=== 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: ef71e334ee8f071d2456394b624a331795da3ecc ################# 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 2022.11.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.1.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.1, # SIRIUS 7.4.3, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Author: Matthias Krack (03.02.2023) # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${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=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.1.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.1 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(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/ef71e334ee8f071d2456394b624a331795da3ecc_performance_tests/01 job id: 45819095 --- 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/ef71e334ee8f071d2456394b624a331795da3ecc_performance_tests/02 job id: 45819098 --- 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/ef71e334ee8f071d2456394b624a331795da3ecc_performance_tests/03 job id: 45819099 --- 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/ef71e334ee8f071d2456394b624a331795da3ecc_performance_tests/04 job id: 45819101 --- 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/ef71e334ee8f071d2456394b624a331795da3ecc_performance_tests/05 job id: 45819103 --- 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/ef71e334ee8f071d2456394b624a331795da3ecc_performance_tests/06 job id: 45819105 --- 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/ef71e334ee8f071d2456394b624a331795da3ecc_performance_tests/07 job id: 45819106 --- 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/ef71e334ee8f071d2456394b624a331795da3ecc_performance_tests/08 job id: 45819107 --- 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/ef71e334ee8f071d2456394b624a331795da3ecc_performance_tests/09 job id: 45819109 --- 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/ef71e334ee8f071d2456394b624a331795da3ecc_performance_tests/10 job id: 45819110 --- 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/ef71e334ee8f071d2456394b624a331795da3ecc_performance_tests/11 job id: 45819112 --- 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/ef71e334ee8f071d2456394b624a331795da3ecc_performance_tests/12 job id: 45819113 --- 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/ef71e334ee8f071d2456394b624a331795da3ecc_performance_tests/13 job id: 45819115 --- 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/ef71e334ee8f071d2456394b624a331795da3ecc_performance_tests/14 job id: 45819117 --- 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/ef71e334ee8f071d2456394b624a331795da3ecc_performance_tests/15 job id: 45819118 --- 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/ef71e334ee8f071d2456394b624a331795da3ecc_performance_tests/16 job id: 45819119 --- 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/ef71e334ee8f071d2456394b624a331795da3ecc_performance_tests/17 job id: 45819122 --- 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/ef71e334ee8f071d2456394b624a331795da3ecc_performance_tests/18 job id: 45819123 --- 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/ef71e334ee8f071d2456394b624a331795da3ecc_performance_tests/19 job id: 45819125 --- 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/ef71e334ee8f071d2456394b624a331795da3ecc_performance_tests/20 job id: 45819127 --- 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/ef71e334ee8f071d2456394b624a331795da3ecc_performance_tests/21 job id: 45819129 --- 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/ef71e334ee8f071d2456394b624a331795da3ecc_performance_tests/22 