 6.400 init_scf_run 11 5.9 0.000 0.001 6.272 6.273 scf_env_initial_rho_setup 11 6.9 0.001 0.002 6.272 6.273 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.960 5.971 calculate_rho_elec 110 8.6 0.039 0.047 5.959 5.970 ot_diis_step 99 11.5 0.010 0.011 5.716 5.716 prepare_preconditioner 11 7.9 0.000 0.000 5.554 5.572 make_preconditioner 11 8.9 0.000 0.001 5.554 5.572 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.766 5.469 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.135 5.287 qs_ot_p2m_diag 48 11.0 0.028 0.044 4.859 4.879 make_m2s 4110 13.4 0.057 0.060 4.363 4.689 make_images 4110 14.4 0.418 0.492 4.253 4.578 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.320 4.320 pw_transfer 1331 11.6 0.066 0.073 3.608 3.752 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.730 3.745 cp_fm_diag_elpa_base 48 14.0 3.683 3.700 3.727 3.742 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.502 3.645 wfi_extrapolate 11 7.9 0.001 0.001 3.519 3.519 grid_integrate_task_list 110 12.3 3.163 3.377 3.163 3.377 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.295 3.362 density_rs2pw 110 9.6 0.004 0.005 3.202 3.354 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.306 3.308 fft_wrap_pw1pw2_140 451 13.1 0.460 0.476 2.958 3.100 calculate_dm_sparse 110 9.5 0.001 0.001 2.995 3.022 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 2.941 2.985 fft3d_ps 1111 14.6 1.110 1.340 2.598 2.755 make_images_data 4110 15.4 0.049 0.053 2.371 2.738 hybrid_alltoall_any 4261 16.3 0.106 0.457 2.062 2.694 calculate_first_density_matrix 1 7.0 0.000 0.003 2.655 2.658 cp_fm_cholesky_invert 11 10.9 2.592 2.599 2.592 2.599 mp_sum_l 6594 12.7 1.830 2.502 1.830 2.502 potential_pw2rs 110 12.3 0.009 0.009 2.290 2.307 grid_collocate_task_list 110 9.6 2.172 2.301 2.172 2.301 acc_transpose_blocks 24660 15.4 0.116 0.119 1.975 2.016 mp_alltoall_d11v 2046 13.8 1.800 2.015 1.800 2.015 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.987 2.006 jit_kernel_multiply 10 16.3 1.164 1.933 1.164 1.933 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 1.915 1.916 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.785 1.798 transfer_rs2pw 451 10.6 0.007 0.007 1.525 1.747 multiply_cannon_metrocomm4 22605 15.4 0.079 0.086 0.784 1.744 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.544 1.557 mp_irecv_dv 57340 16.2 0.654 1.523 0.654 1.523 mp_allgather_i34 2055 14.4 0.483 1.507 0.483 1.507 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.387 1.490 mp_waitany 10164 13.8 1.230 1.427 1.230 1.427 dbcsr_complete_redistribute 325 12.2 0.245 0.309 1.161 1.424 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=69.634000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=561.000000, yerr=6.755469 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5db0f86a89f1e2e120bbf3166966efaf01984442_performance_tests/11/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 404.681598E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 3346752 0.0% 0.0% 100.0% average stack size 0.0 0.0 294.1 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 668.504064E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66421. MP_Allreduce 9774 562. MP_Sync 52 MP_Alltoall 1496 4511006. MP_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.033 60.241 60.243 qs_mol_dyn_low 1 2.0 0.003 0.003 59.957 59.968 qs_forces 11 3.9 0.003 0.003 59.853 59.854 qs_energies 11 4.9 0.010 0.012 56.676 56.680 scf_env_do_scf 11 5.9 0.000 0.001 48.822 48.822 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 40.291 40.292 velocity_verlet 10 3.0 0.001 0.002 32.461 32.463 dbcsr_multiply_generic 2055 12.4 0.107 0.110 28.691 28.912 qs_scf_new_mos 99 7.5 0.001 0.001 25.658 25.761 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.658 25.761 ot_scf_mini 99 9.5 0.002 0.002 24.410 24.525 multiply_cannon 2055 13.4 0.212 0.220 22.394 23.584 multiply_cannon_loop 2055 14.4 0.818 0.851 21.236 22.264 ot_mini 99 10.5 0.001 0.001 13.702 13.817 rebuild_ks_matrix 110 8.3 0.000 0.000 12.114 12.236 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.114 12.236 multiply_cannon_multrec 16440 15.4 3.781 5.098 9.965 11.173 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.666 10.775 mp_waitall_1 139946 16.5 6.700 9.703 6.700 9.703 qs_ot_get_derivative 99 11.5 0.001 0.001 9.292 9.406 init_scf_loop 11 6.9 0.000 0.000 8.495 8.496 multiply_cannon_metrocomm3 16440 15.4 0.046 0.047 4.122 6.898 prepare_preconditioner 11 7.9 0.000 0.000 6.756 6.773 make_preconditioner 11 8.9 0.000 0.000 6.755 6.773 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.053 6.399 sum_up_and_integrate 110 10.3 0.001 0.002 6.308 6.325 integrate_v_rspace 110 11.3 0.003 0.003 6.282 6.299 dbcsr_mm_accdrv_process 34862 16.1 5.341 5.719 6.026 6.133 qs_ot_get_p 110 10.4 0.001 0.001 5.891 6.039 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.759 5.770 calculate_rho_elec 110 8.6 0.058 0.059 5.758 5.770 init_scf_run 11 5.9 0.000 0.001 5.456 5.457 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.456 5.456 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.912 5.357 apply_single 110 13.6 0.000 0.000 4.912 5.357 make_m2s 4110 13.4 0.049 0.051 4.222 4.580 make_images 4110 14.4 0.413 0.590 4.107 4.465 ot_diis_step 99 11.5 0.011 0.011 4.379 4.379 qs_ot_p2m_diag 48 11.0 0.041 0.044 4.220 4.224 multiply_cannon_sync_h2d 16440 15.4 3.283 3.986 3.283 3.986 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.155 3.794 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.755 3.756 pw_transfer 1331 11.6 0.065 0.073 3.573 3.580 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.467 3.476 grid_integrate_task_list 110 12.3 3.205 3.422 3.205 3.422 density_rs2pw 110 9.6 0.004 0.004 3.010 3.210 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.188 3.196 cp_fm_diag_elpa_base 48 14.0 3.121 3.154 3.187 3.194 fft_wrap_pw1pw2_140 451 13.1 0.580 0.590 2.978 2.990 wfi_extrapolate 11 7.9 0.001 0.001 2.964 2.964 make_images_data 4110 15.4 0.045 0.050 2.322 2.823 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.780 2.782 hybrid_alltoall_any 4261 16.3 0.109 0.379 2.099 2.756 cp_fm_cholesky_invert 11 10.9 2.645 2.651 2.645 2.651 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.485 2.552 calculate_dm_sparse 110 9.5 0.001 0.001 2.521 2.551 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.441 2.477 fft3d_ps 1111 14.6 1.089 1.097 2.408 2.423 grid_collocate_task_list 110 9.6 2.215 2.416 2.215 2.416 calculate_first_density_matrix 1 7.0 0.000 0.000 2.401 2.403 multiply_cannon_metrocomm4 14385 15.4 0.049 0.052 0.856 2.382 mp_irecv_dv 48980 15.7 0.780 2.241 0.780 2.241 potential_pw2rs 110 12.3 0.011 0.012 2.168 2.178 mp_sum_l 6594 12.7 1.456 2.088 1.456 2.088 mp_alltoall_d11v 2046 13.8 1.783 2.076 1.783 2.076 qs_energies_init_hamiltonians 11 5.9 0.002 0.008 1.949 1.950 dbcsr_complete_redistribute 325 12.2 0.328 0.351 1.471 1.926 cp_fm_upper_to_full 70 14.2 1.391 1.758 1.391 1.758 acc_transpose_blocks 16440 15.4 0.078 0.079 1.585 1.666 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.641 1.656 cp_fm_cholesky_decompose 22 10.9 1.589 1.613 1.589 1.613 transfer_rs2pw 451 10.6 0.005 0.006 1.304 1.501 mp_allgather_i34 2055 14.4 0.472 1.479 0.472 1.479 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.373 1.473 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.461 1.473 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.019 1.467 mp_waitany 17072 13.8 1.125 1.376 1.125 1.376 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.276 1.286 transfer_pw2rs 451 13.1 0.005 0.005 1.209 1.216 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=60.243000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=630.818182, yerr=8.891448 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5db0f86a89f1e2e120bbf3166966efaf01984442_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 739.753984E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66419. MP_Allreduce 9774 603. MP_Sync 52 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.035 65.698 65.703 qs_mol_dyn_low 1 2.0 0.003 0.003 65.468 65.479 qs_forces 11 3.9 0.003 0.003 65.400 65.401 qs_energies 11 4.9 0.002 0.002 62.015 62.019 scf_env_do_scf 11 5.9 0.000 0.001 53.717 53.720 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 42.038 42.038 velocity_verlet 10 3.0 0.002 0.002 37.286 37.288 dbcsr_multiply_generic 2055 12.4 0.113 0.119 30.642 30.852 qs_scf_new_mos 99 7.5 0.001 0.001 27.494 27.614 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.493 27.614 ot_scf_mini 99 9.5 0.002 0.003 25.803 25.908 multiply_cannon 2055 13.4 0.238 0.254 23.583 24.328 multiply_cannon_loop 2055 14.4 1.432 1.479 22.167 22.848 ot_mini 99 10.5 0.001 0.001 14.840 14.969 multiply_cannon_multrec 24660 15.4 4.076 6.762 12.978 14.016 rebuild_ks_matrix 110 8.3 0.000 0.000 11.918 12.029 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.918 12.028 init_scf_loop 11 6.9 0.000 0.000 11.639 11.639 qs_ot_get_derivative 99 11.5 0.001 0.001 10.648 10.757 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.526 10.625 prepare_preconditioner 11 7.9 0.000 0.000 9.911 9.926 make_preconditioner 11 8.9 0.000 0.000 9.911 9.926 dbcsr_mm_accdrv_process 52304 16.0 7.717 9.138 8.744 9.747 