=== 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: a374e79360e402e0e0c3cfd3175b21c36af7649d ################# 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 2.6.2, ELPA 2022.05.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.0.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.0, # SIRIUS 7.3.2, 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. # # Author: Matthias Krack (26.10.2022) # # \ 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 --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2022.05.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.0.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.0 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.3.2 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MPI_VERSION=3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta_prefixed_scalapack.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/a374e79360e402e0e0c3cfd3175b21c36af7649d_performance_tests/01 job id: 42863080 --- 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/a374e79360e402e0e0c3cfd3175b21c36af7649d_performance_tests/02 job id: 42863081 --- 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/a374e79360e402e0e0c3cfd3175b21c36af7649d_performance_tests/03 job id: 42863082 --- 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/a374e79360e402e0e0c3cfd3175b21c36af7649d_performance_tests/04 job id: 42863083 --- 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/a374e79360e402e0e0c3cfd3175b21c36af7649d_performance_tests/05 job id: 42863084 --- 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/a374e79360e402e0e0c3cfd3175b21c36af7649d_performance_tests/06 job id: 42863086 --- 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/a374e79360e402e0e0c3cfd3175b21c36af7649d_performance_tests/07 job id: 42863087 --- 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/a374e79360e402e0e0c3cfd3175b21c36af7649d_performance_tests/08 job id: 42863088 --- 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/a374e79360e402e0e0c3cfd3175b21c36af7649d_performance_tests/09 job id: 42863089 --- 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/a374e79360e402e0e0c3cfd3175b21c36af7649d_performance_tests/10 job id: 42863090 --- 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/a374e79360e402e0e0c3cfd3175b21c36af7649d_performance_tests/11 job id: 42863091 --- 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/a374e79360e402e0e0c3cfd3175b21c36af7649d_performance_tests/12 job id: 42863092 --- 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/a374e79360e402e0e0c3cfd3175b21c36af7649d_performance_tests/13 job id: 42863094 --- 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/a374e79360e402e0e0c3cfd3175b21c36af7649d_performance_tests/14 job id: 42863095 --- 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/a374e79360e402e0e0c3cfd3175b21c36af7649d_performance_tests/15 job id: 42863096 --- 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/a374e79360e402e0e0c3cfd3175b21c36af7649d_performance_tests/16 job id: 42863097 --- 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/a374e79360e402e0e0c3cfd3175b21c36af7649d_performance_tests/17 job id: 42863098 --- 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/a374e79360e402e0e0c3cfd3175b21c36af7649d_performance_tests/18 job id: 42863099 --- 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/a374e79360e402e0e0c3cfd3175b21c36af7649d_performance_tests/19 job id: 42863100 --- 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/a374e79360e402e0e0c3cfd3175b21c36af7649d_performance_tests/20 job id: 42863101 --- 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/a374e79360e402e0e0c3cfd3175b21c36af7649d_performance_tests/21 job id: 42863103 --- 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/a374e79360e402e0e0c3cfd3175b21c36af7649d_performance_tests/22 job id: 42863104 --- 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/a374e79360e402e0e0c3cfd3175b21c36af7649d_performance_tests/23 job id: 42863105 --- 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/a374e79360e402e0e0c3cfd3175b21c36af7649d_performance_tests/24 job id: 42863106 --- 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/a374e79360e402e0e0c3cfd3175b21c36af7649d_performance_tests/25 job id: 42863108 --- 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/a374e79360e402e0e0c3cfd3175b21c36af7649d_performance_tests/26 job id: 42863109 --- 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 ~~~~~~~ === 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]" === END PLOTS (description) === ============ RESULTS ============ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a374e79360e402e0e0c3cfd3175b21c36af7649d_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 344 9. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.024 0.036 139.293 139.293 farming_run 1 2.0 138.706 138.708 139.254 139.258 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.460097E+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 228 1113141. MP_Allreduce 489 2263609. MP_Sync 27 MP_Alltoall 38 9316958. MP_SendRecv 30 829726. MP_ISendRecv 135 235435. MP_Wait 281 MP_comm_split 8 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 0.013 0.031 117.217 117.218 qs_energies 1 2.0 0.000 0.000 116.948 116.949 mp2_main 1 3.0 0.000 0.000 114.126 114.128 mp2_gpw_main 1 4.0 0.019 0.025 113.022 113.024 mp2_ri_gpw_compute_in 1 5.0 0.171 0.173 93.759 94.196 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 55.453 55.891 mp2_eri_3c_integrate_gpw 272 7.0 0.154 0.168 41.772 47.170 get_2c_integrals 1 6.0 0.000 0.000 37.323 38.134 integrate_v_rspace 273 8.0 0.435 0.448 25.108 30.206 pw_transfer 6555 10.6 0.374 0.400 27.330 27.844 fft_wrap_pw1pw2 5465 11.4 0.044 0.045 25.973 26.551 grid_integrate_task_list 273 9.0 20.924 26.503 20.924 26.503 fft_wrap_pw1pw2_100 2178 12.4 1.179 1.358 23.512 24.101 compute_2c_integrals 1 7.0 0.002 0.002 19.721 19.722 rpa_ri_compute_en 1 5.0 0.000 0.000 19.153 19.337 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.867 19.323 mp2_eri_2c_integrate_gpw 1 9.0 2.393 2.442 18.864 19.320 cp_fm_cholesky_decompose 12 8.2 17.745 18.567 17.745 18.567 cholesky_decomp 1 7.0 0.000 0.000 16.450 17.267 fft3d_s 5443 13.4 16.100 16.581 16.122 16.602 ao_to_mo_and_store_B_mult_1 272 7.0 10.853 15.570 10.853 15.570 calculate_wavefunction 272 8.0 5.471 5.631 12.582 13.190 rpa_num_int 1 6.0 0.000 0.001 10.802 10.802 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.731 10.733 calc_mat_Q 8 8.0 0.000 0.000 9.399 9.491 contract_S_to_Q 8 9.0 0.000 0.000 8.820 8.913 calc_potential_gpw 544 9.5 0.005 0.006 8.275 8.653 parallel_gemm_fm 14 9.1 0.000 0.000 8.405 8.500 parallel_gemm_fm_cosma 14 10.1 8.405 8.500 8.405 8.500 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.207 8.496 potential_pw2rs 545 10.0 0.107 0.109 7.662 8.270 create_integ_mat 1 6.0 0.022 0.027 7.818 7.818 collocate_single_gaussian 272 10.0 0.040 0.043 7.448 7.716 array2fm 1 7.0 0.000 0.000 6.704 7.154 pw_scatter_s 2720 13.7 4.430 4.642 4.430 4.642 pw_gather_s 2722 13.2 3.871 4.211 3.871 4.211 array2fm_buffer_send 1 8.0 2.932 3.148 2.932 3.148 scf_env_do_scf 1 3.0 0.000 0.000 2.529 2.529 pw_poisson_solve 545 10.5 1.111 1.158 2.174 2.394 init_scf_loop 1 4.0 0.000 0.000 2.393 2.393 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=113.023565, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2732.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a374e79360e402e0e0c3cfd3175b21c36af7649d_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 344 10. 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 0.044 0.059 399.440 399.441 farming_run 1 2.0 397.953 397.978 399.369 399.377 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.223008E+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 703 408373. MP_Allreduce 1825 23678. MP_Sync 38 MP_Alltoall 77 17043751. MP_SendRecv 2171 2843495. MP_ISendRecv 1739 144022. MP_Wait 2051 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 0.018 0.051 210.344 210.345 qs_energies 1 2.0 0.000 0.000 210.076 210.083 scf_env_do_scf 1 3.0 0.000 0.000 106.373 106.373 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.360 105.368 rebuild_ks_matrix 4 6.0 0.000 0.000 105.359 105.367 qs_ks_build_kohn_sham_matrix 4 7.0 0.056 0.064 105.359 105.367 hfx_ks_matrix 4 8.0 0.001 0.001 104.938 104.941 integrate_four_center 4 9.0 0.143 0.453 104.937 104.941 mp2_main 1 3.0 0.005 0.023 103.384 103.391 mp2_gpw_main 1 4.0 0.043 0.079 102.182 102.195 integrate_four_center_main 4 10.0 0.129 0.493 96.460 99.245 integrate_four_center_bin 264 11.0 96.331 98.979 96.331 98.979 init_scf_loop 1 4.0 0.000 0.000 92.065 92.065 mp2_ri_gpw_compute_in 1 5.0 0.088 0.142 74.907 75.940 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.436 55.470 mp2_eri_3c_integrate_gpw 91 7.0 0.143 0.155 42.176 47.220 integrate_v_rspace 95 8.0 0.397 0.571 28.585 33.450 pw_transfer 2240 10.6 0.144 0.164 29.913 30.293 fft_wrap_pw1pw2 1868 11.4 0.017 0.020 28.926 29.321 ao_to_mo_and_store_B_mult_1 91 7.0 10.574 29.220 10.574 29.220 grid_integrate_task_list 95 9.0 23.846 28.925 23.846 28.925 mp2_ri_gpw_compute_en 1 5.0 0.056 0.065 27.041 28.769 fft_wrap_pw1pw2_100 730 12.4 1.263 1.417 26.644 27.048 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.851 1.896 25.372 25.382 get_2c_integrals 1 6.0 0.017 0.054 20.320 20.404 compute_2c_integrals 1 7.0 0.018 0.047 19.273 19.286 compute_2c_integrals_loop_lm 1 8.0 0.009 0.022 18.859 19.162 mp2_eri_2c_integrate_gpw 1 9.0 1.743 1.856 18.849 19.160 fft3d_s 1823 13.4 18.420 18.761 18.433 18.775 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.304 14.305 calculate_wavefunction 91 8.0 2.027 2.054 9.750 9.967 potential_pw2rs 186 10.0 0.033 0.035 8.658 9.174 mp2_ri_gpw_compute_en_expansio 172 7.0 0.557 0.592 8.720 9.173 mp2_ri_gpw_compute_en_comm 22 7.0 0.480 0.503 8.405 8.742 local_gemm 172 8.0 8.164 8.627 8.164 8.627 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.220 8.569 collocate_single_gaussian 91 10.0 0.017 0.025 7.885 8.132 calc_potential_gpw 182 9.5 0.002 0.002 7.891 8.121 mp_sendrecv_dm3 2068 8.0 6.491 6.826 6.491 6.826 mp2_ri_gpw_compute_en_ener 172 7.0 6.351 6.422 6.351 6.422 mp_sync 38 10.4 3.159 5.896 3.159 5.896 pw_gather_s 912 13.2 4.924 5.399 4.924 5.399 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=102.180107, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1510.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a374e79360e402e0e0c3cfd3175b21c36af7649d_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 451.325952E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62385. MP_Allreduce 10249 271. MP_Sync 580 MP_Alltoall 2083 MP_ISendRecv 45220 5520. MP_Wait 60486 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 0.168 0.268 53.100 53.113 qs_mol_dyn_low 1 2.0 0.003 0.004 52.290 52.324 qs_forces 11 3.9 0.002 0.002 52.158 52.159 qs_energies 11 4.9 0.032 0.056 50.650 50.665 scf_env_do_scf 11 5.9 0.001 0.001 44.253 44.253 scf_env_do_scf_inner_loop 108 6.5 0.002 0.009 42.158 42.159 dbcsr_multiply_generic 2286 12.5 0.095 0.107 33.788 34.258 qs_scf_new_mos 108 7.5 0.000 0.000 32.084 32.361 qs_scf_loop_do_ot 108 8.5 0.001 0.001 32.083 32.360 ot_scf_mini 108 9.5 0.002 0.002 30.436 30.621 multiply_cannon 2286 13.5 0.182 0.192 26.249 27.898 multiply_cannon_loop 2286 14.5 1.493 1.573 25.524 27.226 velocity_verlet 10 3.0 0.011 0.079 25.765 25.765 ot_mini 108 10.5 0.001 0.001 19.667 19.931 qs_ot_get_derivative 108 11.5 0.001 0.001 16.628 16.827 mp_waitall_1 267858 16.1 9.138 14.932 9.138 14.932 multiply_cannon_metrocomm3 54864 15.5 0.068 0.073 6.176 13.009 multiply_cannon_multrec 54864 15.5 4.216 6.591 7.568 11.107 rebuild_ks_matrix 119 8.3 0.000 0.000 7.938 8.067 qs_ks_build_kohn_sham_matrix 119 9.3 0.011 0.012 7.937 8.067 multiply_cannon_sync_h2d 54864 15.5 5.848 7.171 5.848 7.171 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.997 7.113 mp_sum_l 7207 12.9 5.417 6.982 5.417 6.982 qs_ot_get_p 119 10.4 0.001 0.001 6.161 6.429 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.649 6.098 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.750 5.863 init_scf_run 11 5.9 0.000 0.001 4.954 4.955 scf_env_initial_rho_setup 11 6.9 0.000 0.001 4.954 4.954 dbcsr_mm_accdrv_process 76910 16.1 1.159 1.837 3.276 4.609 sum_up_and_integrate 119 10.3 0.012 0.015 4.508 4.518 integrate_v_rspace 119 11.3 0.002 0.002 4.496 4.507 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.127 4.258 calculate_rho_elec 119 8.7 0.011 0.016 4.126 4.257 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.448 3.500 multiply_cannon_metrocomm1 54864 15.5 0.053 0.058 2.028 3.395 calculate_dm_sparse 119 9.5 0.000 0.000 2.980 3.114 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.900 3.093 apply_single 119 13.6 0.000 0.000 2.900 3.093 rs_pw_transfer 974 11.9 0.011 0.013 2.799 2.895 calculate_first_density_matrix 1 7.0 0.006 0.029 2.811 2.815 ot_diis_step 108 11.5 0.006 0.006 2.751 2.751 jit_kernel_multiply 13 15.8 2.056 2.687 2.056 2.687 cp_dbcsr_syevd 50 12.0 0.002 0.003 2.620 2.620 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.428 2.501 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.411 2.412 cp_fm_redistribute_end 50 14.0 2.190 2.390 2.194 2.392 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.383 2.386 cp_fm_diag_elpa_base 50 14.0 0.195 2.345 0.196 2.354 density_rs2pw 119 9.7 0.004 0.004 2.180 2.291 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.107 2.159 acc_transpose_blocks 54864 15.5 0.223 0.241 1.703 2.147 grid_integrate_task_list 119 12.3 2.007 2.086 2.007 2.086 wfi_extrapolate 11 7.9 0.001 0.001 2.083 2.083 init_scf_loop 11 6.9 0.002 0.013 2.076 2.078 mp_sum_d 4125 12.0 1.477 2.027 1.477 2.027 potential_pw2rs 119 12.3 0.004 0.004 1.885 1.902 make_m2s 4572 13.5 0.055 0.057 1.853 1.900 make_images 4572 14.5 0.134 0.141 1.768 1.816 pw_transfer 1439 11.6 0.051 0.056 1.689 1.763 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.614 1.690 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.498 1.520 mp_alltoall_d11v 2130 13.8 1.375 1.516 1.375 1.516 fft3d_ps 1201 14.6 0.359 0.460 1.392 1.461 mp_waitany 12084 13.8 1.287 1.460 1.287 1.460 grid_collocate_task_list 119 9.7 1.287 1.350 1.287 1.350 fft_wrap_pw1pw2_140 487 13.2 0.080 0.093 1.245 1.323 dbcsr_dot_sd 1205 11.9 0.047 0.057 0.836 1.212 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=53.113000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=430.090909, yerr=1.239835 