job id: 45819130 --- 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/ef71e334ee8f071d2456394b624a331795da3ecc_performance_tests/23 job id: 45819132 --- 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/ef71e334ee8f071d2456394b624a331795da3ecc_performance_tests/24 job id: 45819133 --- 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/ef71e334ee8f071d2456394b624a331795da3ecc_performance_tests/25 job id: 45819135 --- 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/ef71e334ee8f071d2456394b624a331795da3ecc_performance_tests/26 job id: 45819136 --- Point --- name: 510 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/1r/12t) --- Point --- name: 511 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: 512 H2O (4 NVE MD steps on 64 nodes) input file: benchmarks/QS/00512_H2O/H2O-512_md.inp required files: [] output file: result.log # nodes = 64 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ef71e334ee8f071d2456394b624a331795da3ecc_performance_tests/27 job id: 45819138 --- Point --- name: 601 plot: h2o_512_md regex: CP2K label: (64n/12r/1t) --- Point --- name: 602 plot: h2o_512_md_mem regex: Estimated peak process memory label: (64n/12r/1t) ~~~~~~~ END TEST ~~~~~~~ === END TESTS (description) === ===== PLOTS (description) ===== ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2_mem", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md_mem", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md_mem", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md_mem", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls_mem", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_512_md", title="512 H2O (4 NVE MD steps on 64 nodes)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_512_md_mem", title="512 H2O (4 NVE MD steps on 64 nodes)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" === END PLOTS (description) === ============ RESULTS ============ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ef71e334ee8f071d2456394b624a331795da3ecc_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 0.021 0.037 135.289 135.290 farming_run 1 2.0 134.195 134.199 135.260 135.262 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.471189E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 230 1103589. MP_Allreduce 491 2254389. MP_Sync 25 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 10 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.006 0.027 115.219 115.219 qs_energies 1 2.0 0.000 0.000 114.988 114.990 mp2_main 1 3.0 0.000 0.000 112.922 112.924 mp2_gpw_main 1 4.0 0.020 0.026 111.910 111.912 mp2_ri_gpw_compute_in 1 5.0 0.172 0.176 93.285 93.435 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 55.527 55.677 mp2_eri_3c_integrate_gpw 272 7.0 0.155 0.172 41.741 46.753 get_2c_integrals 1 6.0 0.000 0.000 37.072 37.585 integrate_v_rspace 273 8.0 0.437 0.454 25.097 29.764 pw_transfer 6555 10.6 0.375 0.395 27.297 27.584 fft_wrap_pw1pw2 5465 11.4 0.044 0.047 25.964 26.251 grid_integrate_task_list 273 9.0 20.927 26.032 20.927 26.032 fft_wrap_pw1pw2_100 2178 12.4 1.186 1.309 23.488 23.795 compute_2c_integrals 1 7.0 0.002 0.003 19.402 19.403 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.807 19.126 mp2_eri_2c_integrate_gpw 1 9.0 2.382 2.431 18.804 19.123 rpa_ri_compute_en 1 5.0 0.002 0.024 18.517 18.641 cp_fm_cholesky_decompose 12 8.2 17.620 18.162 17.620 18.162 cholesky_decomp 1 7.0 0.000 0.000 16.525 17.075 fft3d_s 5443 13.4 16.088 16.246 16.111 16.269 ao_to_mo_and_store_B_mult_1 272 7.0 10.857 15.601 10.857 15.601 calculate_wavefunction 272 8.0 5.463 5.541 12.569 13.220 rpa_num_int 1 6.0 0.001 0.022 10.574 10.574 