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.130 9.607 sum_up_and_integrate 110 10.3 0.001 0.002 6.215 6.230 integrate_v_rspace 110 11.3 0.003 0.003 6.189 6.204 mp_waitall_1 121746 16.5 4.203 6.161 4.203 6.161 qs_ot_get_p 110 10.4 0.001 0.001 5.847 5.982 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.819 5.827 calculate_rho_elec 110 8.6 0.078 0.081 5.819 5.827 make_m2s 4110 13.4 0.060 0.062 5.418 5.763 make_images 4110 14.4 0.578 0.702 5.277 5.618 init_scf_run 11 5.9 0.000 0.001 5.571 5.572 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.571 5.572 cp_fm_upper_to_full 70 14.2 3.414 4.939 3.414 4.939 ot_diis_step 99 11.5 0.011 0.011 4.151 4.151 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.055 4.128 apply_single 110 13.6 0.000 0.000 4.054 4.128 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.981 4.000 dbcsr_complete_redistribute 325 12.2 0.420 0.465 2.648 3.811 pw_transfer 1331 11.6 0.065 0.073 3.636 3.670 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.526 3.575 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.529 3.568 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.385 3.447 grid_integrate_task_list 110 12.3 3.283 3.426 3.283 3.426 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.405 3.406 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.127 3.273 multiply_cannon_metrocomm3 24660 15.4 0.038 0.039 1.374 3.180 density_rs2pw 110 9.6 0.004 0.004 2.948 3.077 fft_wrap_pw1pw2_140 451 13.1 0.608 0.633 3.034 3.074 calculate_dm_sparse 110 9.5 0.001 0.001 2.987 3.022 make_images_data 4110 15.4 0.048 0.052 2.686 3.002 wfi_extrapolate 11 7.9 0.001 0.001 2.992 2.992 hybrid_alltoall_any 4261 16.3 0.123 0.464 2.306 2.991 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.773 2.893 mp_alltoall_i22 605 13.7 1.668 2.889 1.668 2.889 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.866 2.876 cp_fm_diag_elpa_base 48 14.0 2.717 2.777 2.864 2.874 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.815 2.817 cp_fm_cholesky_invert 11 10.9 2.687 2.695 2.687 2.695 multiply_cannon_sync_h2d 24660 15.4 2.361 2.631 2.361 2.631 acc_transpose_blocks 24660 15.4 0.115 0.118 2.463 2.571 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.461 2.492 calculate_first_density_matrix 1 7.0 0.000 0.000 2.483 2.487 fft3d_ps 1111 14.6 1.089 1.121 2.411 2.429 grid_collocate_task_list 110 9.6 2.261 2.426 2.261 2.426 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 2.229 2.230 potential_pw2rs 110 12.3 0.013 0.014 2.031 2.039 mp_alltoall_d11v 2046 13.8 1.750 1.896 1.750 1.896 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.808 1.844 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.613 1.712 cp_fm_cholesky_decompose 22 10.9 1.631 1.677 1.631 1.677 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.632 1.643 jit_kernel_multiply 9 15.8 0.683 1.582 0.683 1.582 mp_sum_l 6594 12.7 0.955 1.565 0.955 1.565 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.539 1.556 acc_transpose_blocks_sync 73980 16.4 1.434 1.530 1.434 1.530 multiply_cannon_metrocomm4 20550 15.4 0.063 0.066 0.859 1.439 mp_irecv_dv 62702 16.1 0.754 1.355 0.754 1.355 transfer_rs2pw 451 10.6 0.005 0.006 1.202 1.340 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=65.703000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=699.545455, yerr=10.499902 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5db0f86a89f1e2e120bbf3166966efaf01984442_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 861.724672E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66417. MP_Allreduce 9774 644. MP_Sync 52 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.031 54.844 54.845 qs_mol_dyn_low 1 2.0 0.003 0.003 54.607 54.617 qs_forces 11 3.9 0.003 0.003 54.542 54.542 qs_energies 11 4.9 0.002 0.002 50.916 50.921 scf_env_do_scf 11 5.9 0.000 0.001 42.638 42.638 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 34.946 34.947 velocity_verlet 10 3.0 0.002 0.002 30.975 30.978 dbcsr_multiply_generic 2055 12.4 0.105 0.109 23.156 23.273 qs_scf_new_mos 99 7.5 0.001 0.001 20.553 20.601 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.553 20.600 ot_scf_mini 99 9.5 0.002 0.002 19.308 19.335 multiply_cannon 2055 13.4 0.236 0.246 17.655 19.044 multiply_cannon_loop 2055 14.4 0.599 0.618 16.379 16.691 rebuild_ks_matrix 110 8.3 0.000 0.000 11.516 11.540 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 11.516 11.539 ot_mini 99 10.5 0.001 0.001 10.636 10.658 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.243 10.263 multiply_cannon_multrec 8220 15.4 3.195 4.284 7.569 8.434 init_scf_loop 11 6.9 0.000 0.000 7.644 7.646 mp_waitall_1 103326 16.6 5.828 7.517 5.828 7.517 qs_ot_get_derivative 99 11.5 0.001 0.001 6.960 6.987 sum_up_and_integrate 110 10.3 0.001 0.002 6.141 6.158 integrate_v_rspace 110 11.3 0.003 0.005 6.114 6.132 prepare_preconditioner 11 7.9 0.000 0.000 6.008 6.011 make_preconditioner 11 8.9 0.000 0.000 6.008 6.011 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.901 5.914 calculate_rho_elec 110 8.6 0.114 0.114 5.900 5.914 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.599 5.672 init_scf_run 11 5.9 0.000 0.001 5.164 5.164 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.164 5.164 dbcsr_mm_accdrv_process 17442 15.9 3.019 4.092 4.235 5.139 qs_ot_get_p 110 10.4 0.001 0.001 4.722 4.740 make_m2s 4110 13.4 0.038 0.039 4.188 4.480 multiply_cannon_metrocomm3 8220 15.4 0.019 0.020 2.964 4.381 make_images 4110 14.4 0.651 0.711 4.058 4.353 pw_transfer 1331 11.6 0.065 0.070 3.796 3.807 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.718 3.756 apply_single 110 13.6 0.000 0.000 3.717 3.755 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.689 3.703 ot_diis_step 99 11.5 0.012 0.012 3.655 3.655 grid_integrate_task_list 110 12.3 3.382 3.499 3.382 3.499 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.361 3.365 fft_wrap_pw1pw2_140 451 13.1 0.779 0.790 3.195 3.211 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.023 3.024 density_rs2pw 110 9.6 0.004 0.004 2.892 3.001 cp_fm_cholesky_invert 11 10.9 2.847 2.851 2.847 2.851 hybrid_alltoall_any 4261 16.3 0.200 0.853 2.286 2.746 wfi_extrapolate 11 7.9 0.001 0.001 2.698 2.698 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.658 2.659 make_images_data 4110 15.4 0.041 0.046 2.243 2.610 calculate_dm_sparse 110 9.5 0.001 0.001 2.509 2.548 grid_collocate_task_list 110 9.6 2.363 2.548 2.363 2.548 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.504 2.505 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.465 2.472 cp_fm_diag_elpa_base 48 14.0 2.410 2.435 2.463 2.471 multiply_cannon_sync_h2d 8220 15.4 2.385 2.457 2.385 2.457 fft3d_ps 1111 14.6 1.127 1.144 2.353 2.374 calculate_first_density_matrix 1 7.0 0.000 0.000 2.371 2.371 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.142 2.157 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.001 2.028 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.785 1.993 potential_pw2rs 110 12.3 0.015 0.016 1.948 1.953 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.897 1.914 mp_alltoall_d11v 2046 13.8 1.723 1.879 1.723 1.879 cp_fm_cholesky_decompose 22 10.9 1.689 1.708 1.689 1.708 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.487 1.605 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.601 1.605 mp_allgather_i34 2055 14.4 0.446 1.533 0.446 1.533 dbcsr_complete_redistribute 325 12.2 0.561 0.581 1.416 1.498 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.467 1.479 acc_transpose_blocks 8220 15.4 0.039 0.040 1.380 1.419 qs_create_task_list 11 7.9 0.000 0.000 1.209 1.307 generate_qs_task_list 11 8.9 0.374 0.441 1.209 1.307 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.230 1.251 transfer_rs2pw 451 10.6 0.005 0.005 1.109 1.250 multiply_cannon_metrocomm1 8220 15.4 0.022 0.022 0.803 1.182 jit_kernel_multiply 7 15.8 0.904 1.172 0.904 1.172 mp_waitany 9240 13.8 1.041 1.165 1.041 1.165 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=54.845000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=811.000000, yerr=12.613124 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5db0f86a89f1e2e120bbf3166966efaf01984442_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 1.349435E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67098. MP_Allreduce 9752 812. MP_Sync 52 MP_Alltoall 1474 16505187. MP_SendRecv 2310 360267. MP_ISendRecv 2310 360267. MP_Wait 5214 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.034 86.738 86.739 qs_mol_dyn_low 1 2.0 0.003 0.003 86.438 86.448 qs_forces 11 3.9 0.003 0.003 86.370 86.371 qs_energies 11 4.9 0.002 0.004 82.271 82.273 scf_env_do_scf 11 5.9 0.000 0.001 72.357 72.357 velocity_verlet 10 3.0 0.005 0.007 55.353 55.360 scf_env_do_scf_inner_loop 99 6.5 0.004 0.008 44.033 44.033 dbcsr_multiply_generic 2055 12.4 0.120 0.124 29.241 29.335 init_scf_loop 11 6.9 0.000 0.000 28.253 28.254 qs_scf_new_mos 99 7.5 0.001 0.001 26.819 26.855 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.818 26.854 prepare_preconditioner 11 7.9 0.000 0.000 26.236 26.244 make_preconditioner 11 8.9 0.000 0.000 26.236 26.244 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.372 25.726 ot_scf_mini 99 9.5 0.002 0.002 25.034 25.057 multiply_cannon 2055 13.4 