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a374e79360e402e0e0c3cfd3175b21c36af7649d_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 487.469056E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2194560 MPI messages size (bytes): total size 310.646604E+09 min size 0.000000E+00 max size 1.145520E+06 average size 141.553031E+03 MPI breakdown and total messages size (bytes): size <= 128 724648 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 281952 4619501568 32768 < size <= 131072 494448 39143342080 131072 < size <= 4194304 440000 264807943488 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62664. MP_Allreduce 10226 305. MP_Sync 104 MP_Alltoall 2060 1031941. MP_ISendRecv 33558 37093. MP_Wait 40318 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 0.025 0.057 40.030 40.031 qs_mol_dyn_low 1 2.0 0.003 0.003 39.722 39.745 qs_forces 11 3.9 0.002 0.002 39.663 39.664 qs_energies 11 4.9 0.007 0.048 37.978 37.982 scf_env_do_scf 11 5.9 0.000 0.001 32.512 32.513 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 29.986 29.986 dbcsr_multiply_generic 2286 12.5 0.102 0.112 22.080 22.490 qs_scf_new_mos 108 7.5 0.001 0.001 20.731 20.967 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.730 20.967 ot_scf_mini 108 9.5 0.002 0.003 19.813 19.980 velocity_verlet 10 3.0 0.001 0.001 18.306 18.307 multiply_cannon 2286 13.5 0.209 0.217 16.714 18.225 multiply_cannon_loop 2286 14.5 0.906 0.986 15.479 16.919 ot_mini 108 10.5 0.001 0.001 12.306 12.541 mp_waitall_1 217478 16.2 6.497 11.762 6.497 11.762 qs_ot_get_derivative 108 11.5 0.001 0.001 9.563 9.730 multiply_cannon_metrocomm3 27432 15.5 0.068 0.070 4.348 9.624 multiply_cannon_multrec 27432 15.5 1.969 4.533 5.863 8.757 rebuild_ks_matrix 119 8.3 0.000 0.000 7.313 7.466 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.015 7.313 7.466 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.484 6.624 dbcsr_mm_accdrv_process 47894 16.0 2.975 5.156 3.825 5.683 qs_ot_get_p 119 10.4 0.001 0.002 4.440 4.673 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.680 4.522 sum_up_and_integrate 119 10.3 0.024 0.027 4.283 4.289 integrate_v_rspace 119 11.3 0.002 0.002 4.259 4.265 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.219 4.263 apply_single 119 13.6 0.000 0.000 3.219 4.263 mp_sum_l 7207 12.9 2.186 4.138 2.186 4.138 init_scf_run 11 5.9 0.000 0.001 4.129 4.129 scf_env_initial_rho_setup 11 6.9 0.000 0.001 4.128 4.129 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.764 3.799 calculate_rho_elec 119 8.7 0.021 0.024 3.763 3.799 make_m2s 4572 13.5 0.055 0.070 2.743 3.012 make_images 4572 14.5 0.199 0.235 2.651 2.919 qs_ot_p2m_diag 50 11.0 0.008 0.012 2.891 2.910 rs_pw_transfer 974 11.9 0.009 0.011 2.706 2.815 multiply_cannon_sync_h2d 27432 15.5 2.179 2.796 2.179 2.796 ot_diis_step 108 11.5 0.011 0.011 2.679 2.680 calculate_first_density_matrix 1 7.0 0.000 0.003 2.610 2.613 init_scf_loop 11 6.9 0.000 0.002 2.502 2.503 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.447 2.447 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.222 2.312 calculate_dm_sparse 119 9.5 0.000 0.000 2.133 2.216 density_rs2pw 119 9.7 0.004 0.004 2.093 2.203 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.069 2.072 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.035 2.042 potential_pw2rs 119 12.3 0.006 0.006 2.009 2.026 cp_fm_redistribute_end 50 14.0 1.688 2.013 1.693 2.016 grid_integrate_task_list 119 12.3 1.836 1.943 1.836 1.943 pw_transfer 1439 11.6 0.063 0.067 1.881 1.922 cp_fm_diag_elpa_base 50 14.0 0.307 1.857 0.315 1.888 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.791 1.835 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.705 1.745 make_images_data 4572 15.5 0.045 0.051 1.244 1.738 jit_kernel_multiply 8 16.0 0.798 1.615 0.798 1.615 prepare_preconditioner 11 7.9 0.000 0.000 1.567 1.593 make_preconditioner 11 8.9 0.000 0.000 1.567 1.593 acc_transpose_blocks 27432 15.5 0.107 0.111 1.245 1.563 hybrid_alltoall_any 4725 16.4 0.051 0.112 1.067 1.549 fft3d_ps 1201 14.6 0.500 0.553 1.500 1.541 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.468 1.524 mp_allgather_i34 2286 14.5 0.675 1.487 0.675 1.487 wfi_extrapolate 11 7.9 0.001 0.001 1.461 1.461 mp_alltoall_d11v 2130 13.8 1.272 1.385 1.272 1.385 fft_wrap_pw1pw2_140 487 13.2 0.075 0.082 1.344 1.384 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.328 1.336 mp_sum_d 4125 12.0 0.768 1.331 0.768 1.331 grid_collocate_task_list 119 9.7 1.242 1.306 1.242 1.306 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.196 1.246 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 0.971 0.972 acc_transpose_blocks_kernels 27432 16.5 0.179 0.265 0.718 0.951 make_images_sizes 4572 15.5 0.005 0.005 0.681 0.946 mp_alltoall_i44 4572 16.5 0.676 0.941 0.676 0.941 dbcsr_dot_sd 1205 11.9 0.071 0.086 0.510 0.938 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.889 0.904 mp_alltoall_z22v 1201 16.6 0.755 0.845 0.755 0.845 rs_pw_transfer_PW2RS_50 119 14.3 0.589 0.609 0.781 0.838 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=40.031000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=463.727273, yerr=1.958242 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a374e79360e402e0e0c3cfd3175b21c36af7649d_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 521.719808E+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 3672 62660. MP_Allreduce 10225 303. MP_Sync 104 MP_Alltoall 1821 1607811. MP_ISendRecv 22134 57667. MP_Wait 33054 MP_comm_split 50 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 0.101 0.208 34.512 34.513 qs_mol_dyn_low 1 2.0 0.003 0.005 33.950 33.966 qs_forces 11 3.9 0.002 0.002 33.886 33.887 qs_energies 11 4.9 0.012 0.023 32.285 32.297 scf_env_do_scf 11 5.9 0.001 0.002 26.931 26.932 scf_env_do_scf_inner_loop 108 6.5 0.003 0.008 24.226 24.226 dbcsr_multiply_generic 2286 12.5 0.095 0.101 16.984 17.066 qs_scf_new_mos 108 7.5 0.001 0.001 15.791 15.809 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.791 15.808 velocity_verlet 10 3.0 0.003 0.010 15.680 15.681 ot_scf_mini 108 9.5 0.002 0.003 15.014 15.033 multiply_cannon 2286 13.5 0.197 0.205 13.323 14.105 multiply_cannon_loop 2286 14.5 0.639 0.671 12.448 13.254 ot_mini 108 10.5 0.001 0.001 9.086 9.108 qs_ot_get_derivative 108 11.5 0.001 0.001 7.529 7.548 multiply_cannon_multrec 18288 15.5 1.931 2.855 6.786 7.074 rebuild_ks_matrix 119 8.3 0.000 0.000 6.474 6.498 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 6.474 6.497 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.730 5.750 dbcsr_mm_accdrv_process 38222 16.0 3.980 5.329 4.771 5.622 mp_waitall_1 169478 16.3 3.493 4.390 3.493 4.390 sum_up_and_integrate 119 10.3 0.030 0.031 4.050 4.054 integrate_v_rspace 119 11.3 0.002 0.003 4.019 4.027 init_scf_run 11 5.9 0.000 0.001 3.890 3.890 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.889 3.890 qs_ot_get_p 119 10.4 0.001 0.001 3.684 3.707 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.599 3.613 calculate_rho_elec 119 8.7 0.030 0.031 3.598 3.613 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.847 3.483 rs_pw_transfer 974 11.9 0.009 0.009 2.590 2.698 init_scf_loop 11 6.9 0.001 0.003 2.685 2.686 calculate_first_density_matrix 1 7.0 0.002 0.007 2.616 2.618 qs_ot_p2m_diag 50 11.0 0.012 0.012 2.384 2.389 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.071 2.365 apply_single 119 13.6 0.000 0.000 2.071 2.364 multiply_cannon_metrocomm3 18288 15.5 0.044 0.046 1.508 2.268 make_m2s 4572 13.5 0.045 0.047 2.044 2.191 density_rs2pw 119 9.7 0.004 0.004 2.106 2.176 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.110 2.110 make_images 4572 14.5 0.194 0.212 1.957 2.104 calculate_dm_sparse 119 9.5 0.000 0.000 1.923 1.945 jit_kernel_multiply 10 16.0 0.740 1.913 0.740 1.913 grid_integrate_task_list 119 12.3 1.791 1.908 1.791 1.908 prepare_preconditioner 11 7.9 0.000 0.000 1.870 1.872 make_preconditioner 11 8.9 0.001 0.002 1.870 1.872 pw_transfer 1439 11.6 0.064 0.066 1.825 1.841 potential_pw2rs 119 12.3 0.007 0.008 1.820 1.831 make_full_inverse_cholesky 11 9.9 0.000 0.001 1.716 1.804 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.734 1.749 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.746 1.747 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.686 1.703 mp_sum_l 7207 12.9 1.321 1.693 1.321 1.693 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.630 1.647 cp_fm_redistribute_end 50 14.0 1.228 1.619 1.234 1.620 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.558 1.572 multiply_cannon_sync_h2d 18288 15.5 1.376 1.534 1.376 1.534 ot_diis_step 108 11.5 0.011 0.011 1.530 1.532 cp_fm_diag_elpa_base 50 14.0 0.354 1.436 0.368 1.474 fft3d_ps 1201 14.6 0.509 0.526 1.420 1.437 fft_wrap_pw1pw2_140 487 13.2 0.086 0.089 1.329 1.345 grid_collocate_task_list 119 9.7 1.204 1.313 1.204 1.313 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.246 1.251 acc_transpose_blocks 18288 15.5 0.074 0.077 1.224 1.241 wfi_extrapolate 11 7.9 0.001 0.001 1.225 1.225 multiply_cannon_metrocomm1 18288 15.5 0.028 0.030 0.516 1.115 make_images_data 4572 15.5 0.046 0.051 0.906 1.107 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.049 1.051 hybrid_alltoall_any 4725 16.4 0.055 0.113 0.769 0.984 mp_alltoall_d11v 2130 13.8 0.853 0.982 0.853 0.982 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.904 0.932 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.814 0.817 mp_alltoall_z22v 1201 16.6 0.744 0.802 0.744 0.802 acc_transpose_blocks_kernels 18288 16.5 0.207 0.216 0.789 0.801 cp_fm_cholesky_invert 11 10.9 0.790 0.794 0.790 0.794 mp_allgather_i34 2286 14.5 0.333 0.736 0.333 0.736 mp_waitany 9880 13.7 0.595 0.724 0.595 0.724 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.654 0.716 rs_pw_transfer_RS2PW_140 130 11.5 0.121 0.130 0.594 0.715 cp_fm_cholesky_decompose 22 10.9 0.702 0.705 0.702 0.705 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=34.513000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.636364, yerr=1.822722 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a374e79360e402e0e0c3cfd3175b21c36af7649d_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 553.418752E+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 3672 62659. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 2412273. MP_ISendRecv 16422 74133. MP_Wait 24482 MP_comm_split 50 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.047 38.067 38.068 qs_mol_dyn_low 1 2.0 0.003 0.003 37.818 37.826 qs_forces 11 3.9 0.002 0.002 37.756 37.757 qs_energies 11 4.9 0.004 0.011 35.957 35.963 scf_env_do_scf 11 5.9 0.001 0.001 30.507 30.508 scf_env_do_scf_inner_loop 108 6.5 0.003 0.009 26.748 26.749 dbcsr_multiply_generic 2286 12.5 0.101 0.103 19.468 19.573 velocity_verlet 10 3.0 0.001 0.001 19.112 19.114 qs_scf_new_mos 108 7.5 0.001 0.001 17.668 17.727 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.667 17.727 ot_scf_mini 108 9.5 0.002 0.003 16.683 16.736 multiply_cannon 2286 13.5 0.232 0.265 14.810 15.535 multiply_cannon_loop 2286 14.5 0.943 0.973 13.706 14.517 ot_mini 108 10.5 0.001 0.001 10.313 10.377 multiply_cannon_multrec 27432 15.5 2.335 3.047 8.666 9.008 qs_ot_get_derivative 108 11.5 0.001 0.001 8.448 8.502 dbcsr_mm_accdrv_process 47916 15.9 5.427 6.994 6.237 7.415 rebuild_ks_matrix 119 8.3 0.000 0.000 7.062 7.113 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.062 7.113 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.272 6.317 sum_up_and_integrate 119 10.3 0.035 0.037 4.245 4.252 integrate_v_rspace 119 11.3 0.002 0.003 4.210 4.217 mp_waitall_1 145218 16.4 3.126 4.135 3.126 4.135 init_scf_run 11 5.9 0.000 0.001 3.920 3.920 scf_env_initial_rho_setup 11 6.9 0.000 0.001 3.920 3.920 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.784 3.819 calculate_rho_elec 119 8.7 0.040 0.046 3.783 3.819 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.040 3.804 init_scf_loop 11 6.9 0.000 0.002 3.736 3.736 qs_ot_get_p 119 10.4 0.001 0.001 3.631 3.709 make_m2s 4572 13.5 0.055 0.056 2.735 2.863 prepare_preconditioner 11 7.9 0.000 0.000 2.823 2.829 make_preconditioner 11 8.9 0.000 0.002 2.823 2.829 make_images 4572 14.5 0.271 0.331 2.625 2.752 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.407 2.743 rs_pw_transfer 974 11.9 0.009 0.009 2.630 2.717 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.486 2.664 apply_single 119 13.6 0.000 0.000 2.486 2.664 calculate_first_density_matrix 1 7.0 0.001 0.004 2.403 2.405 density_rs2pw 119 9.7 0.004 0.004 2.231 2.323 jit_kernel_multiply 10 16.0 0.750 2.318 0.750 2.318 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.161 2.189 calculate_dm_sparse 119 9.5 0.000 0.000 2.133 2.180 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.156 2.165 pw_transfer 1439 11.6 0.063 0.066 2.095 2.124 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.005 2.033 potential_pw2rs 119 12.3 0.008 0.009 1.977 1.985 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.940 1.941 grid_integrate_task_list 119 12.3 1.815 1.913 1.815 1.913 mp_sum_l 7207 12.9 1.508 1.881 1.508 1.881 ot_diis_step 108 11.5 0.012 0.012 1.822 1.822 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.791 1.791 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.727 1.740 fft3d_ps 1201 14.6 0.537 0.586 1.687 1.713 multiply_cannon_metrocomm3 27432 15.5 0.038 0.040 0.802 1.682 fft_wrap_pw1pw2_140 487 13.2 0.083 0.092 1.566 1.595 acc_transpose_blocks 27432 15.5 0.109 0.111 1.478 1.508 wfi_extrapolate 11 7.9 0.001 0.001 1.463 1.463 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.446 1.456 cp_fm_redistribute_end 50 14.0 0.955 1.420 0.959 1.422 cp_fm_diag_elpa_base 50 14.0 0.431 1.315 0.451 1.356 grid_collocate_task_list 119 9.7 1.220 1.307 1.220 1.307 make_images_data 4572 15.5 0.045 0.048 1.116 1.291 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.280 1.289 dbcsr_complete_redistribute 329 12.2 0.118 0.142 0.927 1.202 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.183 1.196 hybrid_alltoall_any 4725 16.4 0.062 0.150 0.945 1.161 cp_fm_upper_to_full 72 13.5 0.832 1.154 0.832 1.154 qs_energies_init_hamiltonians 11 5.9 0.009 0.018 1.118 1.122 mp_alltoall_d11v 2130 13.8 0.973 1.121 0.973 1.121 multiply_cannon_sync_h2d 27432 15.5 0.982 1.070 0.982 1.070 mp_alltoall_z22v 1201 16.6 1.033 1.058 1.033 1.058 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.645 0.920 cp_fm_cholesky_invert 11 10.9 0.898 0.902 0.898 0.902 acc_transpose_blocks_kernels 27432 16.5 0.265 0.273 0.861 0.891 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.883 0.888 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.793 0.868 make_images_sizes 4572 15.5 0.005 0.005 0.586 0.795 mp_alltoall_i44 4572 16.5 0.581 0.790 0.581 0.790 mp_alltoall_i22 627 13.8 0.464 0.781 0.464 0.781 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=38.068000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=523.818182, yerr=4.238743 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a374e79360e402e0e0c3cfd3175b21c36af7649d_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 601.071616E+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 3672 62658. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 3682667. MP_ISendRecv 10710 94533. MP_Wait 16690 MP_comm_split 50 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 0.041 30.178 30.178 qs_mol_dyn_low 1 2.0 0.003 0.005 29.953 29.965 qs_forces 11 3.9 0.002 0.002 29.725 29.725 qs_energies 11 4.9 0.002 0.004 27.988 27.991 scf_env_do_scf 11 5.9 0.000 0.001 22.770 22.770 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 19.955 19.956 velocity_verlet 10 3.0 0.001 0.001 15.492 15.505 dbcsr_multiply_generic 2286 12.5 0.093 0.096 12.865 12.953 qs_scf_new_mos 108 7.5 0.001 0.001 11.577 11.605 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.576 11.604 ot_scf_mini 108 9.5 0.002 0.002 10.870 10.900 multiply_cannon 2286 13.5 0.234 0.244 9.879 10.437 multiply_cannon_loop 2286 14.5 0.333 0.347 8.904 9.101 rebuild_ks_matrix 119 8.3 0.000 0.000 6.260 6.289 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.260 6.289 ot_mini 108 10.5 0.001 0.001 6.059 6.093 multiply_cannon_multrec 9144 15.5 1.582 1.823 5.786 6.027 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.602 5.629 qs_ot_get_derivative 108 11.5 0.001 0.001 4.716 4.742 dbcsr_mm_accdrv_process 12550 15.8 3.138 3.803 4.102 4.182 sum_up_and_integrate 119 10.3 0.037 0.041 3.903 3.908 integrate_v_rspace 119 11.3 0.002 0.003 3.865 3.871 init_scf_run 11 5.9 0.000 0.001 3.697 3.697 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.696 3.697 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.683 3.692 calculate_rho_elec 119 8.7 0.059 0.061 3.683 3.692 qs_ot_get_p 119 10.4 0.001 0.001 2.928 2.964 mp_waitall_1 121218 16.5 2.421 2.882 2.421 2.882 init_scf_loop 11 6.9 0.000 0.000 2.792 2.793 calculate_first_density_matrix 1 7.0 0.000 0.000 2.424 2.424 make_m2s 4572 13.5 0.035 0.036 2.094 2.269 rs_pw_transfer 974 11.9 0.008 0.008 2.107 2.207 make_images 4572 14.5 0.268 0.301 2.003 2.178 density_rs2pw 119 9.7 0.003 0.004 2.061 2.168 pw_transfer 1439 11.6 0.063 0.066 2.097 2.112 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.007 2.022 prepare_preconditioner 11 7.9 0.000 0.000 1.983 1.989 make_preconditioner 11 8.9 0.000 0.000 1.983 1.989 grid_integrate_task_list 119 12.3 1.831 1.945 1.831 1.945 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.855 1.887 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.834 1.835 calculate_dm_sparse 119 9.5 0.000 0.000 1.803 1.826 jit_kernel_multiply 10 15.6 0.927 1.807 0.927 1.807 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.805 1.806 fft3d_ps 1201 14.6 0.543 0.557 1.673 1.691 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.668 1.683 potential_pw2rs 119 12.3 0.010 0.010 1.653 1.662 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.619 1.620 fft_wrap_pw1pw2_140 487 13.2 0.082 0.085 1.579 1.595 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.546 1.554 grid_collocate_task_list 119 9.7 1.275 1.334 1.275 1.334 ot_diis_step 108 11.5 0.012 0.013 1.326 1.326 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.285 1.301 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.290 1.301 cp_fm_redistribute_end 50 14.0 0.642 1.274 0.645 1.275 qs_energies_init_hamiltonians 11 5.9 0.006 0.010 1.231 1.237 make_images_data 4572 15.5 0.039 0.042 0.982 1.237 cp_fm_diag_elpa_base 50 14.0 0.574 1.177 0.617 1.234 wfi_extrapolate 11 7.9 0.001 0.001 1.219 1.219 hybrid_alltoall_any 4725 16.4 0.062 0.176 0.926 1.213 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.185 1.211 apply_single 119 13.6 0.000 0.000 1.184 1.211 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.200 1.205 mp_alltoall_d11v 2130 13.8 0.959 1.129 0.959 1.129 acc_transpose_blocks 9144 15.5 0.038 0.039 1.014 1.052 mp_alltoall_z22v 1201 16.6 1.003 1.035 1.003 1.035 cp_fm_cholesky_invert 11 10.9 0.975 0.978 0.975 0.978 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.867 0.919 acc_transpose_blocks_kernels 9144 16.5 0.116 0.119 0.775 0.808 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.793 0.802 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.783 0.785 mp_allgather_i34 2286 14.5 0.301 0.775 0.301 0.775 multiply_cannon_sync_h2d 9144 15.5 0.707 0.769 0.707 0.769 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.680 0.727 dbcsr_complete_redistribute 329 12.2 0.191 0.237 0.684 0.714 jit_kernel_transpose 5 15.6 0.660 0.690 0.660 0.690 make_images_sizes 4572 15.5 0.005 0.005 0.436 0.652 mp_alltoall_i44 4572 16.5 0.431 0.647 0.431 0.647 multiply_cannon_metrocomm1 9144 15.5 0.022 0.023 0.476 0.640 yz_to_x 606 15.1 0.053 0.056 0.588 0.607 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=30.178000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=569.000000, yerr=5.257030 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a374e79360e402e0e0c3cfd3175b21c36af7649d_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 750.006272E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63729. MP_Allreduce 10074 433. MP_Sync 54 MP_Alltoall 1582 7383731. MP_ISendRecv 4998 189067. MP_Wait 8898 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 0.017 0.040 45.377 45.378 qs_mol_dyn_low 1 2.0 0.003 0.004 45.007 45.015 qs_forces 11 3.9 0.002 0.003 44.939 44.941 qs_energies 11 4.9 0.020 0.047 42.894 42.901 scf_env_do_scf 11 5.9 0.000 0.001 36.393 36.393 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 27.955 27.956 velocity_verlet 10 3.0 0.001 0.001 24.807 24.813 dbcsr_multiply_generic 2286 12.5 0.102 0.103 19.244 19.493 qs_scf_new_mos 108 7.5 0.001 0.001 18.126 18.223 qs_scf_loop_do_ot 108 8.5 0.001 0.001 18.125 18.222 ot_scf_mini 108 9.5 0.002 0.002 16.977 17.076 multiply_cannon 2286 13.5 0.317 0.332 14.586 15.634 multiply_cannon_loop 2286 14.5 0.346 0.352 13.188 14.184 ot_mini 108 10.5 0.001 0.001 9.880 10.002 multiply_cannon_multrec 9144 15.5 3.441 4.751 8.793 8.909 init_scf_loop 11 6.9 0.000 0.000 8.401 8.406 qs_ot_get_derivative 108 11.5 0.001 0.001 7.654 7.754 rebuild_ks_matrix 119 8.3 0.000 0.000 7.406 7.558 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.406 7.558 prepare_preconditioner 11 7.9 0.000 0.000 7.376 7.390 make_preconditioner 11 8.9 0.000 0.000 7.376 7.390 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.906 7.260 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.689 6.826 dbcsr_mm_accdrv_process 12550 15.8 4.389 6.137 5.230 6.541 cp_fm_upper_to_full 72 14.2 3.191 4.600 3.191 4.600 mp_waitall_1 97218 16.6 3.516 4.534 3.516 4.534 init_scf_run 11 5.9 0.000 0.001 4.350 4.350 scf_env_initial_rho_setup 11 6.9 0.000 0.000 4.349 4.349 qs_ot_get_p 119 10.4 0.001 0.001 4.154 4.296 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.204 4.245 calculate_rho_elec 119 8.7 0.118 0.121 4.204 4.244 sum_up_and_integrate 119 10.3 0.064 0.066 4.120 4.127 integrate_v_rspace 119 11.3 0.003 0.003 4.056 4.064 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.763 3.180 dbcsr_complete_redistribute 329 12.2 0.309 0.314 2.218 3.041 make_m2s 4572 13.5 0.038 0.039 2.765 2.975 make_images 4572 14.5 0.353 0.387 2.642 2.852 calculate_first_density_matrix 1 7.0 0.000 0.000 2.743 2.751 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.806 2.633 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 1.664 2.597 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.297 2.559 apply_single 119 13.6 0.000 0.000 2.296 2.559 qs_ot_p2m_diag 50 11.0 0.043 0.043 2.548 2.550 pw_transfer 1439 11.6 0.066 0.066 2.420 2.434 calculate_dm_sparse 119 9.5 0.000 0.000 2.388 2.413 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.357 2.358 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.324 2.337 mp_alltoall_i22 627 13.8 1.513 2.337 1.513 2.337 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.465 2.284 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.247 2.248 density_rs2pw 119 9.7 0.004 0.004 2.171 2.207 ot_diis_step 108 11.5 0.014 0.014 2.181 2.182 grid_integrate_task_list 119 12.3 2.090 2.158 2.090 2.158 mp_sum_l 7207 12.9 1.353 2.145 1.353 2.145 fft3d_ps 1201 14.6 0.570 0.580 1.958 1.969 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.959 1.964 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.908 1.960 fft_wrap_pw1pw2_140 487 13.2 0.088 0.092 1.859 1.873 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.833 1.833 rs_pw_transfer 974 11.9 0.009 0.009 1.778 1.831 cp_fm_diag_elpa_base 50 14.0 1.682 1.738 1.830 1.831 cp_fm_cholesky_invert 11 10.9 1.812 1.816 1.812 1.816 qs_energies_init_hamiltonians 11 5.9 0.013 0.029 1.808 1.812 make_images_data 4572 15.5 0.043 0.046 1.351 1.642 hybrid_alltoall_any 4725 16.4 0.088 0.150 1.368 1.597 grid_collocate_task_list 119 9.7 1.511 1.553 1.511 1.553 wfi_extrapolate 11 7.9 0.001 0.001 1.537 1.538 potential_pw2rs 119 12.3 0.014 0.014 1.517 1.522 mp_alltoall_d11v 2130 13.8 1.439 1.509 1.439 1.509 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.400 1.418 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.260 1.290 mp_alltoall_z22v 1201 16.6 1.251 1.279 1.251 1.279 acc_transpose_blocks 9144 15.5 0.037 0.038 1.158 1.188 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.105 1.121 jit_kernel_multiply 6 15.6 0.814 1.089 0.814 1.089 multiply_cannon_sync_h2d 9144 15.5 1.048 1.053 1.048 1.053 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.984 1.019 qs_create_task_list 11 7.9 0.001 0.001 0.951 0.961 generate_qs_task_list 11 8.9 0.375 0.394 0.950 0.960 copy_dbcsr_to_fm 153 11.3 0.002 0.002 0.923 0.953 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.931 0.946 acc_transpose_blocks_kernels 9144 16.5 0.116 0.117 0.911 0.942 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=45.378000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=701.636364, yerr=14.156738 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a374e79360e402e0e0c3cfd3175b21c36af7649d_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 499.609600E+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 3521 65372. MP_Allreduce 9840 486. MP_Sync 100 MP_Alltoall 1938 1217550. MP_ISendRecv 41800 9096. MP_Wait 58168 MP_comm_split 48 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.078 0.097 83.220 83.222 qs_mol_dyn_low 1 2.0 0.003 0.006 82.655 82.673 qs_forces 11 3.9 0.005 0.013 82.513 82.532 qs_energies 11 4.9 0.003 0.016 79.717 79.743 scf_env_do_scf 11 5.9 0.001 0.001 70.492 70.495 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 64.964 64.964 dbcsr_multiply_generic 2055 12.4 0.107 0.112 52.360 52.629 qs_scf_new_mos 99 7.5 0.000 0.001 48.156 48.268 qs_scf_loop_do_ot 99 8.5 0.001 0.001 48.156 48.268 ot_scf_mini 99 9.5 0.002 0.002 45.799 45.894 multiply_cannon 2055 13.4 0.178 0.188 42.740 43.809 multiply_cannon_loop 2055 14.4 1.568 1.605 41.758 42.855 velocity_verlet 10 3.0 0.001 0.001 41.743 41.743 ot_mini 99 10.5 0.001 0.001 27.751 27.840 qs_ot_get_derivative 99 11.5 0.001 0.001 20.941 21.050 multiply_cannon_multrec 49320 15.4 12.083 12.845 17.208 17.991 rebuild_ks_matrix 110 8.3 0.000 0.001 14.350 14.465 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.013 14.350 14.465 mp_waitall_1 241148 16.1 11.905 12.754 11.905 12.754 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.622 12.717 multiply_cannon_sync_h2d 49320 15.4 9.925 10.455 9.925 10.455 qs_ot_get_p 110 10.4 0.001 0.001 9.492 9.615 multiply_cannon_metrocomm3 49320 15.4 0.079 0.085 7.008 8.144 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.171 7.696 apply_single 110 13.6 0.000 0.001 7.170 7.695 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.077 7.627 init_scf_run 11 5.9 0.000 0.001 7.086 7.087 scf_env_initial_rho_setup 11 6.9 0.000 0.002 7.086 7.087 sum_up_and_integrate 110 10.3 0.036 0.043 6.958 6.972 integrate_v_rspace 110 11.3 0.002 0.003 6.922 6.945 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.783 6.832 ot_diis_step 99 11.5 0.006 0.006 6.522 6.522 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.198 6.337 calculate_rho_elec 110 8.6 0.020 0.024 6.197 6.336 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.086 6.113 mp_sum_l 6514 12.8 5.258 5.976 5.258 5.976 init_scf_loop 11 6.9 0.001 0.004 5.500 5.501 dbcsr_mm_accdrv_process 87628 16.1 2.116 2.203 5.004 5.283 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.091 5.092 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.615 4.616 cp_fm_redistribute_end 48 14.0 4.016 4.590 4.020 4.591 cp_fm_diag_elpa_base 48 14.0 0.566 4.530 0.569 4.556 make_m2s 4110 13.4 0.061 0.067 3.922 4.054 wfi_extrapolate 11 7.9 0.001 0.001 3.997 3.997 rs_pw_transfer 902 11.9 0.011 0.012 3.729 3.970 make_images 4110 14.4 0.177 0.191 3.826 3.961 calculate_dm_sparse 110 9.5 0.000 0.001 3.795 3.895 multiply_cannon_metrocomm1 49320 15.4 0.059 0.062 2.797 3.813 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.618 3.622 density_rs2pw 110 9.6 0.004 0.004 3.225 3.464 prepare_preconditioner 11 7.9 0.000 0.000 3.356 3.374 make_preconditioner 11 8.9 0.000 0.001 3.356 3.374 grid_integrate_task_list 110 12.3 3.229 3.370 3.229 3.370 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.279 3.309 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.232 3.287 make_full_inverse_cholesky 11 9.9 0.000 0.001 3.130 3.186 calculate_first_density_matrix 1 7.0 0.000 0.002 2.928 3.010 pw_transfer 1331 11.6 0.054 0.062 2.819 2.896 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 2.731 2.808 jit_kernel_multiply 13 15.9 2.603 2.646 2.603 2.646 mp_alltoall_d11v 2046 13.8 2.206 2.610 2.206 2.610 potential_pw2rs 110 12.3 0.006 0.007 2.553 2.580 fft_wrap_pw1pw2_140 451 13.1 0.168 0.189 2.310 2.387 fft3d_ps 1111 14.6 0.800 0.914 2.270 2.337 acc_transpose_blocks 49320 15.4 0.222 0.233 2.182 2.260 grid_collocate_task_list 110 9.6 2.086 2.206 2.086 2.206 mp_waitany 14300 13.8 1.837 2.118 1.837 2.118 mp_sum_d 3879 11.9 1.501 1.983 1.501 1.983 make_images_data 4110 15.4 0.043 0.046 1.787 1.907 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.868 1.891 hybrid_alltoall_any 4261 16.3 0.082 0.477 1.566 1.828 cp_fm_cholesky_invert 