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.506 10.542 calc_mat_Q 8 8.0 0.000 0.000 9.365 9.459 contract_S_to_Q 8 9.0 0.000 0.000 8.789 8.881 calc_potential_gpw 544 9.5 0.005 0.006 8.237 8.606 parallel_gemm_fm 14 9.1 0.000 0.000 8.380 8.482 parallel_gemm_fm_cosma 14 10.1 8.380 8.482 8.380 8.482 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 8.170 8.433 potential_pw2rs 545 10.0 0.107 0.109 7.638 8.301 create_integ_mat 1 6.0 0.023 0.027 7.679 7.680 collocate_single_gaussian 272 10.0 0.040 0.042 7.435 7.662 array2fm 1 7.0 0.000 0.000 6.686 7.113 pw_scatter_s 2720 13.7 4.434 4.636 4.434 4.636 pw_gather_s 2722 13.2 3.849 4.278 3.849 4.278 array2fm_buffer_send 1 8.0 2.946 3.176 2.946 3.176 ao_to_mo_and_store_B_E_Ex_1 272 7.0 1.211 1.356 2.159 2.348 pw_poisson_solve 545 10.5 1.101 1.146 2.141 2.325 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.909156, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2742.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ef71e334ee8f071d2456394b624a331795da3ecc_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 0.051 0.063 397.793 397.794 farming_run 1 2.0 396.328 396.332 397.713 397.743 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.228612E+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 743 386399. MP_Allreduce 1941 22272. MP_Sync 37 MP_Alltoall 77 23480140. MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.009 0.028 209.649 209.650 qs_energies 1 2.0 0.000 0.000 209.419 209.431 scf_env_do_scf 1 3.0 0.000 0.000 106.295 106.295 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.398 105.406 rebuild_ks_matrix 4 6.0 0.000 0.000 105.396 105.405 qs_ks_build_kohn_sham_matrix 4 7.0 0.055 0.064 105.396 105.405 hfx_ks_matrix 4 8.0 0.001 0.001 105.000 105.005 integrate_four_center 4 9.0 0.143 0.454 105.000 105.004 mp2_main 1 3.0 0.000 0.000 102.834 102.846 mp2_gpw_main 1 4.0 0.034 0.056 101.972 101.986 integrate_four_center_main 4 10.0 0.096 0.484 96.747 99.614 integrate_four_center_bin 263 11.0 96.651 99.604 96.651 99.604 init_scf_loop 1 4.0 0.000 0.000 92.061 92.061 mp2_ri_gpw_compute_in 1 5.0 0.065 0.073 75.022 76.238 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.652 55.864 mp2_eri_3c_integrate_gpw 91 7.0 0.145 0.177 42.233 47.334 integrate_v_rspace 95 8.0 0.399 0.570 28.594 33.498 pw_transfer 2240 10.6 0.145 0.160 29.931 30.381 ao_to_mo_and_store_B_mult_1 91 7.0 10.738 29.738 10.738 29.738 fft_wrap_pw1pw2 1868 11.4 0.018 0.020 28.944 29.423 grid_integrate_task_list 95 9.0 23.868 28.979 23.868 28.979 mp2_ri_gpw_compute_en 1 5.0 0.056 0.068 26.800 28.698 fft_wrap_pw1pw2_100 730 12.4 1.267 1.457 26.635 27.078 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.846 1.927 24.939 24.949 get_2c_integrals 1 6.0 0.000 0.000 20.279 20.309 compute_2c_integrals 1 7.0 0.003 0.010 19.263 19.271 compute_2c_integrals_loop_lm 1 8.0 0.003 0.019 18.838 19.123 mp2_eri_2c_integrate_gpw 1 9.0 1.747 1.860 18.835 19.121 fft3d_s 1823 13.4 18.429 18.741 18.442 18.754 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.232 14.232 calculate_wavefunction 91 8.0 2.028 2.064 9.756 9.957 mp2_ri_gpw_compute_en_expansio 172 7.0 0.557 0.581 8.835 9.310 potential_pw2rs 186 10.0 0.033 0.035 8.621 9.306 local_gemm 172 8.0 8.278 8.748 8.278 8.748 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.215 8.580 mp2_ri_gpw_compute_en_comm 22 7.0 0.503 0.527 7.868 8.331 calc_potential_gpw 182 9.5 0.002 0.002 7.919 8.114 collocate_single_gaussian 91 10.0 0.017 0.023 7.874 8.078 mp2_ri_gpw_compute_en_ener 172 7.0 6.346 6.431 6.346 6.431 mp_sendrecv_dm3 2068 8.0 5.887 6.358 5.887 6.358 pw_gather_s 912 13.2 4.908 5.590 4.908 5.590 mp_sync 37 10.5 2.907 4.995 2.907 4.995 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.983430, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1516.