0.338 0.364 22.192 22.986 multiply_cannon_loop 2055 14.4 0.801 0.812 20.421 20.694 cp_fm_upper_to_full 70 14.2 12.750 18.359 12.750 18.359 ot_mini 99 10.5 0.001 0.001 13.977 14.002 rebuild_ks_matrix 110 8.3 0.000 0.000 13.535 13.570 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 13.535 13.569 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.280 12.311 dbcsr_complete_redistribute 325 12.2 1.016 1.036 7.417 10.670 multiply_cannon_multrec 8220 15.4 4.043 4.214 9.679 9.808 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.410 9.660 qs_ot_get_derivative 99 11.5 0.001 0.001 9.421 9.447 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.849 9.076 mp_alltoall_i22 605 13.7 5.479 8.713 5.479 8.713 mp_waitall_1 84994 16.7 7.383 8.191 7.383 8.191 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.081 7.122 calculate_rho_elec 110 8.6 0.224 0.225 7.080 7.121 sum_up_and_integrate 110 10.3 0.002 0.002 6.705 6.720 integrate_v_rspace 110 11.3 0.003 0.004 6.677 6.693 make_m2s 4110 13.4 0.043 0.044 5.353 5.841 qs_ot_get_p 110 10.4 0.001 0.001 5.802 5.831 dbcsr_mm_accdrv_process 11614 15.7 3.811 4.105 5.489 5.710 make_images 4110 14.4 0.888 0.948 5.164 5.650 init_scf_run 11 5.9 0.000 0.001 5.648 5.649 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.648 5.648 cp_fm_cholesky_invert 11 10.9 5.184 5.189 5.184 5.189 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.706 5.166 apply_single 110 13.6 0.000 0.000 4.706 5.165 multiply_cannon_metrocomm3 8220 15.4 0.019 0.020 4.743 5.079 pw_transfer 1331 11.6 0.074 0.075 4.929 4.937 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 4.813 4.821 ot_diis_step 99 11.5 0.015 0.015 4.540 4.540 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.225 4.233 fft_wrap_pw1pw2_140 451 13.1 1.279 1.285 4.205 4.213 grid_integrate_task_list 110 12.3 3.675 3.732 3.675 3.732 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.711 3.711 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.659 3.660 density_rs2pw 110 9.6 0.004 0.004 3.545 3.550 hybrid_alltoall_any 4261 16.3 0.263 0.564 2.805 3.536 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.983 3.437 make_images_data 4110 15.4 0.045 0.048 2.795 3.418 wfi_extrapolate 11 7.9 0.001 0.001 3.305 3.305 multiply_cannon_sync_h2d 8220 15.4 3.161 3.185 3.161 3.185 calculate_dm_sparse 110 9.5 0.001 0.001 3.156 3.182 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.100 3.101 cp_fm_diag_elpa_base 48 14.0 2.553 2.752 3.098 3.099 fft3d_ps 1111 14.6 1.297 1.306 2.867 2.878 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.865 2.868 grid_collocate_task_list 110 9.6 2.666 2.670 2.666 2.670 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.638 2.653 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.359 2.367 qs_env_update_s_mstruct 11 6.9 0.001 0.001 2.175 2.249 potential_pw2rs 110 12.3 0.022 0.022 2.239 2.241 calculate_first_density_matrix 1 7.0 0.000 0.000 2.231 2.231 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.093 2.198 mp_alltoall_d11v 2046 13.8 2.022 2.083 2.022 2.083 cp_fm_cholesky_decompose 22 10.9 1.954 1.972 1.954 1.972 qs_create_task_list 11 7.9 0.000 0.000 1.878 1.922 generate_qs_task_list 11 8.9 0.730 0.782 1.878 1.922 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.854 1.857 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.734 1.769 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=86.739000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1241.363636, yerr=32.184900 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5db0f86a89f1e2e120bbf3166966efaf01984442_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 629.456896E+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 4020 57505. MP_Allreduce 11129 795. MP_Sync 87 MP_Alltoall 2226 3057550. MP_SendRecv 24320 18752. MP_ISendRecv 24320 18752. MP_Wait 42476 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.030 206.125 206.127 qs_mol_dyn_low 1 2.0 0.003 0.003 205.682 205.698 qs_forces 11 3.9 0.005 0.005 205.599 205.599 qs_energies 11 4.9 0.002 0.003 200.024 200.041 scf_env_do_scf 11 5.9 0.001 0.001 183.627 183.632 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 163.166 163.169 dbcsr_multiply_generic 2507 12.6 0.181 0.184 125.627 126.574 qs_scf_new_mos 117 7.6 0.001 0.001 123.852 124.276 qs_scf_loop_do_ot 117 8.6 0.001 0.001 123.851 124.275 velocity_verlet 10 3.0 0.002 0.003 123.959 123.960 ot_scf_mini 117 9.6 0.003 0.003 117.193 117.566 multiply_cannon 2507 13.6 0.251 0.261 101.744 103.874 multiply_cannon_loop 2507 14.6 2.377 2.444 99.582 101.805 ot_mini 117 10.6 0.001 0.001 66.114 66.517 multiply_cannon_multrec 60168 15.6 32.018 34.440 41.712 43.600 qs_ot_get_derivative 117 11.6 0.001 0.001 41.116 41.494 rebuild_ks_matrix 128 8.3 0.001 0.001 33.525 33.871 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 33.524 33.870 mp_waitall_1 267128 16.5 28.823 31.083 28.823 31.083 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.113 30.421 qs_ot_get_p 128 10.4 0.001 0.001 28.741 29.204 multiply_cannon_sync_h2d 60168 15.6 26.444 28.595 26.444 28.595 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.492 25.747 apply_single 128 13.6 0.001 0.001 24.492 25.747 ot_diis_step 117 11.6 0.008 0.008 24.783 24.784 qs_ot_p2m_diag 83 11.4 0.079 0.092 22.157 22.217 init_scf_loop 11 6.9 0.000 0.000 20.384 20.385 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.504 19.796 cp_dbcsr_syevd 83 12.4 0.005 0.005 19.383 19.384 multiply_cannon_metrocomm3 60168 15.6 0.117 0.124 16.078 19.091 cp_fm_diag_elpa 83 13.4 0.000 0.000 16.452 16.484 cp_fm_diag_elpa_base 83 14.4 16.382 16.416 16.449 16.481 prepare_preconditioner 11 7.9 0.000 0.000 15.803 15.858 make_preconditioner 11 8.9 0.000 0.000 15.803 15.858 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.041 15.215 make_m2s 5014 13.6 0.105 0.114 13.790 14.241 make_images 5014 14.6 0.398 0.416 13.607 14.072 sum_up_and_integrate 128 10.3 0.002 0.004 13.931 13.947 integrate_v_rspace 128 11.3 0.004 0.004 13.872 13.889 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.084 13.215 calculate_rho_elec 128 8.7 0.045 0.064 13.083 13.214 init_scf_run 11 5.9 0.000 0.001 12.246 12.247 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.246 12.247 mp_sum_l 7950 12.9 8.828 10.824 8.828 10.824 dbcsr_mm_accdrv_process 124484 16.2 4.663 4.859 9.259 9.863 wfi_extrapolate 11 7.9 0.001 0.001 9.023 9.023 cp_fm_cholesky_invert 11 10.9 9.003 9.011 9.003 9.011 calculate_dm_sparse 128 9.5 0.001 0.001 8.459 8.542 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.304 8.463 qs_ot_get_orbitals 117 10.6 0.001 0.001 8.148 8.233 multiply_cannon_metrocomm1 60168 15.6 0.094 0.099 6.294 8.028 pw_transfer 1547 11.6 0.073 0.093 7.577 7.750 make_images_data 5014 15.6 0.066 0.072 6.763 7.707 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 7.375 7.552 grid_integrate_task_list 128 12.3 7.034 7.444 7.034 7.444 density_rs2pw 128 9.7 0.006 0.007 6.730 7.306 hybrid_alltoall_any 5200 16.5 0.295 2.255 5.914 7.174 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.660 6.672 fft_wrap_pw1pw2_140 523 13.2 1.126 1.182 6.468 6.643 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.821 5.890 fft3d_ps 1291 14.7 2.193 2.842 5.381 5.708 mp_alltoall_d11v 2415 14.1 4.302 5.632 4.302 5.632 grid_collocate_task_list 128 9.7 4.842 5.186 4.842 5.186 cp_fm_cholesky_decompose 22 10.9 4.626 4.641 4.626 4.641 potential_pw2rs 128 12.3 0.009 0.010 4.446 4.469 mp_sum_d 4479 12.1 3.440 4.443 3.440 4.443 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=206.127000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=596.000000, yerr=5.443929 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5db0f86a89f1e2e120bbf3166966efaf01984442_performance_tests/16/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 2.183246E+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 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 839.155712E+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 1463835059104 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4020 57958. MP_Allreduce 11128 958. MP_Sync 87 MP_Alltoall 1969 5618843. 