11 10.9 1.812 1.816 1.812 1.816 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.719 1.806 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=83.222000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=475.181818, yerr=1.748671 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a374e79360e402e0e0c3cfd3175b21c36af7649d_performance_tests/10/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 390.715586E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 5019072 0.0% 0.0% 100.0% average stack size 0.0 0.0 196.1 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 589.266944E+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 3521 65587. MP_Allreduce 9839 562. MP_Sync 100 MP_Alltoall 1717 3254402. MP_ISendRecv 20680 26400. MP_Wait 32692 MP_comm_split 48 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.050 0.151 68.951 68.953 qs_mol_dyn_low 1 2.0 0.003 0.003 68.322 68.370 qs_forces 11 3.9 0.002 0.002 68.216 68.217 qs_energies 11 4.9 0.012 0.045 64.828 64.854 scf_env_do_scf 11 5.9 0.001 0.002 56.106 56.111 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 48.422 48.422 dbcsr_multiply_generic 2055 12.4 0.117 0.143 37.868 37.994 velocity_verlet 10 3.0 0.002 0.009 35.803 35.805 qs_scf_new_mos 99 7.5 0.001 0.001 32.725 32.856 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.724 32.856 multiply_cannon 2055 13.4 0.224 0.245 31.199 32.321 ot_scf_mini 99 9.5 0.003 0.004 31.068 31.195 multiply_cannon_loop 2055 14.4 0.930 0.957 29.904 30.820 ot_mini 99 10.5 0.001 0.001 18.282 18.413 multiply_cannon_multrec 24660 15.4 7.603 9.207 13.901 15.531 rebuild_ks_matrix 110 8.3 0.000 0.001 13.569 13.683 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.024 13.568 13.683 qs_ot_get_derivative 99 11.5 0.001 0.001 12.440 12.572 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.962 12.058 mp_waitall_1 186928 16.3 7.932 10.362 7.932 10.362 multiply_cannon_sync_h2d 24660 15.4 7.044 8.177 7.044 8.177 multiply_cannon_metrocomm3 24660 15.4 0.069 0.071 5.271 7.831 init_scf_loop 11 6.9 0.001 0.004 7.648 7.650 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.595 7.306 apply_single 110 13.6 0.000 0.001 6.595 7.306 sum_up_and_integrate 110 10.3 0.053 0.060 6.457 6.468 integrate_v_rspace 110 11.3 0.002 0.002 6.404 6.417 dbcsr_mm_accdrv_process 52282 16.1 4.769 5.673 6.139 6.411 qs_ot_get_p 110 10.4 0.001 0.001 6.136 6.283 init_scf_run 11 5.9 0.000 0.001 6.237 6.240 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.237 6.240 ot_diis_step 99 11.5 0.010 0.010 5.791 5.791 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.673 5.684 calculate_rho_elec 110 8.6 0.039 0.047 5.672 5.684 prepare_preconditioner 11 7.9 0.000 0.000 5.619 5.633 make_preconditioner 11 8.9 0.000 0.001 5.619 5.633 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.661 5.429 make_full_inverse_cholesky 11 9.9 0.000 0.001 5.197 5.354 make_m2s 4110 13.4 0.057 0.061 4.227 4.704 make_images 4110 14.4 0.395 0.440 4.116 4.591 qs_ot_p2m_diag 48 11.0 0.028 0.044 4.293 4.315 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.859 3.859 wfi_extrapolate 11 7.9 0.001 0.001 3.477 3.478 pw_transfer 1331 11.6 0.065 0.071 3.241 3.380 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.369 3.372 grid_integrate_task_list 110 12.3 3.146 3.348 3.146 3.348 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.329 3.332 cp_fm_redistribute_end 48 14.0 2.525 3.305 2.527 3.305 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.135 3.268 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.186 3.260 density_rs2pw 110 9.6 0.004 0.004 2.969 3.169 rs_pw_transfer 902 11.9 0.012 0.013 2.886 3.121 cp_fm_diag_elpa_base 48 14.0 0.743 3.011 0.774 3.097 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.025 3.063 calculate_dm_sparse 110 9.5 0.001 0.001 2.985 3.017 make_images_data 4110 15.4 0.047 0.051 2.389 2.873 hybrid_alltoall_any 4261 16.3 0.102 0.440 2.093 2.821 fft_wrap_pw1pw2_140 451 13.1 0.200 0.218 2.653 2.787 calculate_first_density_matrix 1 7.0 0.001 0.005 2.667 2.672 cp_fm_cholesky_invert 11 10.9 2.633 2.639 2.633 2.639 fft3d_ps 1111 14.6 1.075 1.277 2.493 2.610 mp_sum_l 6514 12.8 1.873 2.501 1.873 2.501 potential_pw2rs 110 12.3 0.008 0.009 2.323 2.336 grid_collocate_task_list 110 9.6 2.117 2.296 2.117 2.296 mp_alltoall_d11v 2046 13.8 1.798 2.027 1.798 2.027 jit_kernel_multiply 10 16.3 1.019 2.021 1.019 2.021 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.966 1.987 qs_energies_init_hamiltonians 11 5.9 0.002 0.007 1.899 1.902 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.774 1.789 multiply_cannon_metrocomm4 22605 15.4 0.073 0.078 0.780 1.713 mp_allgather_i34 2055 14.4 0.620 1.707 0.620 1.707 mp_irecv_dv 57340 16.2 0.656 1.595 0.656 1.595 acc_transpose_blocks 24660 15.4 0.109 0.113 1.522 1.574 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.561 1.571 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.389 1.542 dbcsr_complete_redistribute 325 12.2 0.245 0.312 1.154 1.419 mp_waitany 10164 13.8 1.199 1.414 1.199 1.414 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=68.953000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=555.090909, yerr=8.084267 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a374e79360e402e0e0c3cfd3175b21c36af7649d_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 661.516288E+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 3521 65578. MP_Allreduce 9838 559. MP_Sync 100 MP_Alltoall 1496 4511006. MP_ISendRecv 13640 27424. MP_Wait 32318 MP_comm_split 48 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 0.040 0.086 61.027 61.029 qs_mol_dyn_low 1 2.0 0.003 0.003 60.485 60.534 qs_forces 11 3.9 0.003 0.005 60.045 60.046 qs_energies 11 4.9 0.004 0.026 56.859 56.879 scf_env_do_scf 11 5.9 0.001 0.001 48.194 48.194 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 39.581 39.582 velocity_verlet 10 3.0 0.001 0.001 32.573 32.586 dbcsr_multiply_generic 2055 12.4 0.109 0.115 28.827 29.097 qs_scf_new_mos 99 7.5 0.001 0.001 25.112 25.232 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.111 25.231 ot_scf_mini 99 9.5 0.002 0.003 23.891 24.027 multiply_cannon 2055 13.4 0.228 0.242 22.455 23.620 multiply_cannon_loop 2055 14.4 0.619 0.643 21.269 22.388 ot_mini 99 10.5 0.001 0.001 13.539 13.673 rebuild_ks_matrix 110 8.3 0.000 0.001 12.117 12.263 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.015 12.116 12.263 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.696 10.830 multiply_cannon_multrec 16440 15.4 3.853 4.768 9.979 10.760 mp_waitall_1 146766 16.3 7.171 10.254 7.171 10.254 qs_ot_get_derivative 99 11.5 0.001 0.001 9.117 9.251 init_scf_loop 11 6.9 0.005 0.043 8.577 8.578 multiply_cannon_metrocomm3 16440 15.4 0.043 0.046 4.318 7.230 prepare_preconditioner 11 7.9 0.000 0.000 6.771 6.789 make_preconditioner 11 8.9 0.000 0.002 6.771 6.789 make_full_inverse_cholesky 11 9.9 0.000 0.001 6.102 6.463 sum_up_and_integrate 110 10.3 0.061 0.062 6.381 6.396 integrate_v_rspace 110 11.3 0.002 0.002 6.320 6.335 dbcsr_mm_accdrv_process 34862 16.1 5.010 5.600 5.981 6.134 init_scf_run 11 5.9 0.000 0.001 6.129 6.129 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.129 6.129 qs_ot_get_p 110 10.4 0.001 0.001 5.575 5.737 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.567 5.577 calculate_rho_elec 110 8.6 0.058 0.058 5.566 5.576 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.944 5.392 apply_single 110 13.6 0.000 0.000 4.944 5.392 make_m2s 4110 13.4 0.050 0.053 4.233 4.573 make_images 4110 14.4 0.391 0.510 4.117 4.456 ot_diis_step 99 11.5 0.011 0.011 4.394 4.394 multiply_cannon_sync_h2d 16440 15.4 3.661 4.139 3.661 4.139 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.924 3.928 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.079 3.728 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.584 3.584 grid_integrate_task_list 110 12.3 3.182 3.396 3.182 3.396 pw_transfer 1331 11.6 0.065 0.071 3.201 3.207 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.130 3.133 calculate_first_density_matrix 1 7.0 0.000 0.003 3.111 3.112 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.095 3.103 density_rs2pw 110 9.6 0.004 0.004 2.861 3.050 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.019 3.021 cp_fm_redistribute_end 48 14.0 1.876 2.992 1.878 2.992 cp_fm_diag_elpa_base 48 14.0 1.053 2.859 1.110 2.964 wfi_extrapolate 11 7.9 0.001 0.001 2.935 2.936 rs_pw_transfer 902 11.9 0.010 0.011 2.731 2.931 make_images_data 4110 15.4 0.044 0.051 2.391 2.828 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.746 2.804 calculate_dm_sparse 110 9.5 0.001 0.001 2.746 2.784 hybrid_alltoall_any 4261 16.3 0.105 0.375 2.081 2.723 cp_fm_cholesky_invert 11 10.9 2.662 2.668 2.662 2.668 fft_wrap_pw1pw2_140 451 13.1 0.210 0.214 2.633 2.644 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.379 2.447 multiply_cannon_metrocomm4 14385 15.4 0.044 0.049 0.890 2.438 fft3d_ps 1111 14.6 1.064 1.085 2.410 2.420 grid_collocate_task_list 110 9.6 2.177 2.402 2.177 2.402 mp_irecv_dv 48980 15.7 0.821 2.312 0.821 2.312 mp_alltoall_d11v 2046 13.8 1.773 2.281 1.773 2.281 potential_pw2rs 110 12.3 0.010 0.011 2.217 2.231 mp_sum_l 6514 12.8 1.536 2.147 1.536 2.147 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.049 2.050 dbcsr_complete_redistribute 325 12.2 0.343 0.383 1.566 2.026 cp_fm_upper_to_full 70 13.6 1.372 1.842 1.372 1.842 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.708 1.722 jit_kernel_multiply 8 16.6 0.584 1.654 0.584 1.654 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.623 1.637 cp_fm_cholesky_decompose 22 10.9 1.596 1.616 1.596 1.616 mp_allgather_i34 2055 14.4 0.494 1.580 0.494 1.580 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.084 1.530 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.368 1.499 mp_waitany 17072 13.8 1.175 1.379 1.175 1.379 rs_gather_matrices 110 12.3 0.137 0.149 0.872 1.357 acc_transpose_blocks 16440 15.4 0.071 0.075 1.292 1.309 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.264 1.273 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=61.029000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=624.000000, yerr=9.657781 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a374e79360e402e0e0c3cfd3175b21c36af7649d_performance_tests/12/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 601.317074E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 4916280 0.0% 0.0% 100.0% average stack size 0.0 0.0 200.2 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 728.670208E+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 3521 65576. MP_Allreduce 9838 600. MP_Sync 100 MP_Alltoall 1496 5863162. MP_ISendRecv 10120 43184. MP_Wait 25102 MP_comm_split 48 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 0.088 0.137 71.784 71.785 qs_mol_dyn_low 1 2.0 0.003 0.004 70.973 71.056 qs_forces 11 3.9 0.002 0.002 70.900 70.901 qs_energies 11 4.9 0.002 0.006 67.390 67.393 scf_env_do_scf 11 5.9 0.001 0.002 58.372 58.374 scf_env_do_scf_inner_loop 99 6.5 0.005 0.025 45.410 45.411 velocity_verlet 10 3.0 0.001 0.001 39.195 39.198 dbcsr_multiply_generic 2055 12.4 0.116 0.118 30.829 31.135 qs_scf_new_mos 99 7.5 0.001 0.001 29.807 29.944 qs_scf_loop_do_ot 99 8.5 0.001 0.001 29.806 29.943 ot_scf_mini 99 9.5 0.003 0.003 28.129 28.259 multiply_cannon 2055 13.4 0.245 0.258 22.819 24.135 multiply_cannon_loop 2055 14.4 0.892 0.911 21.227 22.039 ot_mini 99 10.5 0.001 0.001 15.235 15.395 multiply_cannon_multrec 24660 15.4 4.233 6.733 12.727 13.977 init_scf_loop 11 6.9 0.000 0.002 12.905 12.906 rebuild_ks_matrix 110 8.3 0.000 0.001 12.662 12.755 qs_ks_build_kohn_sham_matrix 110 9.3 0.017 0.052 12.661 12.755 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.206 11.288 qs_ot_get_derivative 99 11.5 0.001 0.001 10.991 11.128 prepare_preconditioner 11 7.9 0.000 0.000 11.065 11.083 make_preconditioner 11 8.9 0.000 0.000 11.065 11.083 make_full_inverse_cholesky 11 9.9 0.000 0.000 9.183 10.704 dbcsr_mm_accdrv_process 52304 16.0 6.933 8.004 8.348 9.304 qs_ot_get_p 110 10.4 0.002 0.010 7.703 7.861 mp_waitall_1 126806 16.4 5.529 7.705 5.529 7.705 sum_up_and_integrate 110 10.3 0.067 0.071 6.880 6.895 integrate_v_rspace 110 11.3 0.002 0.003 6.812 6.827 make_m2s 4110 13.4 0.060 0.061 6.116 6.377 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.267 6.275 calculate_rho_elec 110 8.6 0.077 0.081 6.266 6.275 make_images 4110 14.4 0.570 0.686 5.973 6.227 init_scf_run 11 5.9 0.000 0.001 5.941 5.942 scf_env_initial_rho_setup 11 6.9 0.001 0.004 5.941 5.942 qs_ot_p2m_diag 48 11.0 0.055 0.063 5.457 5.474 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.900 4.902 cp_fm_upper_to_full 70 13.8 3.396 4.847 3.396 4.847 ot_diis_step 99 11.5 0.011 0.011 4.197 4.198 dbcsr_complete_redistribute 325 12.2 0.409 0.448 3.012 4.190 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.058 4.145 apply_single 110 13.6 0.000 0.000 4.058 4.145 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.100 4.105 cp_fm_redistribute_end 48 14.0 2.021 4.004 2.024 4.007 cp_fm_diag_elpa_base 48 14.0 1.892 3.846 1.975 3.953 pw_transfer 1331 11.6 0.064 0.072 3.806 3.837 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.701 3.735 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.505 3.572 multiply_cannon_metrocomm3 24660 15.4 0.036 0.037 1.516 3.512 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.333 3.503 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.417 3.489 density_rs2pw 110 9.6 0.004 0.004 3.321 3.461 grid_integrate_task_list 110 12.3 3.268 3.443 3.268 3.443 make_images_data 4110 15.4 0.046 0.050 3.097 3.411 cp_fm_cholesky_invert 11 10.9 3.312 3.321 3.312 3.321 rs_pw_transfer 902 11.9 0.010 0.011 3.120 3.308 multiply_cannon_sync_h2d 24660 15.4 3.163 3.305 3.163 3.305 hybrid_alltoall_any 4261 16.3 0.119 0.458 2.576 3.304 wfi_extrapolate 11 7.9 0.001 0.001 3.194 3.194 calculate_dm_sparse 110 9.5 0.001 0.001 3.029 3.059 fft_wrap_pw1pw2_140 451 13.1 0.200 0.213 2.995 3.036 fft3d_ps 1111 14.6 1.066 1.097 3.006 3.029 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.873 3.017 mp_alltoall_i22 605 13.7 1.744 2.966 1.744 2.966 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.923 2.925 calculate_first_density_matrix 1 7.0 0.000 0.002 2.651 2.656 mp_alltoall_d11v 2046 13.8 2.273 2.593 2.273 2.593 potential_pw2rs 110 12.3 0.012 0.013 2.559 2.576 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.488 2.520 qs_energies_init_hamiltonians 11 5.9 0.017 0.033 2.458 2.475 grid_collocate_task_list 110 9.6 2.223 2.369 2.223 