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ef71e334ee8f071d2456394b624a331795da3ecc_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 452.718592E+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 272. MP_Sync 530 MP_Alltoall 2083 588782. MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.052 0.101 54.617 54.620 qs_mol_dyn_low 1 2.0 0.004 0.008 54.143 54.151 qs_forces 11 3.9 0.003 0.006 54.051 54.052 qs_energies 11 4.9 0.001 0.003 52.596 52.610 scf_env_do_scf 11 5.9 0.000 0.001 45.628 45.629 scf_env_do_scf_inner_loop 108 6.5 0.003 0.008 43.568 43.569 dbcsr_multiply_generic 2286 12.5 0.106 0.126 33.979 34.454 qs_scf_new_mos 108 7.5 0.000 0.001 33.582 33.886 qs_scf_loop_do_ot 108 8.5 0.000 0.001 33.582 33.886 ot_scf_mini 108 9.5 0.002 0.002 31.905 32.105 multiply_cannon 2286 13.5 0.189 0.197 26.572 28.464 multiply_cannon_loop 2286 14.5 1.463 1.544 25.917 27.785 velocity_verlet 10 3.0 0.001 0.002 25.707 25.710 ot_mini 108 10.5 0.001 0.001 19.305 19.557 qs_ot_get_derivative 108 11.5 0.001 0.001 16.388 16.563 mp_waitall_1 245248 16.5 8.637 15.189 8.637 15.189 multiply_cannon_metrocomm3 54864 15.5 0.069 0.075 6.201 13.581 multiply_cannon_multrec 54864 15.5 4.260 6.558 7.507 10.906 qs_ot_get_p 119 10.4 0.001 0.001 8.037 8.356 rebuild_ks_matrix 119 8.3 0.000 0.000 7.787 7.935 qs_ks_build_kohn_sham_matrix 119 9.3 0.011 0.012 7.787 7.935 mp_sum_l 7207 12.9 5.583 7.297 5.583 7.297 multiply_cannon_sync_h2d 54864 15.5 5.933 7.041 5.933 7.041 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.866 6.995 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.465 5.954 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.619 5.719 init_scf_run 11 5.9 0.000 0.001 5.621 5.622 scf_env_initial_rho_setup 11 6.9 0.000 0.000 5.621 5.621 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.282 5.341 dbcsr_mm_accdrv_process 76910 16.1 1.138 1.815 3.169 4.468 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.465 4.466 sum_up_and_integrate 119 10.3 0.012 0.014 4.426 4.434 integrate_v_rspace 119 11.3 0.002 0.002 4.414 4.423 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.280 4.281 cp_fm_redistribute_end 50 14.0 2.184 4.257 2.189 4.259 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.130 4.250 calculate_rho_elec 119 8.7 0.012 0.017 4.129 4.249 cp_fm_diag_elpa_base 50 14.0 2.065 4.154 2.069 4.163 calculate_first_density_matrix 1 7.0 0.001 0.004 3.482 3.491 calculate_dm_sparse 119 9.5 0.000 0.001 3.258 3.439 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.893 3.078 apply_single 119 13.6 0.000 0.000 2.893 3.078 multiply_cannon_metrocomm1 54864 15.5 0.053 0.058 1.844 2.976 acc_transpose_blocks 54864 15.5 0.231 0.256 2.255 2.823 rs_pw_transfer 974 11.9 0.012 0.013 2.679 2.820 ot_diis_step 108 11.5 0.006 0.006 2.658 2.658 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.005 2.623 2.625 jit_kernel_multiply 13 15.8 1.972 2.587 1.972 2.587 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.426 2.489 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.306 2.354 density_rs2pw 119 9.7 0.004 0.004 2.112 2.244 grid_integrate_task_list 119 12.3 2.024 2.118 2.024 2.118 wfi_extrapolate 11 7.9 0.001 0.001 2.077 2.077 init_scf_loop 11 6.9 0.000 0.000 2.045 2.045 mp_sum_d 4135 12.0 1.283 1.827 1.283 1.827 potential_pw2rs 119 12.3 0.004 0.004 1.785 1.796 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.738 1.773 pw_transfer 1439 11.6 0.052 0.056 1.656 1.727 acc_transpose_blocks_kernels 54864 16.5 0.246 0.383 1.288 1.657 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.581 