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 0.010 0.031 188.238 188.239 qs_mol_dyn_low 1 2.0 0.003 0.003 187.844 187.859 qs_forces 11 3.9 0.004 0.005 187.753 187.758 qs_energies 11 4.9 0.002 0.003 181.028 181.043 scf_env_do_scf 11 5.9 0.001 0.001 164.814 164.824 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 132.185 132.187 velocity_verlet 10 3.0 0.001 0.002 118.585 118.586 dbcsr_multiply_generic 2507 12.6 0.190 0.195 97.307 98.459 qs_scf_new_mos 117 7.6 0.001 0.001 93.721 94.251 qs_scf_loop_do_ot 117 8.6 0.001 0.001 93.720 94.250 ot_scf_mini 117 9.6 0.004 0.004 88.913 89.562 multiply_cannon 2507 13.6 0.502 0.557 77.300 81.645 multiply_cannon_loop 2507 14.6 1.566 1.634 73.927 76.837 ot_mini 117 10.6 0.001 0.001 50.155 50.724 mp_waitall_1 214728 16.6 23.905 39.824 23.905 39.824 multiply_cannon_multrec 30084 15.6 21.036 25.528 31.639 36.439 rebuild_ks_matrix 128 8.3 0.001 0.001 32.210 32.865 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.019 32.210 32.864 init_scf_loop 11 6.9 0.000 0.000 32.539 32.540 qs_ks_update_qs_env 128 7.6 0.001 0.001 28.971 29.571 multiply_cannon_metrocomm3 30084 15.6 0.096 0.101 15.618 29.434 qs_ot_get_derivative 117 11.6 0.001 0.001 28.222 28.868 prepare_preconditioner 11 7.9 0.000 0.000 28.231 28.295 make_preconditioner 11 8.9 0.000 0.000 28.231 28.294 make_full_inverse_cholesky 11 9.9 0.000 0.000 26.959 27.503 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.093 23.229 apply_single 128 13.6 0.001 0.001 22.093 23.229 ot_diis_step 117 11.6 0.015 0.015 21.753 21.755 qs_ot_get_p 128 10.4 0.001 0.001 20.777 21.403 multiply_cannon_sync_h2d 30084 15.6 18.036 20.485 18.036 20.485 cp_fm_cholesky_invert 11 10.9 16.439 16.451 16.439 16.451 qs_ot_p2m_diag 83 11.4 0.188 0.216 15.884 15.918 make_m2s 5014 13.6 0.090 0.095 14.087 15.791 make_images 5014 14.6 1.175 1.376 13.874 15.574 cp_dbcsr_syevd 83 12.4 0.005 0.006 14.684 14.685 sum_up_and_integrate 128 10.3 0.002 0.004 13.939 13.964 integrate_v_rspace 128 11.3 0.003 0.003 13.878 13.905 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.331 13.363 calculate_rho_elec 128 8.7 0.087 0.104 13.330 13.362 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 11.151 11.597 cp_fm_diag_elpa 83 13.4 0.000 0.000 11.539 11.570 cp_fm_diag_elpa_base 83 14.4 11.286 11.379 11.533 11.559 init_scf_run 11 5.9 0.000 0.001 11.430 11.431 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.430 11.431 multiply_cannon_metrocomm4 27577 15.6 0.107 0.120 3.853 10.685 dbcsr_mm_accdrv_process 62242 16.2 5.470 6.219 10.048 10.655 make_images_data 5014 15.6 0.067 0.076 8.297 10.472 mp_irecv_dv 69486 16.3 3.646 10.283 3.646 10.283 hybrid_alltoall_any 5200 16.5 0.352 1.534 7.104 9.764 pw_transfer 1547 11.6 0.084 0.092 8.512 8.580 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.289 8.357 wfi_extrapolate 11 7.9 0.001 0.001 8.353 8.353 density_rs2pw 128 9.7 0.006 0.007 7.052 7.572 grid_integrate_task_list 128 12.3 7.187 7.479 7.187 7.479 fft_wrap_pw1pw2_140 523 13.2 1.213 1.236 7.346 7.432 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.242 7.071 cp_fm_cholesky_decompose 22 10.9 6.876 6.945 6.876 6.945 calculate_dm_sparse 128 9.5 0.001 0.001 6.460 6.608 mp_sum_l 7950 12.9 4.116 6.171 4.116 6.171 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.091 6.102 fft3d_ps 1291 14.7 2.840 3.022 5.850 5.911 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.358 5.429 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.213 5.396 grid_collocate_task_list 128 9.7 5.001 5.351 5.001 5.351 mp_allgather_i34 2507 14.6 1.892 4.966 1.892 4.966 mp_alltoall_d11v 2415 14.1 4.082 4.503 4.082 4.503 potential_pw2rs 128 12.3 0.016 0.019 4.435 4.447 dbcsr_complete_redistribute 395 12.7 0.776 0.893 3.152 3.996 mp_sum_d 4479 12.1 2.590 3.953 2.590 3.953 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=188.239000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=800.272727, yerr=2.299838 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5db0f86a89f1e2e120bbf3166966efaf01984442_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 2.950886E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 4015680 0.0% 0.0% 100.0% average stack size 0.0 0.0 1694.9 marketing flops 145.646636E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 972.173312E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1051232 MPI messages size (bytes): total size 2.737039E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.603648E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 284160 37245419520 131072 < size <= 4194304 665408 1004401590272 4194304 < size <= 16777216 66080 937889764224 16777216 < size 28896 757491302400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4024 57897. MP_Allreduce 11137 997. MP_Sync 87 MP_Alltoall 1724 9394172. MP_SendRecv 7998 75008. MP_ISendRecv 7998 75008. MP_Wait 21986 MP_ISend 11836 275177. MP_IRecv 11836 275177. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.027 176.592 176.593 qs_mol_dyn_low 1 2.0 0.003 0.003 176.145 176.158 qs_forces 11 3.9 0.004 0.004 175.590 175.592 qs_energies 11 4.9 0.002 0.003 169.013 169.024 scf_env_do_scf 11 5.9 0.001 0.001 153.431 153.432 scf_env_do_scf_inner_loop 118 6.6 0.066 0.108 118.473 118.474 velocity_verlet 10 3.0 0.002 0.002 113.747 113.763 dbcsr_multiply_generic 2527 12.6 0.182 0.186 82.675 83.746 qs_scf_new_mos 118 7.6 0.001 0.001 81.068 81.433 qs_scf_loop_do_ot 118 8.6 0.001 0.001 81.068 81.432 ot_scf_mini 118 9.6 0.003 0.004 76.780 77.202 multiply_cannon 2527 13.6 0.510 0.531 62.846 67.820 multiply_cannon_loop 2527 14.6 1.142 1.209 59.706 62.294 ot_mini 118 10.6 0.001 0.001 42.834 43.244 init_scf_loop 11 6.9 0.000 0.000 34.856 34.857 mp_waitall_1 171878 16.6 24.780 34.212 24.780 34.212 rebuild_ks_matrix 129 8.3 0.001 0.001 30.449 30.887 qs_ks_build_kohn_sham_matrix 129 9.3 0.016 0.017 30.449 30.887 prepare_preconditioner 11 7.9 0.000 0.000 30.810 30.858 make_preconditioner 11 8.9 0.000 0.000 30.810 30.858 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.439 29.827 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.440 27.841 multiply_cannon_multrec 20216 15.6 13.174 16.718 22.521 26.058 multiply_cannon_metrocomm3 20216 15.6 0.062 0.066 15.348 24.762 qs_ot_get_derivative 118 11.6 0.001 0.002 23.158 23.582 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.681 20.732 apply_single 129 13.6 0.001 0.001 19.681 20.731 ot_diis_step 118 11.6 0.018 0.018 19.569 19.570 qs_ot_get_p 129 10.4 0.001 0.001 19.082 19.483 make_m2s 5054 13.6 0.080 0.086 14.837 16.224 make_images 5054 14.6 1.162 1.257 14.602 15.984 multiply_cannon_sync_h2d 20216 15.6 13.900 15.562 13.900 15.562 qs_ot_p2m_diag 83 11.4 0.265 0.272 14.728 14.739 cp_fm_cholesky_invert 11 10.9 14.363 14.372 14.363 14.372 sum_up_and_integrate 129 10.3 0.002 0.003 13.952 13.980 integrate_v_rspace 129 11.3 0.003 0.003 13.892 13.917 cp_dbcsr_syevd 83 12.4 0.005 0.006 13.671 13.672 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.610 13.645 calculate_rho_elec 129 8.7 0.131 0.146 13.609 13.645 make_images_data 5054 15.6 0.063 0.071 8.882 10.700 init_scf_run 11 5.9 0.000 0.001 10.504 10.504 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.504 10.504 cp_fm_diag_elpa 83 13.4 0.000 0.000 10.462 10.484 cp_fm_diag_elpa_base 83 14.4 10.046 10.219 10.459 10.481 hybrid_alltoall_any 5240 16.5 0.453 2.048 7.713 10.287 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.074 9.349 multiply_cannon_metrocomm4 17689 15.6 0.069 0.080 3.498 9.246 mp_irecv_dv 50625 16.2 3.366 8.983 3.366 8.983 dbcsr_mm_accdrv_process 41830 16.2 5.627 5.960 8.798 8.941 pw_transfer 1559 11.6 0.085 0.102 8.729 8.837 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 8.504 8.617 grid_integrate_task_list 129 12.3 7.392 7.857 7.392 7.857 fft_wrap_pw1pw2_140 527 13.2 1.314 1.347 7.568 7.691 wfi_extrapolate 11 7.9 0.001 0.001 7.453 7.453 density_rs2pw 129 9.7 0.006 0.006 6.925 7.408 cp_fm_cholesky_decompose 22 10.9 7.333 7.403 7.333 7.403 cp_fm_upper_to_full 105 14.8 5.688 7.152 5.688 7.152 dbcsr_complete_redistribute 395 12.7 1.173 1.207 4.589 6.314 calculate_dm_sparse 129 9.5 0.001 0.001 5.884 6.000 fft3d_ps 1301 14.7 2.785 3.007 5.859 5.940 grid_collocate_task_list 129 9.7 5.198 5.554 5.198 5.554 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 4.804 5.475 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.446 5.453 mp_alltoall_d11v 2423 14.1 4.354 5.188 4.354 5.188 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.428 5.144 mp_allgather_i34 2527 14.6 1.626 4.947 1.626 4.947 mp_sum_l 8010 12.9 3.142 4.783 3.142 4.783 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.560 4.675 potential_pw2rs 129 12.3 0.021 0.023 4.315 4.322 transfer_fm_to_dbcsr 11 9.9 0.018 0.026 2.352 4.030 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.982 4.015 qs_energies_init_hamiltonians 11 5.9 0.015 0.025 3.780 3.781 mp_alltoall_i22 718 14.1 1.872 3.742 1.872 3.742 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.512 3.548 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=176.593000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=921.545455, yerr=11.633522 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5db0f86a89f1e2e120bbf3166966efaf01984442_performance_tests/18/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 4.353791E+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 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 1.187901E+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 665379475120 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 58205. MP_Allreduce 11082 1082. 