2.369 cp_fm_cholesky_decompose 22 10.9 2.010 2.051 2.010 2.051 mp_allgather_i34 2055 14.4 0.695 1.918 0.695 1.918 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.826 1.857 mp_sum_l 6514 12.8 1.232 1.847 1.232 1.847 mp_alltoall_z22v 1111 16.6 1.718 1.790 1.718 1.790 jit_kernel_multiply 9 15.7 1.086 1.745 1.086 1.745 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.708 1.718 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.601 1.701 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.604 1.616 multiply_cannon_metrocomm4 20550 15.4 0.056 0.059 0.835 1.594 mp_irecv_dv 62702 16.1 0.739 1.514 0.739 1.514 acc_transpose_blocks 24660 15.4 0.102 0.103 1.486 1.502 mp_waitany 13376 13.8 1.265 1.499 1.265 1.499 qs_env_update_s_mstruct 11 6.9 0.001 0.013 1.376 1.478 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=71.785000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=690.909091, yerr=9.802176 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a374e79360e402e0e0c3cfd3175b21c36af7649d_performance_tests/13/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 807.299199E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 1438408 0.0% 0.0% 100.0% average stack size 0.0 0.0 684.2 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 834.686976E+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 3521 65574. MP_Allreduce 9838 640. MP_Sync 100 MP_Alltoall 1496 8504061. MP_ISendRecv 6600 54848. MP_Wait 17226 MP_comm_split 48 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 0.017 0.030 58.043 58.043 qs_mol_dyn_low 1 2.0 0.003 0.003 57.626 57.635 qs_forces 11 3.9 0.006 0.007 57.146 57.147 qs_energies 11 4.9 0.001 0.002 53.452 53.461 scf_env_do_scf 11 5.9 0.000 0.001 44.098 44.098 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 35.914 35.915 velocity_verlet 10 3.0 0.001 0.001 31.989 32.018 dbcsr_multiply_generic 2055 12.4 0.105 0.107 23.967 24.138 qs_scf_new_mos 99 7.5 0.001 0.001 21.599 21.636 qs_scf_loop_do_ot 99 8.5 0.001 0.001 21.598 21.635 ot_scf_mini 99 9.5 0.002 0.002 20.340 20.368 multiply_cannon 2055 13.4 0.252 0.268 18.270 19.482 multiply_cannon_loop 2055 14.4 0.325 0.338 16.916 17.192 rebuild_ks_matrix 110 8.3 0.000 0.000 11.571 11.592 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 11.570 11.591 ot_mini 99 10.5 0.001 0.001 10.691 10.711 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.282 10.302 multiply_cannon_multrec 8220 15.4 3.244 4.679 8.088 9.095 init_scf_loop 11 6.9 0.000 0.000 8.127 8.129 mp_waitall_1 106626 16.5 6.327 7.913 6.327 7.913 qs_ot_get_derivative 99 11.5 0.001 0.001 6.907 6.935 prepare_preconditioner 11 7.9 0.000 0.000 6.476 6.482 make_preconditioner 11 8.9 0.000 0.000 6.476 6.482 sum_up_and_integrate 110 10.3 0.080 0.082 6.320 6.331 integrate_v_rspace 110 11.3 0.003 0.003 6.239 6.251 init_scf_run 11 5.9 0.000 0.001 6.162 6.163 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.162 6.162 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.062 6.141 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.768 5.781 calculate_rho_elec 110 8.6 0.115 0.116 5.767 5.781 qs_ot_get_p 110 10.4 0.001 0.001 5.705 5.738 dbcsr_mm_accdrv_process 17442 15.9 3.059 4.263 4.713 5.690 make_m2s 4110 13.4 0.039 0.040 4.353 4.608 make_images 4110 14.4 0.633 0.688 4.222 4.475 multiply_cannon_metrocomm3 8220 15.4 0.017 0.018 2.943 4.278 qs_ot_p2m_diag 48 11.0 0.081 0.084 4.212 4.216 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.897 3.897 ot_diis_step 99 11.5 0.012 0.012 3.760 3.760 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.674 3.699 apply_single 110 13.6 0.000 0.000 3.674 3.698 grid_integrate_task_list 110 12.3 3.349 3.499 3.349 3.499 pw_transfer 1331 11.6 0.065 0.071 3.350 3.378 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.321 3.322 cp_fm_redistribute_end 48 14.0 0.842 3.280 0.847 3.282 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.243 3.274 calculate_first_density_matrix 1 7.0 0.000 0.000 3.251 3.251 cp_fm_diag_elpa_base 48 14.0 2.254 3.078 2.424 3.239 cp_fm_cholesky_invert 11 10.9 3.210 3.214 3.210 3.214 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.130 3.140 multiply_cannon_sync_h2d 8220 15.4 2.913 3.080 2.913 3.080 density_rs2pw 110 9.6 0.004 0.004 2.808 2.919 calculate_dm_sparse 110 9.5 0.001 0.001 2.843 2.885 make_images_data 4110 15.4 0.039 0.044 2.430 2.836 wfi_extrapolate 11 7.9 0.001 0.001 2.816 2.816 fft_wrap_pw1pw2_140 451 13.1 0.212 0.216 2.733 2.766 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.722 2.733 hybrid_alltoall_any 4261 16.3 0.200 0.862 2.359 2.711 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 2.666 2.670 fft3d_ps 1111 14.6 1.118 1.167 2.501 2.534 rs_pw_transfer 902 11.9 0.010 0.010 2.365 2.492 grid_collocate_task_list 110 9.6 2.330 2.421 2.330 2.421 jit_kernel_multiply 9 15.7 1.348 2.144 1.348 2.144 mp_alltoall_d11v 2046 13.8 1.757 2.125 1.757 2.125 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.074 2.084 potential_pw2rs 110 12.3 0.015 0.015 2.041 2.059 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.772 1.993 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.856 1.873 cp_fm_cholesky_decompose 22 10.9 1.807 1.822 1.807 1.822 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.801 1.813 dbcsr_complete_redistribute 325 12.2 0.556 0.599 1.560 1.663 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.514 1.631 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.620 1.624 mp_allgather_i34 2055 14.4 0.540 1.592 0.540 1.592 qs_create_task_list 11 7.9 0.000 0.001 1.236 1.332 generate_qs_task_list 11 8.9 0.381 0.449 1.235 1.331 acc_transpose_blocks 8220 15.4 0.035 0.036 1.249 1.259 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.217 1.234 mp_waitany 9240 13.8 1.066 1.218 1.066 1.218 mp_alltoall_z22v 1111 16.6 1.149 1.173 1.149 1.173 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=58.043000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=781.272727, yerr=16.608272 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a374e79360e402e0e0c3cfd3175b21c36af7649d_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 1.391297E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67104. MP_Allreduce 9672 819. MP_Sync 52 MP_Alltoall 1474 16505187. MP_ISendRecv 4620 360267. MP_Wait 7524 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.022 0.042 87.834 87.835 qs_mol_dyn_low 1 2.0 0.003 0.003 87.362 87.371 qs_forces 11 3.9 0.002 0.002 87.257 87.259 qs_energies 11 4.9 0.001 0.001 83.153 83.156 scf_env_do_scf 11 5.9 0.001 0.001 72.418 72.419 velocity_verlet 10 3.0 0.001 0.001 55.635 55.646 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 43.721 43.722 dbcsr_multiply_generic 2055 12.4 0.120 0.122 29.769 29.857 init_scf_loop 11 6.9 0.000 0.000 28.624 28.627 qs_scf_new_mos 99 7.5 0.001 0.001 26.906 26.991 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.905 26.990 prepare_preconditioner 11 7.9 0.000 0.000 26.623 26.629 make_preconditioner 11 8.9 0.000 0.000 26.623 26.629 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.807 26.088 ot_scf_mini 99 9.5 0.002 0.002 25.126 25.212 multiply_cannon 2055 13.4 0.371 0.415 22.365 23.301 multiply_cannon_loop 2055 14.4 0.342 0.344 20.485 20.854 cp_fm_upper_to_full 70 14.2 12.613 18.140 12.613 18.140 ot_mini 99 10.5 0.001 0.001 13.880 13.969 rebuild_ks_matrix 110 8.3 0.001 0.001 13.506 13.545 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 13.505 13.545 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.263 12.300 dbcsr_complete_redistribute 325 12.2 1.023 1.041 7.455 10.639 multiply_cannon_multrec 8220 15.4 4.371 4.554 9.911 10.067 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.398 9.573 qs_ot_get_derivative 99 11.5 0.001 0.001 9.258 9.344 mp_waitall_1 87304 16.6 8.504 9.227 8.504 9.227 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.801 8.953 mp_alltoall_i22 605 13.7 5.424 8.608 5.424 8.608 sum_up_and_integrate 110 10.3 0.150 0.152 6.716 6.731 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.663 6.718 calculate_rho_elec 110 8.6 0.227 0.227 6.662 6.717 integrate_v_rspace 110 11.3 0.003 0.004 6.566 6.581 init_scf_run 11 5.9 0.000 0.001 6.381 6.381 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.381 6.381 make_m2s 4110 13.4 0.043 0.044 5.570 6.085 qs_ot_get_p 110 10.4 0.001 0.001 6.030 6.045 make_images 4110 14.4 0.872 0.910 5.381 5.895 dbcsr_mm_accdrv_process 11614 15.7 3.205 3.608 5.396 5.709 cp_fm_cholesky_invert 11 10.9 5.556 5.561 5.556 5.561 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 5.110 5.489 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.723 5.167 apply_single 110 13.6 0.000 0.000 4.723 5.167 ot_diis_step 99 11.5 0.015 0.016 4.596 4.596 qs_ot_p2m_diag 48 11.0 0.151 0.155 4.362 4.369 multiply_cannon_sync_h2d 8220 15.4 3.953 3.959 3.953 3.959 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.912 3.913 hybrid_alltoall_any 4261 16.3 0.255 0.550 3.098 3.882 pw_transfer 1331 11.6 0.073 0.074 3.851 3.858 make_images_data 4110 15.4 0.042 0.045 3.134 3.834 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 3.735 3.742 qs_energies_init_hamiltonians 11 5.9 0.035 0.052 3.729 3.731 grid_integrate_task_list 110 12.3 3.661 3.718 3.661 3.718 calculate_dm_sparse 110 9.5 0.001 0.001 3.496 3.574 wfi_extrapolate 11 7.9 0.001 0.001 3.385 3.385 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.920 3.347 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.300 3.300 cp_fm_diag_elpa_base 48 14.0 2.764 2.960 3.298 3.298 fft_wrap_pw1pw2_140 451 13.1 0.214 0.217 3.224 3.232 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.133 3.140 density_rs2pw 110 9.6 0.004 0.004 3.087 3.103 fft3d_ps 1111 14.6 1.268 1.277 2.946 2.953 calculate_first_density_matrix 1 7.0 0.000 0.000 2.891 2.891 grid_collocate_task_list 110 9.6 2.631 2.663 2.631 2.663 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.596 2.626 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.612 2.621 rs_pw_transfer 902 11.9 0.010 0.011 2.264 2.348 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.208 2.262 mp_alltoall_d11v 2046 13.8 2.187 2.245 2.187 2.245 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.087 2.201 jit_kernel_multiply 10 15.4 1.991 2.155 1.991 2.155 potential_pw2rs 110 12.3 0.021 0.021 2.142 2.148 cp_fm_cholesky_decompose 22 10.9 2.083 2.100 2.083 2.100 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 2.023 2.091 qs_create_task_list 11 7.9 0.001 0.002 1.897 1.944 generate_qs_task_list 11 8.9 0.740 0.794 1.896 1.943 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.847 1.849 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.824 1.846 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=87.835000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1246.818182, yerr=72.108275 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a374e79360e402e0e0c3cfd3175b21c36af7649d_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430458527744 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 11613083000832 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 94.228655E+12 0.0% 0.0% 100.0% flops max/rank 1.103326E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806314816 0.0% 0.0% 100.0% number of processed stacks 12044928 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.1 marketing flops 145.647559E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 625.721344E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10431456 MPI messages size (bytes): total size 4.526012E+12 min size 0.000000E+00 max size 4.537280E+06 average size 433.881094E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3605352 96403587072 32768 < size <= 131072 1305088 74666999808 131072 < size <= 4194304 5190114 3200148065288 4194304 < size <= 16777216 263934 1154745839544 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4113 56379. MP_Allreduce 11323 782. MP_Sync 170 MP_Alltoall 2242 2668832. MP_ISendRecv 49020 18752. MP_Wait 67312 MP_comm_split 83 MP_ISend 16140 108019. MP_IRecv 16140 108019. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.049 0.062 203.467 203.468 qs_mol_dyn_low 1 2.0 0.003 0.004 202.937 202.951 qs_forces 11 3.9 0.003 0.003 202.852 202.853 qs_energies 11 4.9 0.001 0.002 197.384 197.398 scf_env_do_scf 11 5.9 0.001 0.001 180.359 180.363 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 159.710 159.712 dbcsr_multiply_generic 2527 12.6 0.179 0.183 123.900 124.889 velocity_verlet 10 3.0 0.001 0.001 123.023 123.024 qs_scf_new_mos 118 7.6 0.001 0.001 120.738 121.076 qs_scf_loop_do_ot 118 8.6 0.001 0.001 120.737 121.075 ot_scf_mini 118 9.6 0.003 0.003 114.120 114.446 multiply_cannon 2527 13.6 0.250 0.329 101.236 104.025 multiply_cannon_loop 2527 14.6 2.089 2.197 98.937 101.491 ot_mini 118 10.6 0.001 0.001 65.149 65.487 multiply_cannon_multrec 60648 15.6 33.761 35.505 41.930 43.971 qs_ot_get_derivative 118 11.6 0.001 0.001 40.128 40.475 rebuild_ks_matrix 129 8.3 0.001 0.001 33.445 33.946 qs_ks_build_kohn_sham_matrix 129 9.3 0.015 0.017 33.445 33.946 mp_waitall_1 293764 16.2 27.864 31.371 27.864 31.371 qs_ks_update_qs_env 129 7.6 0.001 0.001 30.129 30.576 multiply_cannon_sync_h2d 60648 15.6 28.244 30.358 28.244 30.358 qs_ot_get_p 129 10.4 0.001 0.001 27.075 27.375 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 24.543 25.556 apply_single 129 13.6 0.001 0.001 24.543 25.555 ot_diis_step 118 11.6 0.008 0.008 24.692 24.693 init_scf_loop 11 6.9 0.000 0.000 20.567 20.568 qs_ot_p2m_diag 83 11.4 0.077 0.091 20.466 20.520 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 18.215 18.461 cp_dbcsr_syevd 83 12.4 0.004 0.005 18.195 18.196 multiply_cannon_metrocomm3 60648 15.6 0.118 0.124 15.008 17.109 prepare_preconditioner 11 7.9 0.000 0.000 15.963 16.006 make_preconditioner 11 8.9 0.000 0.000 15.963 16.006 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.207 15.390 cp_fm_diag_elpa 83 13.4 0.000 0.001 15.333 15.338 cp_fm_redistribute_end 83 14.4 12.081 15.257 12.093 15.260 cp_fm_diag_elpa_base 83 14.4 3.121 15.032 3.155 15.153 make_m2s 5054 13.6 0.104 0.110 14.045 14.322 make_images 5054 14.6 0.404 0.422 13.862 14.153 sum_up_and_integrate 129 10.3 0.090 0.107 13.951 13.966 integrate_v_rspace 129 11.3 0.003 0.004 13.861 13.879 init_scf_run 11 5.9 0.000 0.001 12.822 12.822 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.821 12.822 qs_rho_update_rho_low 129 7.7 0.001 0.001 12.686 12.793 calculate_rho_elec 129 8.7 0.045 0.064 12.685 12.792 wfi_extrapolate 11 7.9 0.001 0.001 9.124 9.124 cp_fm_cholesky_invert 