1.654 make_m2s 4572 13.5 0.053 0.056 1.542 1.601 make_images 4572 14.5 0.133 0.138 1.460 1.519 mp_alltoall_d11v 2130 13.8 1.259 1.427 1.259 1.427 mp_waitany 12084 13.8 1.290 1.418 1.290 1.418 grid_collocate_task_list 119 9.7 1.298 1.365 1.298 1.365 fft3d_ps 1201 14.6 0.372 0.478 1.224 1.295 fft_wrap_pw1pw2_140 487 13.2 0.184 0.200 1.223 1.295 jit_kernel_transpose 5 15.5 1.041 1.283 1.041 1.283 parallel_gemm_fm 81 9.0 0.000 0.000 1.152 1.167 parallel_gemm_fm_cosma 81 10.0 1.151 1.167 1.151 1.167 dbcsr_dot_sd 1205 11.9 0.049 0.060 0.708 1.111 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=54.620000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=431.181818, yerr=1.113404 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ef71e334ee8f071d2456394b624a331795da3ecc_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 489.099264E+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 54 MP_Alltoall 2060 1400941. MP_SendRecv 16779 37093. MP_ISendRecv 16779 37093. MP_Wait 23539 MP_comm_split 50 MP_ISend 5720 128509. MP_IRecv 5720 128509. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.062 0.140 41.255 41.256 qs_mol_dyn_low 1 2.0 0.003 0.003 40.707 40.715 qs_forces 11 3.9 0.005 0.007 40.649 40.649 qs_energies 11 4.9 0.009 0.052 38.943 38.949 scf_env_do_scf 11 5.9 0.001 0.002 33.011 33.015 scf_env_do_scf_inner_loop 108 6.5 0.002 0.008 30.459 30.459 dbcsr_multiply_generic 2286 12.5 0.100 0.103 22.345 22.718 qs_scf_new_mos 108 7.5 0.001 0.001 21.089 21.332 qs_scf_loop_do_ot 108 8.5 0.001 0.001 21.088 21.331 ot_scf_mini 108 9.5 0.002 0.003 20.179 20.353 velocity_verlet 10 3.0 0.001 0.002 18.762 18.762 multiply_cannon 2286 13.5 0.208 0.214 17.213 18.717 multiply_cannon_loop 2286 14.5 0.888 0.966 16.065 17.526 ot_mini 108 10.5 0.001 0.001 12.534 12.772 mp_waitall_1 200699 16.5 6.246 11.970 6.246 11.970 multiply_cannon_metrocomm3 27432 15.5 0.067 0.070 4.736 10.405 qs_ot_get_derivative 108 11.5 0.001 0.001 10.060 10.236 multiply_cannon_multrec 27432 15.5 1.998 4.398 6.173 9.265 rebuild_ks_matrix 119 8.3 0.000 0.000 7.310 7.447 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 7.310 7.446 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.485 6.610 dbcsr_mm_accdrv_process 47894 16.0 3.069 5.570 4.105 6.340 qs_ot_get_p 119 10.4 0.001 0.001 4.705 4.944 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 4.093 4.929 sum_up_and_integrate 119 10.3 0.024 0.027 4.286 4.291 integrate_v_rspace 119 11.3 0.002 0.003 4.261 4.269 init_scf_run 11 5.9 0.000 0.001 4.223 4.225 scf_env_initial_rho_setup 11 6.9 0.003 0.024 4.222 4.224 mp_sum_l 7207 12.9 2.155 4.126 2.155 4.126 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.056 4.126 apply_single 119 13.6 0.000 0.000 3.056 4.126 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.897 3.923 calculate_rho_elec 119 8.7 0.021 0.024 3.897 3.922 rs_pw_transfer 974 11.9 0.010 0.011 2.757 3.236 qs_ot_p2m_diag 50 11.0 0.009 0.013 3.179 3.201 make_m2s 4572 13.5 0.052 0.054 2.576 2.807 multiply_cannon_sync_h2d 27432 15.5 2.205 2.796 2.205 2.796 cp_dbcsr_syevd 50 12.0 0.003 0.004 2.733 2.734 make_images 4572 14.5 0.200 0.238 2.488 2.717 density_rs2pw 119 9.7 0.004 0.004 2.219 2.707 calculate_first_density_matrix 1 7.0 0.001 0.002 2.692 2.699 init_scf_loop 11 6.9 0.003 0.022 2.531 2.535 ot_diis_step 108 11.5 0.010 0.011 2.422 2.423 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.382 2.383 cp_fm_redistribute_end 50 14.0 1.206 2.347 1.209 2.350 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.223 2.314 cp_fm_diag_elpa_base 50 14.0 1.107 2.246 1.136 2.288 jit_kernel_multiply 9 16.0 0.985 2.235 0.985 2.235 calculate_dm_sparse 119 9.5 