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 0.022 0.049 188.008 188.010 qs_mol_dyn_low 1 2.0 0.003 0.003 187.593 187.607 qs_forces 11 3.9 0.007 0.009 187.445 187.455 qs_energies 11 4.9 0.002 0.002 180.360 180.373 scf_env_do_scf 11 5.9 0.001 0.001 163.584 163.596 velocity_verlet 10 3.0 0.002 0.002 124.816 124.820 scf_env_do_scf_inner_loop 117 6.6 0.010 0.026 116.957 116.959 dbcsr_multiply_generic 2507 12.6 0.188 0.195 80.576 81.407 qs_scf_new_mos 117 7.6 0.001 0.001 80.777 81.089 qs_scf_loop_do_ot 117 8.6 0.001 0.001 80.776 81.089 ot_scf_mini 117 9.6 0.003 0.004 76.193 76.544 multiply_cannon 2507 13.6 0.552 0.583 55.915 58.289 multiply_cannon_loop 2507 14.6 1.850 1.937 52.082 54.190 init_scf_loop 11 6.9 0.000 0.000 46.504 46.505 ot_mini 117 10.6 0.001 0.001 42.981 43.329 prepare_preconditioner 11 7.9 0.000 0.000 42.378 42.396 make_preconditioner 11 8.9 0.000 0.000 42.378 42.396 make_full_inverse_cholesky 11 9.9 0.011 0.021 35.968 41.039 multiply_cannon_multrec 30084 15.6 13.520 19.056 26.394 31.598 rebuild_ks_matrix 128 8.3 0.001 0.001 29.248 29.604 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.019 29.248 29.604 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.407 26.725 mp_waitall_1 147882 16.7 17.026 26.372 17.026 26.372 qs_ot_get_derivative 117 11.6 0.001 0.002 23.349 23.701 make_m2s 5014 13.6 0.097 0.102 20.198 21.306 make_images 5014 14.6 1.955 2.299 19.893 21.002 qs_ot_get_p 128 10.4 0.001 0.001 19.345 19.711 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 18.978 19.581 apply_single 128 13.6 0.001 0.001 18.977 19.581 ot_diis_step 117 11.6 0.018 0.018 19.496 19.497 cp_fm_upper_to_full 105 14.8 11.594 17.093 11.594 17.093 cp_fm_cholesky_invert 11 10.9 16.131 16.140 16.131 16.140 qs_ot_p2m_diag 83 11.4 0.342 0.390 15.154 15.204 multiply_cannon_metrocomm3 30084 15.6 0.050 0.054 6.284 15.175 sum_up_and_integrate 128 10.3 0.002 0.003 13.903 13.928 integrate_v_rspace 128 11.3 0.003 0.004 13.844 13.870 cp_dbcsr_syevd 83 12.4 0.005 0.006 13.753 13.754 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.688 13.739 calculate_rho_elec 128 8.7 0.173 0.189 13.688 13.738 dbcsr_complete_redistribute 395 12.7 1.505 1.643 9.137 12.957 dbcsr_mm_accdrv_process 62264 16.2 8.379 9.178 12.450 12.946 make_images_data 5014 15.6 0.066 0.072 10.809 12.643 multiply_cannon_sync_h2d 30084 15.6 10.476 11.709 10.476 11.709 hybrid_alltoall_any 5200 16.5 0.529 2.222 9.661 11.573 copy_fm_to_dbcsr 209 11.7 0.001 0.002 7.767 11.566 init_scf_run 11 5.9 0.000 0.001 10.863 10.865 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.863 10.865 cp_fm_diag_elpa 83 13.4 0.000 0.000 10.523 10.535 cp_fm_diag_elpa_base 83 14.4 9.540 9.873 10.517 10.528 transfer_fm_to_dbcsr 11 9.9 0.001 0.007 6.387 10.058 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.800 10.053 mp_alltoall_i22 716 14.1 5.672 9.454 5.672 9.454 pw_transfer 1547 11.6 0.085 0.099 8.965 9.066 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.739 8.846 grid_integrate_task_list 128 12.3 7.565 7.996 7.565 7.996 fft_wrap_pw1pw2_140 523 13.2 1.444 1.480 7.772 7.890 wfi_extrapolate 11 7.9 0.001 0.001 7.723 7.724 cp_fm_cholesky_decompose 22 10.9 7.551 7.648 7.551 7.648 multiply_cannon_metrocomm4 25070 15.6 0.086 0.097 2.872 7.497 density_rs2pw 128 9.7 0.006 0.006 6.752 7.271 mp_irecv_dv 76098 16.2 2.713 7.216 2.713 7.216 calculate_dm_sparse 128 9.5 0.001 0.001 6.265 6.350 fft3d_ps 1291 14.7 2.848 2.960 5.924 6.006 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.485 5.543 grid_collocate_task_list 128 9.7 5.293 5.541 5.293 5.541 mp_alltoall_d11v 2415 14.1 4.765 5.294 4.765 5.294 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.485 4.595 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.361 4.494 qs_energies_init_hamiltonians 11 5.9 0.012 0.018 4.422 4.429 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.238 4.301 potential_pw2rs 128 12.3 0.023 0.024 4.219 4.236 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=188.010000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1103.181818, yerr=26.866059 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5db0f86a89f1e2e120bbf3166966efaf01984442_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430457200640 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986252263424 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613077360640 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239167967232 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239167967232 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.233007E+12 0.0% 0.0% 100.0% flops max/rank 5.910121E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383584 0.0% 0.0% 100.0% number of processed stacks 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 1.532899E+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 726592338480 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 9010 51. MP_Alltoall 9724 794507. MP_ISend 40420 2100460. MP_IRecv 40420 2099564. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 57624. MP_Allreduce 11184 1163. MP_Sync 88 MP_Alltoall 1724 18848050. 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 0.080 0.107 170.174 170.180 qs_mol_dyn_low 1 2.0 0.003 0.003 169.701 169.723 qs_forces 11 3.9 0.004 0.004 169.604 169.608 qs_energies 11 4.9 0.002 0.002 162.211 162.222 scf_env_do_scf 11 5.9 0.001 0.001 145.123 145.136 velocity_verlet 10 3.0 0.002 0.002 112.305 112.309 scf_env_do_scf_inner_loop 118 6.6 0.003 0.009 109.409 109.410 dbcsr_multiply_generic 2529 12.6 0.183 0.195 72.694 73.114 qs_scf_new_mos 118 7.6 0.001 0.001 72.568 72.695 qs_scf_loop_do_ot 118 8.6 0.001 0.001 72.567 72.695 ot_scf_mini 118 9.6 0.003 0.004 68.119 68.271 multiply_cannon 2529 13.6 0.565 0.579 54.014 57.161 multiply_cannon_loop 2529 14.6 0.814 0.848 50.819 51.856 ot_mini 118 10.6 0.001 0.001 38.106 38.249 init_scf_loop 11 6.9 0.000 0.000 35.564 35.565 prepare_preconditioner 11 7.9 0.000 0.000 31.705 31.730 make_preconditioner 11 8.9 0.000 0.000 31.705 31.730 mp_waitall_1 126876 16.7 25.021 31.603 25.021 31.603 make_full_inverse_cholesky 11 9.9 0.015 0.027 29.571 29.828 rebuild_ks_matrix 129 8.3 0.001 0.001 29.127 29.275 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.018 29.126 29.275 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.585 26.721 multiply_cannon_multrec 10116 15.6 10.396 16.106 17.965 22.451 qs_ot_get_derivative 118 11.6 0.001 0.002 20.688 20.840 multiply_cannon_metrocomm3 10116 15.6 0.025 0.027 13.101 20.614 cp_fm_cholesky_invert 11 10.9 18.133 18.139 18.133 18.139 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 17.474 17.634 apply_single 129 13.6 0.001 0.001 17.474 17.634 ot_diis_step 118 11.6 0.020 0.020 17.346 17.346 qs_ot_get_p 129 10.4 0.001 0.001 17.171 17.299 make_m2s 5058 13.6 0.066 0.068 14.777 15.687 make_images 5058 14.6 2.189 2.635 14.467 15.373 sum_up_and_integrate 129 10.3 0.002 0.002 14.165 14.214 integrate_v_rspace 129 11.3 0.004 0.004 14.104 14.155 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.995 14.022 calculate_rho_elec 129 8.7 0.258 0.269 13.994 14.021 qs_ot_p2m_diag 84 11.4 0.501 0.507 13.402 13.419 cp_dbcsr_syevd 84 12.4 0.005 0.006 12.222 12.222 multiply_cannon_sync_h2d 10116 15.6 10.817 11.253 10.817 11.253 init_scf_run 11 5.9 0.000 0.001 10.284 10.285 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.284 10.284 make_images_data 5058 15.6 0.056 0.065 8.440 9.794 hybrid_alltoall_any 5245 16.5 0.847 3.806 8.280 9.367 cp_fm_diag_elpa 84 13.4 0.000 0.000 9.303 9.314 cp_fm_diag_elpa_base 84 14.4 9.062 9.144 9.298 9.308 pw_transfer 1559 11.6 0.085 0.094 9.277 9.307 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 9.053 9.089 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 8.141 8.261 grid_integrate_task_list 129 12.3 7.802 8.160 7.802 8.160 fft_wrap_pw1pw2_140 527 13.2 1.815 1.853 8.021 8.063 cp_fm_cholesky_decompose 22 10.9 7.850 7.971 7.850 7.971 dbcsr_mm_accdrv_process 20934 16.1 3.334 4.466 7.212 7.896 multiply_cannon_metrocomm1 10116 15.6 0.030 0.031 4.259 7.454 wfi_extrapolate 11 7.9 0.001 0.001 7.316 7.316 density_rs2pw 129 9.7 0.006 0.006 6.663 6.865 calculate_dm_sparse 129 9.5 0.001 0.001 6.108 6.187 grid_collocate_task_list 129 9.7 5.608 5.875 5.608 5.875 multiply_cannon_metrocomm4 7587 15.6 0.027 0.030 1.887 5.820 fft3d_ps 1301 14.7 2.785 2.861 5.763 5.798 mp_irecv_dv 29102 15.9 1.845 5.725 1.845 5.725 dbcsr_complete_redistribute 397 12.7 2.113 2.158 5.267 5.589 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.243 5.247 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.116 5.142 mp_alltoall_d11v 2429 14.1 4.689 5.079 4.689 5.079 mp_allgather_i34 2529 14.6 1.366 4.672 1.366 4.672 potential_pw2rs 129 12.3 0.027 0.028 4.258 4.264 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.020 4.068 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.592 3.896 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.680 3.730 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.624 3.722 copy_fm_to_dbcsr 210 11.7 0.001 0.002 3.407 3.708 copy_dbcsr_to_fm 187 11.8 0.004 0.004 3.458 3.556 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.451 3.464 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=170.180000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1441.000000, yerr=19.840271 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5db0f86a89f1e2e120bbf3166966efaf01984442_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430454546432 0.0% 0.0% 100.0% flops 32 x 32 x 32 1973537472512 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992006770688 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613065416704 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239182565376 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239182565376 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.243766E+12 0.0% 0.0% 100.0% flops max/rank 11.787405E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806547424 0.0% 0.0% 100.0% number of processed stacks 1982128 0.0% 0.0% 100.0% average stack size 0.0 0.0 3434.0 marketing flops 145.661668E+12 ------------------------------------------------------------------------------- # multiplications 2534 max memory usage/rank 3.116646E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101360 MPI messages size (bytes): total size 1.145137E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.297725E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 45848 35601252352 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592540656 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 58558. MP_Allreduce 11184 1500. MP_Sync 88 MP_Alltoall 1724 36993632. MP_SendRecv 1806 218624. MP_ISendRecv 1806 218624. MP_Wait 9876 MP_ISend 6456 1080169. MP_IRecv 6456 1080169. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.037 0.056 289.871 289.873 qs_mol_dyn_low 1 2.0 0.003 0.003 289.250 289.264 qs_forces 11 3.9 0.004 0.004 289.152 289.162 qs_energies 11 4.9 0.002 0.002 280.353 280.361 scf_env_do_scf 11 5.9 0.001 0.001 258.058 258.063 velocity_verlet 10 3.0 0.003 0.005 210.111 210.119 scf_env_do_scf_inner_loop 118 6.6 0.004 0.009 134.543 134.544 init_scf_loop 11 6.9 0.000 0.000 123.240 123.242 prepare_preconditioner 11 7.9 0.000 0.000 118.485 118.506 make_preconditioner 11 8.9 0.000 0.000 118.485 118.506 make_full_inverse_cholesky 11 9.9 0.037 0.039 95.083 115.626 qs_scf_new_mos 118 7.6 0.001 0.001 90.935 90.987 qs_scf_loop_do_ot 118 8.6 0.001 0.001 90.934 90.986 ot_scf_mini 118 9.6 0.004 0.004 86.039 86.072 dbcsr_multiply_generic 2534 12.6 0.215 0.229 82.522 83.004 cp_fm_upper_to_full 106 14.8 52.610 74.987 52.610 74.987 multiply_cannon 2534 13.6 0.686 0.749 59.075 60.568 multiply_cannon_loop 2534 14.6 1.036 1.068 54.988 56.268 ot_mini 118 10.6 0.001 0.001 44.949 45.003 dbcsr_complete_redistribute 397 12.7 4.005 4.041 29.355 42.061 copy_fm_to_dbcsr 210 11.7 0.001 0.002 25.832 38.499 transfer_fm_to_dbcsr 11 9.9 0.030 0.030 23.359 35.904 rebuild_ks_matrix 129 8.3 0.001 0.001 34.053 34.078 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.018 34.053 34.077 mp_alltoall_i22 720 14.1 21.128 33.661 21.128 33.661 cp_fm_cholesky_invert 11 10.9 33.447 33.455 33.447 33.455 mp_waitall_1 104780 16.8 28.192 32.381 28.192 32.381 qs_ks_update_qs_env 129 7.6 0.001 0.001 31.630 31.659 qs_ot_get_p 129 10.4 0.001 0.001 26.097 26.186 qs_ot_get_derivative 118 11.6 0.002 0.002 24.809 24.840 qs_ot_p2m_diag 84 11.4 0.890 0.895 21.961 21.991 make_m2s 5068 13.6 0.076 0.080 18.810 20.316 cp_dbcsr_syevd 84 12.4 0.006 0.006 20.141 20.141 ot_diis_step 118 11.6 0.022 0.024 20.101 20.102 multiply_cannon_metrocomm3 10136 15.6 0.025 0.027 18.818 20.087 make_images 5068 14.6 3.096 3.293 18.325 19.832 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.403 19.536 apply_single 129 13.6 0.001 0.001 19.402 19.535 multiply_cannon_multrec 10136 15.6 10.357 12.193 18.149 18.202 cp_fm_diag_elpa 84 13.4 0.000 0.000 16.910 16.911 cp_fm_diag_elpa_base 84 14.4 12.440 14.058 16.907 16.907 qs_rho_update_rho_low 129 7.7 0.001 0.001 16.512 16.523 calculate_rho_elec 129 8.7 0.484 0.486 16.512 16.523 sum_up_and_integrate 129 10.3 0.002 0.002 16.094 16.185 integrate_v_rspace 129 11.3 0.004 0.005 16.031 16.123 multiply_cannon_sync_h2d 10136 15.6 14.450 14.484 14.450 14.484 hybrid_alltoall_any 5255 16.5 1.324 3.072 10.847 13.032 make_images_data 5068 15.6 0.065 0.069 10.671 12.903 init_scf_run 11 5.9 0.000 0.001 11.974 11.975 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.974 11.975 pw_transfer 1559 11.6 0.095 0.096 11.362 11.369 fft_wrap_pw1pw2 1301 12.7 0.012 0.012 11.123 11.129 fft_wrap_pw1pw2_140 527 13.2 3.065 3.114 9.914 9.922 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.854 9.875 dbcsr_mm_accdrv_process 20954 16.1 4.289 6.184 7.540 9.426 cp_fm_cholesky_decompose 22 10.9 8.902 8.938 8.902 8.938 wfi_extrapolate 11 7.9 0.001 0.001 8.870 8.870 grid_integrate_task_list 129 12.3 8.619 8.810 8.619 8.810 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 8.143 8.148 density_rs2pw 129 9.7 0.005 0.006 7.498 7.546 mp_alltoall_d11v 2429 14.1 6.845 6.965 6.845 6.965 calculate_dm_sparse 129 9.5 0.001 0.001 6.750 6.805 grid_collocate_task_list 129 9.7 6.460 6.544 6.460 6.544 fft3d_ps 1301 14.7 2.867 2.877 6.340 6.397 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.248 6.291 copy_dbcsr_to_fm 187 11.8 0.004 0.004 6.165 6.230 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=289.873000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2776.363636, yerr=177.304706 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5db0f86a89f1e2e120bbf3166966efaf01984442_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 1.260020E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255646. MP_Allreduce 3139 6114. MP_Sync 4 MP_Alltoall 54 15435264. 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 0.025 0.040 84.730 84.731 qs_energies 1 2.0 0.000 0.000 84.240 84.254 ls_scf 1 3.0 0.000 0.000 83.355 83.369 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.341 72.493 multiply_cannon 111 7.7 0.018 0.021 55.723 56.826 multiply_cannon_loop 111 8.7 0.227 0.244 52.377 53.875 ls_scf_main 1 4.0 0.000 0.000 52.183 52.184 density_matrix_trs4 2 5.0 0.002 0.003 46.611 46.682 ls_scf_init_scf 1 4.0 0.000 0.000 28.114 28.115 ls_scf_init_matrix_S 1 5.0 0.000 0.000 26.990 27.041 mp_waitall_1 11031 10.9 22.190 24.980 22.190 24.980 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 24.884 24.902 multiply_cannon_multrec 2664 9.7 8.170 8.942 15.543 17.283 multiply_cannon_sync_h2d 2664 9.7 13.657 15.972 13.657 15.972 make_m2s 222 7.7 0.009 0.012 13.019 13.475 make_images 222 8.7 0.099 0.108 12.997 13.454 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.609 12.741 make_images_data 222 9.7 0.004 0.005 7.557 8.076 dbcsr_mm_accdrv_process 4760 10.4 0.587 0.703 6.989 7.947 hybrid_alltoall_any 227 10.6 0.216 1.842 6.536 7.720 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.352 7.159 dbcsr_mm_accdrv_process_sort 4760 11.4 6.203 7.088 6.203 7.088 calculate_norms 4752 9.8 5.517 6.257 5.517 6.257 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.022 5.164 mp_sum_l 887 5.1 3.064 4.381 3.064 4.381 multiply_cannon_metrocomm4 2442 9.7 0.012 0.016 2.048 3.850 mp_irecv_dv 6231 10.9 2.031 3.815 2.031 3.815 make_images_sizes 222 9.7 0.000 0.000 0.671 3.700 mp_alltoall_i44 222 10.7 0.671 3.700 0.671 3.700 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.280 3.432 arnoldi_extremal 4 6.8 0.000 0.000 3.208 3.240 arnoldi_normal_ev 4 7.8 0.001 0.002 3.208 3.240 build_subspace 16 8.4 0.009 0.012 3.114 3.115 ls_scf_post 1 4.0 0.000 0.000 3.057 3.071 ls_scf_store_result 1 5.0 0.000 0.000 2.863 2.903 dbcsr_special_finalize 555 9.7 0.005 0.006 2.436 2.788 dbcsr_merge_single_wm 555 10.7 0.460 0.588 2.427 2.778 make_images_pack 222 9.7 2.207 2.623 2.208 2.625 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.319 2.561 dbcsr_sort_data 658 11.4 2.224 2.518 2.224 2.518 dbcsr_matrix_vector_mult_local 304 10.0 2.068 2.459 2.070 2.461 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.321 2.398 buffer_matrices_ensure_size 222 8.7 1.749 2.023 1.749 2.023 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.770 1.771 rebuild_ks_matrix 3 7.3 0.000 0.000 1.760 1.762 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.760 1.762 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=84.731000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1140.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5db0f86a89f1e2e120bbf3166966efaf01984442_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 2.139111E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266673. MP_Allreduce 3138 10075. MP_Sync 4 MP_Alltoall 47 15335933. MP_SendRecv 141 57600. MP_ISendRecv 141 57600. MP_Wait 687 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.023 0.039 89.957 89.958 qs_energies 1 2.0 0.000 0.000 89.502 89.506 ls_scf 1 3.0 0.000 0.000 88.176 88.179 dbcsr_multiply_generic 111 6.7 0.015 0.016 74.356 74.613 multiply_cannon 111 7.7 0.027 0.041 52.565 57.275 ls_scf_main 1 4.0 0.000 0.000 54.652 54.652 multiply_cannon_loop 111 8.7 0.135 0.145 49.798 53.260 density_matrix_trs4 2 5.0 0.002 0.003 48.969 49.224 ls_scf_init_scf 1 4.0 0.000 0.000 29.830 29.831 mp_waitall_1 9105 10.9 20.676 29.611 20.676 29.611 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.627 28.712 multiply_cannon_multrec 1332 9.7 13.410 17.059 22.682 27.489 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.222 26.235 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.352 20.579 make_m2s 222 7.7 0.006 0.007 15.226 15.755 make_images 222 8.7 1.365 1.696 15.195 15.725 dbcsr_mm_accdrv_process 4041 10.4 0.324 0.522 8.867 10.384 dbcsr_mm_accdrv_process_sort 4041 11.4 8.388 9.863 8.388 9.863 make_images_data 222 9.7 0.004 0.005 8.839 9.743 hybrid_alltoall_any 227 10.6 0.541 2.534 8.154 9.649 mp_sum_l 887 5.1 5.095 8.245 5.095 8.245 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.204 7.778 mp_irecv_dv 3311 11.0 3.183 7.723 3.183 7.723 calculate_norms 2376 9.8 6.062 6.670 6.062 6.670 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.849 6.523 multiply_cannon_sync_h2d 1332 9.7 4.877 6.278 4.877 6.278 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.976 5.205 arnoldi_extremal 4 6.8 0.000 0.000 4.635 4.660 arnoldi_normal_ev 4 7.8 0.001 0.004 4.635 4.659 build_subspace 16 8.4 0.014 0.021 4.387 4.389 ls_scf_post 1 4.0 0.000 0.000 3.694 3.699 ls_scf_store_result 1 5.0 0.000 0.000 3.413 3.519 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.156 3.374 dbcsr_matrix_vector_mult_local 304 10.0 2.751 3.232 2.753 3.234 mp_allgather_i34 111 8.7 0.758 3.031 0.758 3.031 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.591 2.696 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.132 2.652 dbcsr_data_new 4174 10.1 2.113 2.415 2.113 2.415 make_images_pack 222 9.7 1.838 2.123 1.841 2.126 dbcsr_sort_data 436 11.2 1.825 2.108 1.825 2.108 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.873 1.875 rebuild_ks_matrix 3 7.3 0.000 0.000 1.860 1.862 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.005 1.860 1.862 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=89.958000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1774.