11 10.9 9.089 9.097 9.089 9.097 mp_sum_l 7930 13.1 7.333 8.700 7.333 8.700 calculate_dm_sparse 129 9.5 0.001 0.001 8.468 8.582 dbcsr_mm_accdrv_process 125468 16.2 3.103 3.458 7.716 8.463 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 8.110 8.319 make_images_data 5054 15.6 0.068 0.074 6.872 7.824 qs_ot_get_orbitals 118 10.6 0.001 0.001 7.699 7.759 hybrid_alltoall_any 5240 16.5 0.295 2.283 6.000 7.504 grid_integrate_task_list 129 12.3 7.154 7.492 7.154 7.492 multiply_cannon_metrocomm1 60648 15.6 0.088 0.091 5.892 7.316 pw_transfer 1559 11.6 0.075 0.104 6.839 7.125 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.984 6.994 density_rs2pw 129 9.7 0.006 0.007 6.448 6.950 fft_wrap_pw1pw2 1301 12.7 0.010 0.014 6.635 6.891 rs_pw_transfer 1054 12.0 0.016 0.018 5.585 6.155 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 6.029 6.152 fft_wrap_pw1pw2_140 527 13.2 0.449 0.514 5.694 5.867 fft3d_ps 1301 14.7 2.112 2.584 5.419 5.611 mp_alltoall_d11v 2423 14.1 4.182 5.303 4.182 5.303 grid_collocate_task_list 129 9.7 4.762 5.156 4.762 5.156 cp_fm_cholesky_decompose 22 10.9 4.689 4.701 4.689 4.701 potential_pw2rs 129 12.3 0.009 0.011 4.408 4.437 mp_sum_d 4492 12.2 3.489 4.124 3.489 4.124 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=203.468000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=592.363636, yerr=5.693261 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a374e79360e402e0e0c3cfd3175b21c36af7649d_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 2.166472E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 5925696 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.6 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 830.951424E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2385600 MPI messages size (bytes): total size 4.069300E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.705776E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70188 2295595008 32768 < size <= 131072 716032 54973693952 131072 < size <= 4194304 1363760 1386318135296 4194304 < size <= 16777216 153648 1453842923456 16777216 < size 67056 1171888537600 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4087 57173. MP_Allreduce 11257 946. MP_Sync 168 MP_Alltoall 1955 5598961. MP_ISendRecv 23876 47072. MP_Wait 37656 MP_comm_split 82 MP_ISend 11660 212488. MP_IRecv 11660 212488. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.041 0.056 188.684 188.685 qs_mol_dyn_low 1 2.0 0.003 0.003 188.192 188.206 qs_forces 11 3.9 0.004 0.005 188.109 188.111 qs_energies 11 4.9 0.009 0.016 181.398 181.410 scf_env_do_scf 11 5.9 0.001 0.001 164.989 164.998 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 131.805 131.808 velocity_verlet 10 3.0 0.001 0.001 118.185 118.186 dbcsr_multiply_generic 2485 12.5 0.185 0.195 96.599 97.885 qs_scf_new_mos 116 7.6 0.001 0.001 93.625 94.119 qs_scf_loop_do_ot 116 8.6 0.001 0.001 93.624 94.118 ot_scf_mini 116 9.6 0.004 0.004 88.871 89.516 multiply_cannon 2485 13.5 0.477 0.531 76.774 81.270 multiply_cannon_loop 2485 14.5 1.246 1.290 73.522 76.225 ot_mini 116 10.6 0.001 0.001 49.228 49.790 mp_waitall_1 224796 16.4 24.615 37.680 24.615 37.680 multiply_cannon_multrec 29820 15.5 21.937 26.684 31.517 36.760 init_scf_loop 11 6.9 0.000 0.000 33.093 33.094 rebuild_ks_matrix 127 8.3 0.001 0.001 32.483 33.065 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.020 32.483 33.065 qs_ks_update_qs_env 127 7.6 0.001 0.001 29.243 29.774 prepare_preconditioner 11 7.9 0.000 0.000 28.699 28.759 make_preconditioner 11 8.9 0.000 0.000 28.699 28.759 multiply_cannon_metrocomm3 29820 15.5 0.092 0.098 15.481 28.505 qs_ot_get_derivative 116 11.6 0.001 0.002 27.555 28.188 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.365 27.912 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 21.795 22.870 apply_single 127 13.6 0.001 0.001 21.795 22.870 qs_ot_get_p 127 10.4 0.001 0.001 21.540 22.135 ot_diis_step 116 11.6 0.014 0.015 21.501 21.503 multiply_cannon_sync_h2d 29820 15.5 19.143 21.446 19.143 21.446 qs_ot_p2m_diag 82 11.4 0.185 0.213 16.774 16.806 cp_fm_cholesky_invert 11 10.9 16.750 16.762 16.750 16.762 make_m2s 4970 13.5 0.089 0.094 14.161 15.724 cp_dbcsr_syevd 82 12.4 0.005 0.005 15.653 15.654 make_images 4970 14.5 1.136 1.317 13.952 15.517 sum_up_and_integrate 127 10.3 0.116 0.133 14.223 14.254 integrate_v_rspace 127 11.3 0.003 0.003 14.106 14.144 qs_rho_update_rho_low 127 7.7 0.001 0.001 12.784 12.816 calculate_rho_elec 127 8.7 0.087 0.104 12.784 12.816 cp_fm_diag_elpa 82 13.4 0.000 0.001 12.596 12.604 cp_fm_redistribute_end 82 14.4 7.320 12.527 7.336 12.534 cp_fm_diag_elpa_base 82 14.4 4.967 12.084 5.181 12.444 init_scf_run 11 5.9 0.000 0.001 11.644 11.646 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.644 11.646 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 10.740 11.158 multiply_cannon_metrocomm4 27335 15.5 0.096 0.110 3.708 10.432 make_images_data 4970 15.5 0.064 0.071 8.650 10.359 mp_irecv_dv 68888 16.3 3.515 10.050 3.515 10.050 hybrid_alltoall_any 5155 16.4 0.340 1.499 7.333 9.881 dbcsr_mm_accdrv_process 61726 16.2 4.518 5.202 9.046 9.661 wfi_extrapolate 11 7.9 0.001 0.001 8.315 8.315 pw_transfer 1535 11.6 0.084 0.106 7.662 7.718 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 7.441 7.495 grid_integrate_task_list 127 12.3 7.133 7.493 7.133 7.493 density_rs2pw 127 9.7 0.006 0.007 6.623 7.252 cp_fm_cholesky_decompose 22 10.9 7.059 7.141 7.059 7.141 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.214 7.054 fft_wrap_pw1pw2_140 519 13.2 0.464 0.514 6.496 6.545 calculate_dm_sparse 127 9.5 0.001 0.001 6.407 6.533 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.288 6.298 rs_pw_transfer 1038 11.9 0.014 0.016 5.514 6.172 mp_sum_l 7804 13.0 3.897 5.955 3.897 5.955 fft3d_ps 1281 14.7 2.758 2.897 5.832 5.867 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.323 5.475 qs_ot_get_orbitals 116 10.6 0.001 0.001 5.250 5.332 grid_collocate_task_list 127 9.7 4.876 5.184 4.876 5.184 mp_allgather_i34 2485 14.5 1.830 5.034 1.830 5.034 mp_alltoall_d11v 2401 14.1 4.074 5.013 4.074 5.013 potential_pw2rs 127 12.3 0.015 0.017 4.759 4.780 dbcsr_complete_redistribute 393 12.7 0.867 0.976 3.251 4.104 mp_sum_d 4443 12.2 2.598 3.870 2.598 3.870 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=188.685000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=788.818182, yerr=1.991718 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a374e79360e402e0e0c3cfd3175b21c36af7649d_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 2.928533E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 3984192 0.0% 0.0% 100.0% average stack size 0.0 0.0 1695.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 963.600384E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1042912 MPI messages size (bytes): total size 2.716210E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.604448E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 281856 36943429632 131072 < size <= 4194304 660064 996105256960 4194304 < size <= 16777216 65632 931531265168 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4085 57197. MP_Allreduce 11251 985. MP_Sync 170 MP_Alltoall 1712 9388896. MP_ISendRecv 15872 75008. MP_Wait 29756 MP_comm_split 83 MP_ISend 11748 275205. MP_IRecv 11748 275205. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.095 0.114 176.622 176.623 qs_mol_dyn_low 1 2.0 0.003 0.003 175.850 175.864 qs_forces 11 3.9 0.003 0.003 175.736 175.741 qs_energies 11 4.9 0.004 0.024 169.287 169.297 scf_env_do_scf 11 5.9 0.001 0.001 153.610 153.611 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 118.115 118.116 velocity_verlet 10 3.0 0.001 0.001 113.091 113.093 dbcsr_multiply_generic 2507 12.6 0.182 0.186 81.477 82.509 qs_scf_new_mos 117 7.6 0.001 0.001 82.037 82.414 qs_scf_loop_do_ot 117 8.6 0.001 0.002 82.036 82.413 ot_scf_mini 117 9.6 0.005 0.014 77.894 78.315 multiply_cannon 2507 13.6 0.498 0.515 61.562 65.887 multiply_cannon_loop 2507 14.6 0.859 0.890 58.399 61.011 ot_mini 117 10.6 0.001 0.001 42.605 43.043 init_scf_loop 11 6.9 0.001 0.006 35.396 35.397 mp_waitall_1 178456 16.5 25.052 34.483 25.052 34.483 prepare_preconditioner 11 7.9 0.000 0.000 31.424 31.473 make_preconditioner 11 8.9 0.000 0.000 31.424 31.473 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.066 30.461 rebuild_ks_matrix 128 8.3 0.001 0.001 29.712 30.160 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.017 29.712 30.160 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.858 27.268 multiply_cannon_multrec 20056 15.6 13.484 16.931 22.144 25.622 multiply_cannon_metrocomm3 20056 15.6 0.061 0.065 15.046 24.253 qs_ot_get_derivative 117 11.6 0.001 0.002 22.866 23.284 qs_ot_get_p 128 10.4 0.001 0.001 20.684 21.140 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.838 20.815 apply_single 128 13.6 0.001 0.001 19.838 20.814 ot_diis_step 117 11.6 0.018 0.018 19.625 19.625 qs_ot_p2m_diag 83 11.4 0.266 0.272 16.407 16.418 multiply_cannon_sync_h2d 20056 15.6 14.302 16.045 14.302 16.045 make_m2s 5014 13.6 0.081 0.086 14.795 15.707 make_images 5014 14.6 1.172 1.261 14.561 15.462 cp_dbcsr_syevd 83 12.4 0.005 0.005 15.370 15.372 cp_fm_cholesky_invert 11 10.9 14.748 14.759 14.748 14.759 sum_up_and_integrate 128 10.3 0.132 0.145 13.954 13.976 integrate_v_rspace 128 11.3 0.003 0.004 13.821 13.850 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.846 12.892 calculate_rho_elec 128 8.7 0.132 0.146 12.845 12.892 cp_fm_diag_elpa 83 13.4 0.000 0.001 12.238 12.239 cp_fm_redistribute_end 83 14.4 4.598 12.165 4.615 12.168 cp_fm_diag_elpa_base 83 14.4 7.122 11.577 7.530 12.063 init_scf_run 11 5.9 0.000 0.001 10.511 10.514 scf_env_initial_rho_setup 11 6.9 0.000 0.001 10.511 10.514 make_images_data 5014 15.6 0.061 0.067 9.047 10.329 hybrid_alltoall_any 5200 16.5 0.435 1.986 7.873 9.540 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.013 9.303 multiply_cannon_metrocomm4 17549 15.6 0.059 0.069 3.484 8.940 mp_irecv_dv 50230 16.2 3.364 8.679 3.364 8.679 dbcsr_mm_accdrv_process 41502 16.2 4.467 5.250 8.114 8.266 grid_integrate_task_list 128 12.3 7.326 7.748 7.326 7.748 cp_fm_cholesky_decompose 22 10.9 7.533 7.600 7.533 7.600 pw_transfer 1547 11.6 0.084 0.104 7.473 7.576 cp_fm_upper_to_full 105 14.5 5.740 7.482 5.740 7.482 wfi_extrapolate 11 7.9 0.001 0.001 7.365 7.365 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.251 7.360 density_rs2pw 128 9.7 0.006 0.006 6.349 6.756 fft_wrap_pw1pw2_140 523 13.2 0.477 0.528 6.332 6.449 dbcsr_complete_redistribute 395 12.7 1.187 1.233 4.535 6.294 calculate_dm_sparse 128 9.5 0.001 0.001 5.760 5.839 fft3d_ps 1291 14.7 2.693 2.897 5.522 5.587 rs_pw_transfer 1046 11.9 0.013 0.014 5.123 5.526 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.425 5.430 grid_collocate_task_list 128 9.7 5.072 5.412 5.072 5.412 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.574 5.275 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.338 5.095 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.514 4.638 mp_allgather_i34 2507 14.6 1.688 4.554 1.688 4.554 mp_sum_l 7870 13.0 3.297 4.521 3.297 4.521 mp_alltoall_d11v 2415 14.1 4.013 4.451 4.013 4.451 potential_pw2rs 128 12.3 0.020 0.022 4.411 4.426 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.060 4.091 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.338 4.032 qs_energies_init_hamiltonians 11 5.9 0.027 0.038 3.848 3.859 mp_alltoall_i22 716 14.1 1.969 3.812 1.969 3.812 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=176.623000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=904.727273, yerr=20.525332 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a374e79360e402e0e0c3cfd3175b21c36af7649d_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410023282688 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444707676160 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796579E+12 0.0% 0.0% 100.0% flops max/rank 4.320339E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500928 0.0% 0.0% 100.0% number of processed stacks 5927808 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.2 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.168581E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1133160 MPI messages size (bytes): total size 2.008142E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.772161E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 315952 35695099904 131072 < size <= 4194304 709496 778939400192 4194304 < size <= 16777216 69840 660837789680 16777216 < size 30480 532676608000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4085 57191. MP_Allreduce 11251 1068. MP_Sync 168 MP_Alltoall 1700 12496381. MP_ISendRecv 11684 75008. MP_Wait 28114 MP_comm_split 82 MP_ISend 14840 244848. MP_IRecv 14840 244848. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.093 0.179 186.899 186.900 qs_mol_dyn_low 1 2.0 0.003 0.004 186.221 186.235 qs_forces 11 3.9 0.005 0.006 186.101 186.109 qs_energies 11 4.9 0.005 0.026 179.097 179.109 scf_env_do_scf 11 5.9 0.002 0.010 162.087 162.103 velocity_verlet 10 3.0 0.001 0.001 123.278 123.281 scf_env_do_scf_inner_loop 116 6.6 0.003 0.010 114.987 114.989 qs_scf_new_mos 116 7.6 0.001 0.001 79.799 80.081 qs_scf_loop_do_ot 116 8.6 0.001 0.001 79.799 80.080 dbcsr_multiply_generic 2485 12.5 0.190 0.199 79.013 79.638 ot_scf_mini 116 9.6 0.003 0.004 75.349 75.653 multiply_cannon 2485 13.5 0.567 0.617 54.299 56.551 multiply_cannon_loop 2485 14.5 1.179 1.203 50.621 52.183 init_scf_loop 11 6.9 0.001 0.002 46.975 46.975 prepare_preconditioner 11 7.9 0.000 0.000 42.834 42.864 make_preconditioner 11 8.9 0.000 0.002 42.834 42.864 ot_mini 116 10.6 0.001 0.001 41.951 42.234 make_full_inverse_cholesky 11 9.9 0.000 0.000 36.356 41.440 multiply_cannon_multrec 29820 15.5 14.034 19.010 25.867 30.302 rebuild_ks_matrix 127 8.3 0.001 0.001 28.835 29.034 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.018 28.835 29.033 mp_waitall_1 152434 16.5 17.686 27.719 17.686 27.719 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.057 26.240 qs_ot_get_derivative 116 11.6 0.001 0.002 22.544 22.843 make_m2s 4970 13.5 0.095 0.103 20.388 21.336 make_images 4970 14.5 1.932 2.274 20.083 21.032 qs_ot_get_p 127 10.4 0.001 0.001 19.741 20.033 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 18.878 19.425 apply_single 127 13.6 0.001 0.001 18.878 19.425 ot_diis_step 116 11.6 0.017 0.018 19.285 19.287 cp_fm_upper_to_full 104 14.7 11.216 16.632 11.216 16.632 cp_fm_cholesky_invert 11 10.9 16.400 16.409 16.400 16.409 qs_ot_p2m_diag 82 11.4 0.339 0.385 15.630 15.681 multiply_cannon_metrocomm3 29820 15.5 0.047 0.050 6.475 15.284 cp_dbcsr_syevd 82 12.4 0.005 0.005 14.357 14.358 sum_up_and_integrate 127 10.3 0.140 0.151 13.960 13.987 integrate_v_rspace 127 11.3 0.003 0.004 13.820 13.852 qs_rho_update_rho_low 127 7.7 0.001 0.001 12.973 13.010 calculate_rho_elec 127 8.7 0.173 0.189 12.972 13.009 dbcsr_complete_redistribute 393 12.7 1.519 1.661 9.118 12.929 make_images_data 4970 15.5 0.063 0.066 11.040 12.763 multiply_cannon_sync_h2d 29820 15.5 11.629 12.370 11.629 12.370 dbcsr_mm_accdrv_process 61748 16.2 7.365 8.389 11.412 11.880 hybrid_alltoall_any 5155 16.4 0.519 2.186 9.887 11.636 copy_fm_to_dbcsr 208 11.6 0.002 0.002 7.741 11.554 cp_fm_diag_elpa 82 13.4 0.000 0.001 11.309 11.311 cp_fm_redistribute_end 82 14.4 1.924 11.241 1.938 11.246 cp_fm_diag_elpa_base 82 14.4 8.692 10.649 9.278 11.141 init_scf_run 11 5.9 0.000 0.001 10.914 10.915 scf_env_initial_rho_setup 11 6.9 0.000 0.002 10.913 10.915 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.454 10.180 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.326 9.552 mp_alltoall_i22 712 14.1 5.682 9.518 5.682 9.518 pw_transfer 1535 11.6 0.084 0.101 7.813 7.884 grid_integrate_task_list 127 12.3 7.482 7.795 7.482 7.795 cp_fm_cholesky_decompose 22 10.9 7.653 7.763 7.653 7.763 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 7.590 7.663 wfi_extrapolate 11 7.9 0.001 0.001 7.599 7.599 multiply_cannon_metrocomm4 24850 15.5 0.074 0.085 2.726 7.071 mp_irecv_dv 75445 16.2 2.583 6.800 2.583 6.800 fft_wrap_pw1pw2_140 519 13.2 0.478 0.485 6.701 6.788 density_rs2pw 127 9.7 0.005 0.006 6.166 6.589 calculate_dm_sparse 127 9.5 0.001 0.001 6.149 6.231 fft3d_ps 1281 14.7 2.772 2.851 5.839 5.908 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.626 5.688 grid_collocate_task_list 127 9.7 5.169 5.523 5.169 5.523 mp_alltoall_d11v 2401 14.1 4.774 5.284 4.774 5.284 rs_pw_transfer 1038 11.9 0.013 0.014 4.607 5.008 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.507 4.589 qs_energies_init_hamiltonians 11 5.9 0.015 0.022 4.554 4.559 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.351 4.435 potential_pw2rs 127 12.3 0.022 0.023 4.243 4.263 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.156 4.204 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=186.900000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1090.000000, yerr=30.582229 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a374e79360e402e0e0c3cfd3175b21c36af7649d_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 5.820059E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 1944496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3448.5 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.508393E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 238560 MPI messages size (bytes): total size 1.321104E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.537828E+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 112800 59139686400 4194304 < size <= 16777216 104112 545846722560 16777216 < size 20064 716108580288 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8852 52. MP_Alltoall 9584 804353. MP_ISend 39716 2104723. MP_IRecv 39716 2103824. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4007 58136. MP_Allreduce 11015 1175. MP_Sync 86 MP_Alltoall 1700 18828153. MP_ISendRecv 7620 122880. MP_Wait 19810 MP_ISend 10600 423612. MP_IRecv 10600 423612. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.048 170.523 170.523 qs_mol_dyn_low 1 2.0 0.003 0.004 169.966 169.999 qs_forces 11 3.9 0.003 0.003 169.872 169.875 qs_energies 11 4.9 0.001 0.002 162.591 162.600 scf_env_do_scf 11 5.9 0.001 0.001 145.427 145.437 velocity_verlet 10 3.0 0.001 0.001 110.845 110.849 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 109.757 109.758 qs_scf_new_mos 116 7.6 0.001 0.001 74.650 74.753 qs_scf_loop_do_ot 116 8.6 0.001 0.001 74.650 74.753 dbcsr_multiply_generic 2485 12.5 0.182 0.187 73.643 73.969 ot_scf_mini 116 9.6 0.004 0.004 70.194 70.251 multiply_cannon 2485 13.5 0.580 0.617 53.900 58.321 multiply_cannon_loop 2485 14.5 0.443 0.451 49.259 49.865 ot_mini 116 10.6 0.001 0.001 39.277 39.340 init_scf_loop 11 6.9 0.000 0.000 35.516 35.518 mp_waitall_1 128490 16.6 25.843 33.019 25.843 33.019 prepare_preconditioner 11 7.9 0.000 0.000 31.683 31.720 make_preconditioner 11 8.9 0.000 0.000 31.683 31.720 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.619 29.885 rebuild_ks_matrix 127 8.3 0.001 0.001 28.164 28.260 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.017 28.164 28.259 qs_ks_update_qs_env 127 7.6 0.001 0.001 25.704 25.788 multiply_cannon_multrec 9940 15.5 10.334 14.157 17.558 20.794 qs_ot_get_derivative 116 11.6 0.002 0.002 19.619 19.685 ot_diis_step 116 11.6 0.019 0.022 19.581 19.581 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.244 19.483 apply_single 127 13.6 0.001 0.001 19.244 19.483 multiply_cannon_metrocomm3 9940 15.5 0.023 0.024 12.144 18.906 make_m2s 4970 13.5 0.066 0.071 16.087 18.210 qs_ot_get_p 127 10.4 0.001 0.001 18.066 18.143 cp_fm_cholesky_invert 11 10.9 18.115 18.121 18.115 18.121 make_images 4970 14.5 2.299 2.802 15.782 17.908 qs_ot_p2m_diag 82 11.4 0.489 0.495 14.309 14.326 sum_up_and_integrate 127 10.3 0.180 0.191 13.815 13.862 integrate_v_rspace 127 11.3 0.004 0.004 13.635 13.691 cp_dbcsr_syevd 82 12.4 0.005 0.005 13.187 13.187 qs_rho_update_rho_low 127 7.7 0.001 0.001 12.957 13.013 calculate_rho_elec 127 8.7 0.256 0.266 12.956 13.013 make_images_data 4970 15.5 0.053 0.059 9.790 12.199 multiply_cannon_sync_h2d 9940 15.5 11.511 12.085 11.511 12.085 hybrid_alltoall_any 5155 16.4 0.834 3.779 9.481 12.078 init_scf_run 11 5.9 0.000 0.001 10.310 10.310 scf_env_initial_rho_setup 11 6.9 0.000 0.001 10.310 10.310 cp_fm_diag_elpa 82 13.4 0.000 0.000 10.205 10.206 cp_fm_diag_elpa_base 82 14.4 9.967 10.038 10.201 10.202 cp_fm_cholesky_decompose 22 10.9 7.974 8.077 7.974 8.077 qs_ot_get_derivative_diag 76 12.4 0.002 0.003 7.957 8.006 grid_integrate_task_list 127 12.3 7.708 7.937 7.708 7.937 pw_transfer 1535 11.6 0.082 0.088 7.644 7.662 dbcsr_mm_accdrv_process 20590 16.1 2.640 3.411 6.860 7.472 fft_wrap_pw1pw2 1281 12.7 0.010 0.010 7.425 7.442 multiply_cannon_metrocomm1 9940 15.5 0.028 0.029 4.474 7.277 wfi_extrapolate 11 7.9 0.001 0.001 7.258 7.258 mp_allgather_i34 2485 14.5 2.861 7.046 2.861 7.046 fft_wrap_pw1pw2_140 519 13.2 0.499 0.522 6.542 6.566 density_rs2pw 127 9.7 0.005 0.005 5.859 6.234 calculate_dm_sparse 127 9.5 0.001 0.001 6.137 6.204 grid_collocate_task_list 127 9.7 5.434 5.741 5.434 5.741 fft3d_ps 1281 14.7 2.706 2.770 5.600 5.633 dbcsr_complete_redistribute 393 12.7 2.119 2.165 5.159 5.543 qs_energies_init_hamiltonians 11 5.9 0.040 0.060 5.276 5.284 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.105 5.111 mp_alltoall_d11v 2401 14.1 4.521 4.817 4.521 4.817 rs_pw_transfer 1038 11.9 0.012 0.013 4.058 4.459 multiply_cannon_metrocomm4 7455 15.5 0.022 0.025 1.841 4.233 mp_irecv_dv 28618 15.9 1.805 4.163 1.805 4.163 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.025 4.066 potential_pw2rs 127 12.3 0.026 0.026 4.004 4.016 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.559 3.871 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.734 3.781 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.373 3.687 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.499 3.568 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.416 3.437 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=170.523000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1413.000000, yerr=45.803930 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a374e79360e402e0e0c3cfd3175b21c36af7649d_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 11.606413E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 1947808 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.6 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 2.950562E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 99400 MPI messages size (bytes): total size 1.127422E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.342275E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 44768 34745614336 4194304 < size <= 16777216 43984 376564613120 16777216 < size 10032 716108580288 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4003 59127. MP_Allreduce 11005 1515. MP_Sync 86 MP_Alltoall 1700 36954383. MP_ISendRecv 3556 218624. MP_Wait 11506 MP_ISend 6360 1080477. MP_IRecv 6360 1080477. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.028 0.050 287.504 287.505 qs_mol_dyn_low 1 2.0 0.003 0.004 286.728 286.745 qs_forces 11 3.9 0.003 0.003 286.626 286.629 qs_energies 11 4.9 0.002 0.002 278.136 278.146 scf_env_do_scf 11 5.9 0.001 0.001 256.245 256.259 velocity_verlet 10 3.0 0.001 0.001 206.969 206.978 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 131.847 131.848 init_scf_loop 11 6.9 0.000 0.000 124.147 124.149 prepare_preconditioner 11 7.9 0.000 0.000 119.434 119.459 make_preconditioner 11 8.9 0.000 0.000 119.434 119.459 make_full_inverse_cholesky 11 9.9 0.000 0.000 95.615 116.566 qs_scf_new_mos 116 7.6 0.001 0.001 90.513 90.585 qs_scf_loop_do_ot 116 8.6 0.001 0.001 90.512 90.584 ot_scf_mini 116 9.6 0.004 0.004 85.779 85.804 dbcsr_multiply_generic 2485 12.5 0.212 0.219 82.679 83.243 cp_fm_upper_to_full 104 14.8 52.845 75.643 52.845 75.643 multiply_cannon 2485 13.5 0.712 0.820 58.253 59.167 multiply_cannon_loop 2485 14.5 0.470 0.481 54.526 55.982 ot_mini 116 10.6 0.001 0.001 44.122 44.153 dbcsr_complete_redistribute 393 12.7 3.995 4.023 29.487 41.947 copy_fm_to_dbcsr 208 11.6 0.001 0.002 26.125 38.585 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 23.773 36.147 mp_alltoall_i22 712 14.1 21.585 34.015 21.585 34.015 cp_fm_cholesky_invert 11 10.9 33.647 33.653 33.647 33.653 rebuild_ks_matrix 127 8.3 0.001 0.001 33.016 33.081 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.017 33.016 33.081 mp_waitall_1 104546 16.7 28.242 31.869 28.242 31.869 qs_ks_update_qs_env 127 7.6 0.001 0.001 30.799 30.858 qs_ot_get_p 127 10.4 0.001 0.001 26.271 26.353 qs_ot_get_derivative 116 11.6 0.002 0.002 23.944 23.972 qs_ot_p2m_diag 82 11.4 0.867 0.872 22.225 22.255 make_m2s 4970 13.5 0.076 0.078 19.825 20.866 cp_dbcsr_syevd 82 12.4 0.005 0.006 20.521 20.522 make_images 4970 14.5 3.726 3.840 19.349 20.392 ot_diis_step 116 11.6 0.022 0.022 20.144 20.145 multiply_cannon_metrocomm3 9940 15.5 0.024 0.025 18.624 19.865 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.395 19.613 apply_single 127 13.6 0.001 0.001 19.395 19.612 multiply_cannon_multrec 9940 15.5 10.396 12.164 17.817 17.907 cp_fm_diag_elpa 82 13.4 0.000 0.000 17.347 17.347 cp_fm_diag_elpa_base 82 14.4 12.986 14.623 17.343 17.343 multiply_cannon_sync_h2d 9940 15.5 15.533 15.543 15.533 15.543 sum_up_and_integrate 127 10.3 0.318 0.320 15.428 15.510 integrate_v_rspace 127 11.3 0.004 0.004 15.109 15.190 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.746 14.763 calculate_rho_elec 127 8.7 0.479 0.479 14.745 14.762 hybrid_alltoall_any 5155 16.4 1.288 3.000 10.833 12.610 make_images_data 4970 15.5 0.059 0.063 10.727 12.527 init_scf_run 11 5.9 0.000 0.001 11.924 11.925 scf_env_initial_rho_setup 11 6.9 0.000 0.000 11.924 11.925 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.320 9.349 dbcsr_mm_accdrv_process 20590 16.0 3.689 5.310 7.186 8.960 cp_fm_cholesky_decompose 22 10.9 8.832 8.854 8.832 8.854 wfi_extrapolate 11 7.9 0.001 0.001 8.679 8.679 grid_integrate_task_list 127 12.3 8.478 8.663 8.478 8.663 pw_transfer 1535 11.6 0.089 0.090 8.177 8.189 qs_energies_init_hamiltonians 11 5.9 0.002 0.002 7.988 7.990 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 7.946 7.956 fft_wrap_pw1pw2_140 519 13.2 0.535 0.538 7.045 7.057 mp_alltoall_d11v 2401 14.1 6.778 6.911 6.778 6.911 calculate_dm_sparse 127 9.5 0.001 0.001 6.561 6.625 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.439 6.517 grid_collocate_task_list 127 9.7 6.290 6.335 6.290 6.335 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.013 6.075 fft3d_ps 1281 14.7 2.739 2.749 6.007 6.016 density_rs2pw 127 9.7 0.005 0.005 5.899 5.926 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=287.505000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2613.090909, yerr=172.109085 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a374e79360e402e0e0c3cfd3175b21c36af7649d_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 1.261769E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255669. MP_Allreduce 3059 6274. MP_Sync 4 MP_Alltoall 54 MP_ISendRecv 570 19200. MP_Wait 1302 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 0.079 0.114 85.735 85.736 qs_energies 1 2.0 0.002 0.019 85.026 85.042 ls_scf 1 3.0 0.000 0.001 83.894 83.901 dbcsr_multiply_generic 111 6.7 0.015 0.015 72.904 73.149 multiply_cannon 111 7.7 0.018 0.021 56.171 57.148 multiply_cannon_loop 111 8.7 0.205 0.219 52.721 53.861 ls_scf_main 1 4.0 0.000 0.001 52.365 52.366 density_matrix_trs4 2 5.0 0.002 0.005 46.863 46.943 ls_scf_init_scf 1 4.0 0.000 0.002 28.426 28.428 ls_scf_init_matrix_S 1 5.0 0.000 0.002 27.353 27.410 mp_waitall_1 11316 10.9 22.785 25.698 22.785 25.698 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.222 25.243 multiply_cannon_multrec 2664 9.7 8.194 9.111 15.497 17.435 multiply_cannon_sync_h2d 2664 9.7 13.604 15.376 13.604 15.376 make_m2s 222 7.7 0.008 0.011 13.121 13.688 make_images 222 8.7 0.099 0.110 13.099 13.667 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 9.940 12.334 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.499 8.910 make_images_data 222 9.7 0.004 0.005 7.656 8.152 dbcsr_mm_accdrv_process 4760 10.4 0.508 0.633 6.922 7.931 hybrid_alltoall_any 227 10.6 0.216 1.843 6.556 7.872 dbcsr_mm_accdrv_process_sort 4760 11.4 6.207 7.142 6.207 7.142 calculate_norms 4752 9.8 5.515 6.249 5.515 6.249 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.097 5.259 mp_sum_l 807 5.4 3.102 4.214 3.102 4.214 multiply_cannon_metrocomm4 2442 9.7 0.012 0.014 2.058 3.426 mp_irecv_dv 6231 10.9 2.040 3.399 2.040 3.399 make_images_sizes 222 9.7 0.000 0.000 0.726 3.386 mp_alltoall_i44 222 10.7 0.725 3.386 0.725 3.386 arnoldi_extremal 4 6.8 0.000 0.000 3.311 3.338 arnoldi_normal_ev 4 7.8 0.002 0.008 3.311 3.338 build_subspace 16 8.4 0.009 0.012 3.211 3.213 ls_scf_post 1 4.0 0.000 0.001 3.103 3.109 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.274 3.048 ls_scf_store_result 1 5.0 0.000 0.000 2.909 2.953 dbcsr_special_finalize 555 9.7 0.005 0.006 2.386 2.777 dbcsr_merge_single_wm 555 10.7 0.467 0.617 2.378 2.769 make_images_pack 222 9.7 2.206 2.640 2.208 2.641 dbcsr_matrix_vector_mult 304 9.0 0.003 0.010 2.330 2.567 dbcsr_sort_data 658 11.4 2.171 2.518 2.171 2.518 dbcsr_matrix_vector_mult_local 304 10.0 2.062 2.447 2.064 2.449 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.234 2.322 buffer_matrices_ensure_size 222 8.7 1.787 2.161 1.787 2.161 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.728 1.730 rebuild_ks_matrix 3 7.3 0.000 0.000 1.719 1.720 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.003 1.719 1.720 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.736000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1130.