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5db0f86a89f1e2e120bbf3166966efaf01984442_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 2.896638E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265448. MP_Allreduce 3138 10896. MP_Sync 4 MP_Alltoall 47 23526250. MP_SendRecv 93 57600. MP_ISendRecv 93 57600. MP_Wait 639 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.034 0.051 91.796 91.797 qs_energies 1 2.0 0.004 0.029 91.212 91.238 ls_scf 1 3.0 0.000 0.001 89.790 89.794 dbcsr_multiply_generic 111 6.7 0.016 0.019 74.511 74.821 ls_scf_main 1 4.0 0.000 0.000 56.005 56.009 multiply_cannon 111 7.7 0.037 0.073 51.585 55.151 multiply_cannon_loop 111 8.7 0.117 0.131 48.796 53.067 density_matrix_trs4 2 5.0 0.002 0.003 50.044 50.290 mp_waitall_1 7281 11.0 23.164 33.515 23.164 33.515 ls_scf_init_scf 1 4.0 0.000 0.002 30.163 30.166 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.963 29.051 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.503 26.517 multiply_cannon_multrec 888 9.7 12.633 15.305 21.195 24.660 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.632 22.895 make_m2s 222 7.7 0.006 0.007 16.173 16.992 make_images 222 8.7 1.584 1.898 16.135 16.952 make_images_data 222 9.7 0.004 0.005 9.533 10.598 hybrid_alltoall_any 227 10.6 0.642 2.956 9.074 10.197 dbcsr_mm_accdrv_process 3754 10.4 0.303 0.492 8.081 9.345 dbcsr_mm_accdrv_process_sort 3754 11.4 7.648 8.854 7.648 8.854 mp_sum_l 887 5.1 5.060 8.397 5.060 8.397 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.490 7.290 mp_irecv_dv 2335 11.1 2.474 7.232 2.474 7.232 multiply_cannon_sync_h2d 888 9.7 6.037 7.204 6.037 7.204 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.642 6.930 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.906 6.498 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.896 5.122 arnoldi_extremal 4 6.8 0.000 0.000 5.065 5.083 arnoldi_normal_ev 4 7.8 0.001 0.005 5.065 5.083 build_subspace 16 8.4 0.014 0.020 4.768 4.774 calculate_norms 1584 9.8 4.337 4.762 4.337 4.762 mp_allgather_i34 111 8.7 0.900 3.784 0.900 3.784 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.442 3.766 ls_scf_post 1 4.0 0.000 0.000 3.622 3.627 dbcsr_matrix_vector_mult_local 304 10.0 3.038 3.609 3.041 3.611 ls_scf_store_result 1 5.0 0.000 0.000 3.396 3.446 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.886 3.010 dbcsr_data_new 4116 9.9 2.106 2.443 2.106 2.443 make_images_sizes 222 9.7 0.000 0.000 0.777 2.133 mp_alltoall_i44 222 10.7 0.776 2.132 0.776 2.132 dbcsr_sort_data 325 11.1 1.880 2.113 1.880 2.113 make_images_pack 222 9.7 1.648 1.955 1.651 1.958 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.887 1.889 rebuild_ks_matrix 3 7.3 0.000 0.000 1.868 1.870 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.868 1.870 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=91.797000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2314.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5db0f86a89f1e2e120bbf3166966efaf01984442_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 3.344200E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266673. MP_Allreduce 3138 13030. MP_Sync 4 MP_Alltoall 47 30278988. MP_SendRecv 69 86400. MP_ISendRecv 69 86400. MP_Wait 531 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.029 0.046 96.826 96.827 qs_energies 1 2.0 0.000 0.000 96.223 96.231 ls_scf 1 3.0 0.000 0.000 94.519 94.528 dbcsr_multiply_generic 111 6.7 0.017 0.017 78.113 78.353 ls_scf_main 1 4.0 0.000 0.000 58.709 58.709 multiply_cannon 111 7.7 0.044 0.094 51.536 55.779 density_matrix_trs4 2 5.0 0.002 0.003 52.615 52.735 multiply_cannon_loop 111 8.7 0.154 0.167 46.412 49.560 ls_scf_init_scf 1 4.0 0.000 0.000 32.572 32.574 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.379 31.435 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.886 28.896 mp_waitall_1 6369 11.0 22.308 28.852 22.308 28.852 multiply_cannon_multrec 1332 9.7 14.205 17.226 22.100 24.639 make_m2s 222 7.7 0.007 0.008 20.953 22.458 make_images 222 8.7 3.150 3.610 20.903 22.410 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.105 16.932 make_images_data 222 9.7 0.004 0.004 11.682 13.233 hybrid_alltoall_any 227 10.6 0.798 3.767 11.048 12.928 dbcsr_mm_accdrv_process 3641 10.4 0.286 0.479 7.530 9.061 dbcsr_mm_accdrv_process_sort 3641 11.4 7.090 8.577 7.090 8.577 mp_sum_l 887 5.1 4.037 7.426 4.037 7.426 multiply_cannon_sync_h2d 1332 9.7 5.451 6.059 5.451 6.059 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.092 6.028 mp_irecv_dv 3229 10.9 2.066 5.943 2.066 5.943 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.070 5.797 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.352 5.349 arnoldi_extremal 4 6.8 0.000 0.000 5.281 5.295 arnoldi_normal_ev 4 7.8 0.001 0.005 5.281 5.295 build_subspace 16 8.4 0.014 0.021 4.950 4.960 mp_allgather_i34 111 8.7 2.244 4.944 2.244 4.944 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.542 4.715 calculate_norms 2376 9.8 4.206 4.593 4.206 4.593 dbcsr_matrix_vector_mult 304 9.0 0.011 0.022 3.618 3.928 dbcsr_matrix_vector_mult_local 304 10.0 3.215 3.715 3.217 3.717 dbcsr_sort_data 658 11.4 3.073 3.422 3.073 3.422 ls_scf_post 1 4.0 0.000 0.000 3.238 3.245 dbcsr_special_finalize 555 9.7 0.006 0.007 2.816 3.148 dbcsr_merge_single_wm 555 10.7 0.543 0.664 2.807 3.140 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.014 3.065 ls_scf_store_result 1 5.0 0.000 0.000 2.963 3.041 dbcsr_data_release 10477 10.7 1.591 2.469 1.591 2.469 dbcsr_finalize 304 7.8 0.049 0.061 1.812 2.019 make_images_pack 222 9.7 1.607 1.937 1.609 1.940 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=96.827000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2711.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5db0f86a89f1e2e120bbf3166966efaf01984442_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 4.757361E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265536. MP_Allreduce 3129 15263. MP_Sync 4 MP_Alltoall 47 46208988. MP_SendRecv 45 115200. MP_ISendRecv 45 115200. MP_Wait 528 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.041 0.059 96.216 96.217 qs_energies 1 2.0 0.000 0.000 95.233 95.238 ls_scf 1 3.0 0.000 0.000 93.295 93.299 dbcsr_multiply_generic 111 6.7 0.018 0.019 74.782 74.991 multiply_cannon 111 7.7 0.081 0.131 56.575 59.367 ls_scf_main 1 4.0 0.000 0.000 56.462 56.462 multiply_cannon_loop 111 8.7 0.088 0.094 53.961 55.065 density_matrix_trs4 2 5.0 0.002 0.003 49.476 49.556 ls_scf_init_scf 1 4.0 0.000 0.000 33.438 33.440 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.188 32.228 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 30.101 30.114 mp_waitall_1 5436 11.0 24.274 29.020 24.274 29.020 multiply_cannon_multrec 444 9.7 13.731 16.440 20.790 22.265 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.586 15.386 make_m2s 222 7.7 0.005 0.005 13.497 14.446 make_images 222 8.7 2.039 2.479 13.430 14.378 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.269 14.372 make_images_data 222 9.7 0.003 0.004 8.234 9.752 hybrid_alltoall_any 227 10.6 0.802 3.830 8.147 9.463 multiply_cannon_sync_h2d 444 9.7 6.754 8.251 6.754 8.251 dbcsr_mm_accdrv_process 3003 10.4 0.363 0.400 6.755 7.859 dbcsr_mm_accdrv_process_sort 3003 11.4 6.392 7.464 6.392 7.464 arnoldi_extremal 4 6.8 0.000 0.000 5.898 5.910 arnoldi_normal_ev 4 7.8 0.001 0.004 5.898 5.910 build_subspace 16 8.4 0.015 0.020 5.501 5.512 compute_matrix_preconditioner 1 6.0 0.001 0.002 4.706 4.714 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.427 4.702 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.221 4.394 acc_transpose_blocks 444 9.7 0.003 0.003 4.272 4.277 acc_transpose_blocks_kernels 444 10.7 0.006 0.007 4.207 4.211 jit_kernel_transpose 1 13.0 4.201 4.205 4.201 4.205 dbcsr_matrix_vector_mult_local 304 10.0 3.734 4.194 3.736 4.196 mp_sum_l 887 5.1 2.698 4.189 2.698 4.189 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.531 3.839 mp_irecv_dv 1241 11.2 1.515 3.795 1.515 3.795 calculate_norms 792 9.8 3.616 3.733 3.616 3.733 mp_allgather_i34 111 8.7 1.198 3.641 1.198 3.641 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.529 3.597 ls_scf_post 1 4.0 0.000 0.000 3.395 3.399 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.904 3.327 make_images_sizes 222 9.7 0.000 0.000 0.882 3.294 mp_alltoall_i44 222 10.7 0.882 3.294 0.882 3.294 ls_scf_store_result 1 5.0 0.000 0.000 3.175 3.219 dbcsr_finalize 304 7.8 0.062 0.077 2.199 2.287 dbcsr_data_new 4608 9.7 1.777 2.185 1.777 2.185 dbcsr_merge_all 275 8.9 0.478 0.528 2.057 2.129 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.028 2.029 rebuild_ks_matrix 3 7.3 0.000 0.000 1.995 1.996 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.000 1.995 1.996 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=96.217000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3744.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5db0f86a89f1e2e120bbf3166966efaf01984442_performance_tests/26/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 23 x 23 x 23 234439235724792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 234.439236E+12 0.0% 0.0% 100.0% flops max/rank 30.358840E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 9634225188 0.0% 0.0% 100.0% number of processed stacks 339931 0.0% 0.0% 100.0% average stack size 0.0 0.0 28341.7 marketing flops 1.742118E+15 ------------------------------------------------------------------------------- # multiplications 111 max memory usage/rank 8.792408E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284089. MP_Allreduce 3123 21388. MP_Sync 4 MP_Alltoall 47 88727262. MP_SendRecv 42 732600. MP_ISendRecv 42 732600. MP_Wait 267 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.075 0.092 105.783 105.783 qs_energies 1 2.0 0.000 0.000 104.348 104.353 ls_scf 1 3.0 0.000 0.000 101.406 101.410 dbcsr_multiply_generic 111 6.7 0.024 0.026 74.969 75.116 ls_scf_main 1 4.0 0.000 0.000 63.969 63.969 density_matrix_trs4 2 5.0 0.002 0.003 55.003 55.061 multiply_cannon 111 7.7 0.143 0.262 48.407 50.537 multiply_cannon_loop 111 8.7 0.097 0.099 45.461 46.098 ls_scf_init_scf 1 4.0 0.000 0.000 33.657 33.658 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.151 32.174 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.461 29.469 mp_waitall_1 4527 11.1 22.053 25.599 22.053 25.599 make_m2s 222 7.7 0.005 0.005 22.782 23.825 make_images 222 8.7 3.584 3.874 22.675 23.716 multiply_cannon_multrec 444 9.7 17.815 18.480 22.443 23.158 hybrid_alltoall_any 227 10.6 1.656 3.612 12.920 15.907 make_images_data 222 9.7 0.003 0.004 13.171 15.278 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.294 10.644 multiply_cannon_sync_h2d 444 9.7 8.796 8.841 8.796 8.841 arnoldi_extremal 4 6.8 0.000 0.000 7.558 7.567 arnoldi_normal_ev 4 7.8 0.002 0.008 7.558 7.567 build_subspace 16 8.4 0.026 0.036 7.010 7.021 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.593 5.752 dbcsr_matrix_vector_mult_local 304 10.0 5.138 5.453 5.141 5.455 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.087 5.184 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.855 5.117 dbcsr_mm_accdrv_process 1814 10.4 0.242 0.354 4.441 4.576 dbcsr_mm_accdrv_process_sort 1814 11.4 4.108 4.245 4.108 4.245 ls_scf_post 1 4.0 0.000 0.000 3.781 3.786 make_images_sizes 222 9.7 0.000 0.000 1.475 3.670 mp_alltoall_i44 222 10.7 1.475 3.670 1.475 3.670 ls_scf_store_result 1 5.0 0.000 0.000 3.503 3.543 mp_allgather_i34 111 8.7 1.061 3.357 1.061 3.357 calculate_norms 792 9.8 3.227 3.277 3.227 3.277 dbcsr_finalize 304 7.8 0.082 0.089 3.072 3.159 dbcsr_merge_all 275 8.9 0.885 0.918 2.857 2.938 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.912 2.912 dbcsr_complete_redistribute 5 7.6 1.446 1.487 2.754 2.875 matrix_ls_to_qs 2 6.0 0.000 0.000 2.407 2.536 dbcsr_sort_data 325 11.1 2.440 2.495 2.440 2.495 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.391 2.393 dbcsr_data_new 6591 9.6 1.863 2.353 1.863 2.353 rebuild_ks_matrix 3 7.3 0.000 0.000 2.324 2.326 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.324 2.326 dbcsr_new_transposed 4 7.5 0.240 0.250 2.280 2.294 dbcsr_frobenius_norm 74 6.6 2.057 2.136 2.208 2.246 dbcsr_add_d 103 6.2 0.000 0.000 2.125 2.200 dbcsr_add_anytype 103 7.2 0.859 0.891 2.125 2.200 dbcsr_data_release 12724 10.6 1.975 2.179 1.975 2.179 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=105.783000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=7029.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5db0f86a89f1e2e120bbf3166966efaf01984442_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 595.742720E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 102097920 MPI messages size (bytes): total size 37.227590E+12 min size 0.000000E+00 max size 4.551360E+06 average size 364.626312E+03 MPI breakdown and total messages size (bytes): size <= 128 731472 0 128 < size <= 8192 11922720 97670922240 8192 < size <= 32768 24718992 614677610496 32768 < size <= 131072 20000256 1970081366016 131072 < size <= 4194304 42515668 24886801223040 4194304 < size <= 16777216 2208812 9656099886720 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4640 78072. MP_Allreduce 13232 2081. MP_Sync 1064 MP_Alltoall 2588 4805008. 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 0.031 0.162 227.979 227.980 qs_mol_dyn_low 1 2.0 0.003 0.030 226.827 226.881 qs_forces 5 3.8 0.004 0.005 226.721 226.722 qs_energies 5 4.8 0.002 0.028 223.769 223.798 scf_env_do_scf 5 5.8 0.000 0.001 208.721 208.724 scf_env_do_scf_inner_loop 105 6.6 0.002 0.006 182.287 182.290 qs_scf_new_mos 105 7.6 0.000 0.001 142.629 142.782 qs_scf_loop_do_ot 105 8.6 0.001 0.001 142.629 142.781 dbcsr_multiply_generic 1445 12.2 0.127 0.137 133.875 134.459 ot_scf_mini 105 9.6 0.003 0.004 132.601 132.723 multiply_cannon 1445 13.2 0.276 0.287 114.951 117.387 multiply_cannon_loop 1445 14.2 2.731 2.890 113.224 114.298 velocity_verlet 4 3.0 0.005 0.020 106.746 106.746 ot_mini 105 10.6 0.001 0.001 60.104 60.201 multiply_cannon_multrec 69360 15.2 30.072 35.286 39.460 44.539 qs_ot_get_p 112 10.4 0.001 0.001 42.427 42.729 mp_waitall_1 488190 16.1 34.813 42.023 34.813 42.023 qs_ot_get_derivative 55 11.6 0.001 0.001 38.262 38.386 multiply_cannon_sync_h2d 69360 15.2 29.543 33.759 29.543 33.759 multiply_cannon_metrocomm3 69360 15.2 0.203 0.213 25.809 33.405 qs_ot_p2m_diag 40 11.0 0.020 0.030 31.226 31.324 rebuild_ks_matrix 110 8.4 0.000 0.000 28.835 29.018 qs_ks_build_kohn_sham_matrix 110 9.4 0.011 0.013 28.834 29.018 cp_dbcsr_syevd 40 12.0 0.002 0.002 28.187 28.187 qs_ks_update_qs_env 112 7.6 0.001 0.001 26.465 26.631 init_scf_loop 7 6.6 0.000 0.004 26.404 26.406 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 23.119 23.406 apply_single 62 13.6 0.000 0.000 23.119 23.405 cp_fm_syevd 40 13.0 0.000 0.001 23.059 23.225 prepare_preconditioner 7 7.6 0.000 0.000 21.531 21.562 make_preconditioner 7 8.6 0.000 0.004 21.531 21.562 ot_new_cg_direction 55 11.6 0.001 0.001 21.118 21.118 cp_fm_redistribute_end 40 14.0 9.180 18.313 9.184 18.314 cp_fm_syevd_base 40 14.0 9.122 18.260 9.122 18.260 qs_rho_update_rho_low 110 7.6 0.001 0.001 16.519 16.894 calculate_rho_elec 110 8.6 0.030 0.032 16.518 16.894 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.851 15.035 make_full_inverse_cholesky 7 9.6 0.000 0.000 14.395 14.453 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 13.988 14.087 mp_sum_l 4764 12.2 12.098 12.762 12.098 12.762 init_scf_run 5 5.8 0.000 0.001 12.433 12.434 scf_env_initial_rho_setup 5 6.8 0.001 0.003 12.433 12.434 pw_transfer 1645 12.4 0.079 0.101 11.748 11.996 fft_wrap_pw1pw2 1425 13.5 0.013 0.016 11.609 11.862 calculate_dm_sparse 110 9.5 0.000 0.001 11.584 11.786 density_rs2pw 110 9.6 0.006 0.007 10.775 11.263 qs_vxc_create 110 10.4 0.008 0.037 10.524 10.572 fft_wrap_pw1pw2_240 915 15.0 1.144 1.255 10.221 10.460 dbcsr_mm_accdrv_process 154766 15.8 5.784 6.151 9.255 10.243 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 10.086 10.165 cp_fm_cholesky_invert 7 10.6 10.094 10.102 10.094 10.102 check_diag 80 13.5 8.690 8.949 9.538 9.671 multiply_cannon_metrocomm1 69360 15.2 0.097 0.104 4.661 9.152 fft3d_pb 915 16.0 2.394 2.681 8.420 8.697 acc_transpose_blocks 69360 15.2 0.334 0.363 7.888 8.458 sum_up_and_integrate 60 10.3 0.001 0.002 8.290 8.301 integrate_v_rspace 60 11.3 0.001 0.002 8.273 8.284 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.001 8.014 8.031 transfer_rs2pw 445 10.6 0.007 0.008 7.460 7.995 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.001 7.623 7.676 xc_rho_set_and_dset_create 110 12.4 0.079 0.126 7.368 7.629 calculate_first_density_matrix 1 7.0 0.000 0.002 7.600 7.615 xc_vxc_pw_create 60 11.3 0.038 0.052 7.035 7.089 make_m2s 2890 13.2 0.078 0.088 6.362 6.922 make_full_single_inverse 7 9.6 0.001 0.005 6.870 6.902 make_images 2890 14.2 0.240 0.271 6.255 6.816 xc_pw_derive 510 13.4 0.006 0.007 6.255 6.350 acc_transpose_blocks_kernels 69360 16.2 0.809 0.873 5.328 5.722 mp_alltoall_z22v 2340 17.7 5.441 5.708 5.441 5.708 mp_waitany 7680 13.5 4.392 5.010 4.392 5.010 jit_kernel_transpose 5 15.0 4.519 4.891 4.519 4.891 multiply_cannon_metrocomm4 67915 15.2 0.191 0.210 2.021 4.748 potential_pw2rs 60 12.3 0.003 0.003 4.543 4.561 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=227.980000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=564.400000, yerr=4.029888 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 5db0f86a89f1e2e120bbf3166966efaf01984442 Summary: empty Status: OK