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a374e79360e402e0e0c3cfd3175b21c36af7649d_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 2.104869E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 10339. MP_Sync 4 MP_Alltoall 47 15335933. MP_ISendRecv 282 57600. MP_Wait 828 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 0.055 0.184 91.939 91.941 qs_energies 1 2.0 0.000 0.000 91.170 91.225 ls_scf 1 3.0 0.003 0.020 89.855 89.860 dbcsr_multiply_generic 111 6.7 0.015 0.016 75.808 76.200 multiply_cannon 111 7.7 0.029 0.045 53.470 57.065 ls_scf_main 1 4.0 0.000 0.002 55.370 55.379 multiply_cannon_loop 111 8.7 0.116 0.124 50.183 53.373 density_matrix_trs4 2 5.0 0.002 0.003 49.733 49.919 mp_waitall_1 9246 10.9 21.661 30.934 21.661 30.934 ls_scf_init_scf 1 4.0 0.000 0.002 30.917 30.918 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.741 29.850 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.318 27.338 multiply_cannon_multrec 1332 9.7 13.077 16.973 22.261 27.306 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 12.089 21.310 make_m2s 222 7.7 0.006 0.008 15.504 16.183 make_images 222 8.7 1.575 1.946 15.474 16.153 dbcsr_mm_accdrv_process 4041 10.4 0.273 0.448 8.784 10.365 dbcsr_mm_accdrv_process_sort 4041 11.4 8.379 9.917 8.379 9.917 make_images_data 222 9.7 0.004 0.004 8.942 9.893 hybrid_alltoall_any 227 10.6 0.521 2.476 8.234 9.525 mp_sum_l 807 5.4 5.626 9.062 5.626 9.062 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.235 7.780 mp_irecv_dv 3311 11.0 3.216 7.712 3.216 7.712 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.242 7.143 calculate_norms 2376 9.8 5.997 6.829 5.997 6.829 multiply_cannon_sync_h2d 1332 9.7 4.780 6.030 4.780 6.030 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.069 5.266 arnoldi_extremal 4 6.8 0.000 0.000 4.814 4.842 arnoldi_normal_ev 4 7.8 0.001 0.005 4.814 4.842 build_subspace 16 8.4 0.014 0.021 4.549 4.552 ls_scf_post 1 4.0 0.000 0.001 3.565 3.572 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.185 3.424 ls_scf_store_result 1 5.0 0.000 0.000 3.266 3.389 dbcsr_matrix_vector_mult_local 304 10.0 2.739 3.224 2.741 3.226 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.255 2.906 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.517 2.606 make_images_pack 222 9.7 2.023 2.393 2.025 2.395 mp_allgather_i34 111 8.7 0.984 2.320 0.984 2.320 dbcsr_sort_data 436 11.2 1.830 2.067 1.830 2.067 dbcsr_data_new 4174 10.1 1.612 1.848 1.612 1.848 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.840 1.842 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.941000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1726.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a374e79360e402e0e0c3cfd3175b21c36af7649d_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 2.682458E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265470. MP_Allreduce 3058 11181. MP_Sync 4 MP_Alltoall 47 23526250. MP_ISendRecv 186 57600. MP_Wait 732 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 0.032 0.048 93.998 93.999 qs_energies 1 2.0 0.000 0.000 93.446 93.450 ls_scf 1 3.0 0.000 0.001 92.022 92.055 dbcsr_multiply_generic 111 6.7 0.015 0.016 76.837 77.114 ls_scf_main 1 4.0 0.000 0.002 57.809 57.813 multiply_cannon 111 7.7 0.039 0.071 53.041 57.034 multiply_cannon_loop 111 8.7 0.100 0.114 49.481 53.034 density_matrix_trs4 2 5.0 0.002 0.003 51.920 52.074 mp_waitall_1 7374 11.0 24.205 33.701 24.205 33.701 ls_scf_init_scf 1 4.0 0.000 0.001 30.657 30.688 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.538 29.616 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 27.056 27.075 multiply_cannon_multrec 888 9.7 12.576 15.203 21.077 24.299 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.484 22.691 make_m2s 222 7.7 0.006 0.007 17.200 18.469 make_images 222 8.7 1.974 2.287 17.162 18.428 hybrid_alltoall_any 227 10.6 0.616 2.855 9.426 10.937 make_images_data 222 9.7 0.003 0.004 9.830 10.806 dbcsr_mm_accdrv_process 3754 10.4 0.251 0.415 8.023 9.235 dbcsr_mm_accdrv_process_sort 3754 11.4 7.654 8.821 7.654 8.821 mp_sum_l 807 5.4 5.034 8.595 5.034 8.595 multiply_cannon_sync_h2d 888 9.7 6.086 7.158 6.086 7.158 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.451 6.993 mp_irecv_dv 2335 11.1 2.437 6.949 2.437 6.949 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.502 6.639 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.785 6.527 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.937 5.148 arnoldi_extremal 4 6.8 0.000 0.000 5.062 5.088 arnoldi_normal_ev 4 7.8 0.001 0.006 5.062 5.088 build_subspace 16 8.4 0.014 0.020 4.757 4.763 calculate_norms 1584 9.8 4.268 4.594 4.268 4.594 mp_allgather_i34 111 8.7 1.402 3.918 1.402 3.918 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.435 3.760 dbcsr_matrix_vector_mult_local 304 10.0 3.021 3.597 3.023 3.599 ls_scf_post 1 4.0 0.000 0.001 3.556 3.562 ls_scf_store_result 1 5.0 0.000 0.000 3.279 3.382 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.744 2.850 dbcsr_sort_data 325 11.1 1.878 2.174 1.878 2.174 make_images_pack 222 9.7 1.814 2.136 1.817 2.139 make_images_sizes 222 9.7 0.000 0.000 1.057 2.138 mp_alltoall_i44 222 10.7 1.057 2.138 1.057 2.138 dbcsr_data_release 9322 10.9 1.315 2.005 1.315 2.005 dbcsr_finalize 304 7.8 0.026 0.032 1.613 1.925 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=93.999000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2182.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a374e79360e402e0e0c3cfd3175b21c36af7649d_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 3.410915E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 13371. MP_Sync 4 MP_Alltoall 47 30278988. MP_ISendRecv 138 86400. MP_Wait 600 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 0.052 0.087 98.643 98.644 qs_energies 1 2.0 0.000 0.000 97.554 97.560 ls_scf 1 3.0 0.000 0.002 95.799 95.805 dbcsr_multiply_generic 111 6.7 0.017 0.021 79.541 79.782 ls_scf_main 1 4.0 0.000 0.003 59.426 59.427 multiply_cannon 111 7.7 0.053 0.120 52.360 56.796 density_matrix_trs4 2 5.0 0.002 0.003 53.373 53.498 multiply_cannon_loop 111 8.7 0.114 0.126 47.323 50.791 ls_scf_init_scf 1 4.0 0.000 0.002 33.104 33.107 ls_scf_init_matrix_S 1 5.0 0.000 0.001 31.964 32.039 mp_waitall_1 6438 11.0 23.535 30.181 23.535 30.181 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.170 29.183 multiply_cannon_multrec 1332 9.7 14.096 17.180 21.969 24.432 make_m2s 222 7.7 0.007 0.009 21.080 22.487 make_images 222 8.7 3.125 3.592 21.030 22.439 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.916 17.860 make_images_data 222 9.7 0.004 0.004 11.781 13.287 hybrid_alltoall_any 227 10.6 0.796 3.797 11.100 12.976 dbcsr_mm_accdrv_process 3641 10.4 0.220 0.399 7.509 9.324 dbcsr_mm_accdrv_process_sort 3641 11.4 7.094 8.575 7.094 8.575 mp_sum_l 807 5.4 4.544 8.101 4.544 8.101 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.383 6.521 multiply_cannon_sync_h2d 1332 9.7 5.573 6.219 5.573 6.219 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.088 5.931 mp_irecv_dv 3229 10.9 2.065 5.880 2.065 5.880 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.659 5.536 arnoldi_extremal 4 6.8 0.000 0.000 5.430 5.448 arnoldi_normal_ev 4 7.8 0.001 0.005 5.430 5.447 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.920 5.123 build_subspace 16 8.4 0.014 0.021 5.088 5.098 calculate_norms 2376 9.8 4.170 4.536 4.170 4.536 mp_allgather_i34 111 8.7 2.163 4.474 2.163 4.474 dbcsr_matrix_vector_mult 304 9.0 0.006 0.016 3.653 3.952 dbcsr_matrix_vector_mult_local 304 10.0 3.187 3.688 3.189 3.690 dbcsr_sort_data 658 11.4 3.078 3.395 3.078 3.395 ls_scf_post 1 4.0 0.001 0.004 3.268 3.274 dbcsr_special_finalize 555 9.7 0.006 0.007 2.829 3.242 dbcsr_merge_single_wm 555 10.7 0.536 0.673 2.821 3.235 ls_scf_store_result 1 5.0 0.000 0.000 3.018 3.070 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.903 2.959 dbcsr_data_release 10477 10.7 1.580 2.377 1.580 2.377 dbcsr_finalize 304 7.8 0.049 0.061 1.795 1.986 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=98.644000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2705.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a374e79360e402e0e0c3cfd3175b21c36af7649d_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 4.638155E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265558. MP_Allreduce 3049 15663. MP_Sync 4 MP_Alltoall 47 46208988. MP_ISendRecv 90 115200. MP_Wait 573 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 0.061 0.073 101.331 101.331 qs_energies 1 2.0 0.000 0.000 100.394 100.400 ls_scf 1 3.0 0.000 0.000 98.466 98.471 dbcsr_multiply_generic 111 6.7 0.017 0.019 79.953 80.177 ls_scf_main 1 4.0 0.000 0.000 63.300 63.300 multiply_cannon 111 7.7 0.095 0.181 56.634 61.226 density_matrix_trs4 2 5.0 0.002 0.003 56.128 56.228 multiply_cannon_loop 111 8.7 0.070 0.077 52.208 54.369 mp_waitall_1 5481 11.0 28.140 33.655 28.140 33.655 ls_scf_init_scf 1 4.0 0.000 0.000 31.486 31.490 ls_scf_init_matrix_S 1 5.0 0.000 0.000 30.295 30.327 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.094 28.105 multiply_cannon_multrec 444 9.7 14.071 16.916 21.078 23.805 make_m2s 222 7.7 0.004 0.005 18.171 20.639 make_images 222 8.7 3.711 4.424 18.109 20.578 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 12.604 17.445 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 5.889 15.656 make_images_data 222 9.7 0.003 0.004 10.436 12.799 hybrid_alltoall_any 227 10.6 0.788 3.770 10.079 12.600 multiply_cannon_sync_h2d 444 9.7 6.571 8.428 6.571 8.428 dbcsr_mm_accdrv_process 3003 10.4 0.164 0.349 6.718 7.850 dbcsr_mm_accdrv_process_sort 3003 11.4 6.403 7.504 6.403 7.504 mp_allgather_i34 111 8.7 2.658 7.070 2.658 7.070 arnoldi_extremal 4 6.8 0.000 0.000 5.992 6.003 arnoldi_normal_ev 4 7.8 0.002 0.005 5.992 6.003 build_subspace 16 8.4 0.015 0.020 5.580 5.591 mp_sum_l 807 5.4 3.198 4.942 3.198 4.942 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.685 4.849 dbcsr_matrix_vector_mult 304 9.0 0.007 0.016 4.264 4.437 dbcsr_matrix_vector_mult_local 304 10.0 3.700 4.156 3.702 4.158 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.637 3.879 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.283 3.816 mp_irecv_dv 1241 11.2 1.619 3.813 1.619 3.813 make_images_sizes 222 9.7 0.000 0.000 1.106 3.776 mp_alltoall_i44 222 10.7 1.105 3.776 1.105 3.776 ls_scf_post 1 4.0 0.000 0.000 3.680 3.686 calculate_norms 792 9.8 3.528 3.663 3.528 3.663 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.449 3.547 ls_scf_store_result 1 5.0 0.000 0.000 3.469 3.500 dbcsr_finalize 304 7.8 0.062 0.078 2.203 2.333 dbcsr_merge_all 275 8.9 0.473 0.532 2.050 2.164 dbcsr_data_release 10123 10.8 1.339 2.032 1.339 2.032 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=101.331000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3607.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a374e79360e402e0e0c3cfd3175b21c36af7649d_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 8.722596E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284111. MP_Allreduce 3043 21950. MP_Sync 4 MP_Alltoall 47 88727262. MP_ISendRecv 84 732600. MP_Wait 309 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 0.114 0.151 106.679 106.679 qs_energies 1 2.0 0.000 0.000 105.231 105.243 ls_scf 1 3.0 0.000 0.000 102.330 102.342 dbcsr_multiply_generic 111 6.7 0.023 0.026 76.535 76.661 ls_scf_main 1 4.0 0.000 0.000 64.960 64.961 density_matrix_trs4 2 5.0 0.002 0.003 56.156 56.211 multiply_cannon 111 7.7 0.131 0.214 49.384 51.462 multiply_cannon_loop 111 8.7 0.067 0.070 45.941 46.488 ls_scf_init_scf 1 4.0 0.000 0.000 33.685 33.686 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.360 32.369 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.521 29.536 mp_waitall_1 4569 11.1 21.668 25.046 21.668 25.046 make_m2s 222 7.7 0.005 0.005 23.690 24.652 make_images 222 8.7 4.561 4.942 23.584 24.544 multiply_cannon_multrec 444 9.7 17.792 18.587 22.379 23.058 hybrid_alltoall_any 227 10.6 1.654 3.611 12.663 15.566 make_images_data 222 9.7 0.003 0.003 12.955 15.320 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.130 10.541 multiply_cannon_sync_h2d 444 9.7 8.847 8.895 8.847 8.895 arnoldi_extremal 4 6.8 0.000 0.000 7.382 7.403 arnoldi_normal_ev 4 7.8 0.004 0.010 7.382 7.403 build_subspace 16 8.4 0.026 0.037 6.833 6.843 dbcsr_matrix_vector_mult 304 9.0 0.009 0.025 5.481 5.624 dbcsr_matrix_vector_mult_local 304 10.0 5.037 5.332 5.040 5.335 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.010 5.220 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.784 4.865 dbcsr_mm_accdrv_process 1814 10.4 0.184 0.306 4.418 4.553 dbcsr_mm_accdrv_process_sort 1814 11.4 4.121 4.256 4.121 4.256 ls_scf_post 1 4.0 0.000 0.000 3.685 3.696 make_images_sizes 222 9.7 0.000 0.000 1.447 3.597 mp_alltoall_i44 222 10.7 1.447 3.597 1.447 3.597 ls_scf_store_result 1 5.0 0.000 0.000 3.401 3.407 calculate_norms 792 9.8 3.240 3.285 3.240 3.285 mp_allgather_i34 111 8.7 1.082 3.276 1.082 3.276 dbcsr_finalize 304 7.8 0.082 0.089 3.076 3.135 dbcsr_merge_all 275 8.9 0.889 0.911 2.863 2.913 dbcsr_complete_redistribute 5 7.6 1.449 1.481 2.776 2.885 qs_energies_init_hamiltonians 1 3.0 0.002 0.002 2.871 2.871 dbcsr_data_release 12724 10.6 2.319 2.867 2.319 2.867 matrix_ls_to_qs 2 6.0 0.000 0.000 2.424 2.541 dbcsr_sort_data 325 11.1 2.436 2.499 2.436 2.499 dbcsr_new_transposed 4 7.5 0.243 0.253 2.287 2.300 dbcsr_frobenius_norm 74 6.6 2.056 2.138 2.193 2.238 dbcsr_add_d 103 6.2 0.000 0.000 2.132 2.202 dbcsr_add_anytype 103 7.2 0.859 0.891 2.132 2.202 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.190 2.191 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=106.679000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6807.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: a374e79360e402e0e0c3cfd3175b21c36af7649d Summary: empty Status: OK