0.000 0.001 2.046 2.123 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.108 2.110 pw_transfer 1439 11.6 0.065 0.069 2.012 2.056 potential_pw2rs 119 12.3 0.006 0.006 2.009 2.019 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.920 1.966 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.902 1.942 grid_integrate_task_list 119 12.3 1.829 1.915 1.829 1.915 acc_transpose_blocks 27432 15.5 0.108 0.112 1.240 1.764 make_images_data 4572 15.5 0.045 0.051 1.161 1.589 prepare_preconditioner 11 7.9 0.000 0.000 1.557 1.583 make_preconditioner 11 8.9 0.000 0.001 1.556 1.583 fft_wrap_pw1pw2_140 487 13.2 0.201 0.210 1.473 1.515 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.456 1.513 fft3d_ps 1201 14.6 0.522 0.578 1.472 1.512 mp_sum_d 4135 12.0 0.844 1.502 0.844 1.502 wfi_extrapolate 11 7.9 0.001 0.001 1.478 1.478 hybrid_alltoall_any 4725 16.4 0.051 0.111 1.010 1.470 mp_alltoall_d11v 2130 13.8 1.210 1.434 1.210 1.434 qs_energies_init_hamiltonians 11 5.9 0.007 0.016 1.379 1.382 mp_allgather_i34 2286 14.5 0.592 1.363 0.592 1.363 grid_collocate_task_list 119 9.7 1.226 1.354 1.226 1.354 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.211 1.262 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.243 1.251 acc_transpose_blocks_kernels 27432 16.5 0.181 0.270 0.714 1.150 rs_pw_transfer_RS2PW_140 130 11.5 0.143 0.154 0.594 1.081 mp_waitany 5720 13.7 0.568 1.069 0.568 1.069 qs_env_update_s_mstruct 11 6.9 0.186 0.373 0.739 0.954 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.901 0.914 jit_kernel_transpose 5 15.5 0.533 0.881 0.533 0.881 make_images_sizes 4572 15.5 0.005 0.005 0.603 0.845 mp_alltoall_i44 4572 16.5 0.599 0.841 0.599 0.841 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=41.256000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=465.090909, yerr=1.504813 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ef71e334ee8f071d2456394b624a331795da3ecc_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 523.571200E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63497. MP_Allreduce 10075 307. MP_Sync 54 MP_Alltoall 1821 1607820. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.097 0.178 34.682 34.708 qs_mol_dyn_low 1 2.0 0.003 0.003 33.095 33.103 qs_forces 11 3.9 0.015 0.026 33.030 33.031 qs_energies 11 4.9 0.001 0.001 31.439 31.453 scf_env_do_scf 11 5.9 0.000 0.001 26.035 26.035 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 23.328 23.328 dbcsr_multiply_generic 2286 12.5 0.094 0.097 16.845 16.938 velocity_verlet 10 3.0 0.001 0.002 15.295 15.296 qs_scf_new_mos 108 7.5 0.001 0.001 15.111 15.136 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.111 15.135 ot_scf_mini 108 9.5 0.002 0.002 14.387 14.414 multiply_cannon 2286 13.5 0.195 0.203 13.579 14.284 multiply_cannon_loop 2286 14.5 0.629 0.661 12.814 13.484 ot_mini 108 10.5 0.001 0.001 8.934 8.966 multiply_cannon_multrec 18288 15.5 1.975 3.105 7.147 7.517 qs_ot_get_derivative 108 11.5 0.001 0.001 7.428 7.453 rebuild_ks_matrix 119 8.3 0.000 0.000 6.445 6.464 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.445 6.463 dbcsr_mm_accdrv_process 38222 16.0 4.275 5.768 5.089 6.050 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.739 5.789 init_scf_run 11 5.9 0.000 0.001 4.214 4.214 scf_env_initial_rho_setup 11 6.9 0.000 0.000 4.213 4.213 sum_up_and_integrate 119 10.3 0.030 0.031 3.983 3.990 integrate_v_rspace 119 11.3 0.002 0.003 3.953 3.961 mp_waitall_1 158411 16.6 2.746 3.610 2.746 3.610 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.849 3.470 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.450 3.455 calculate_rho_elec 119 8.7 0.031 0.032 3.450 3.455 qs_ot_get_p 119 10.4 0.001 0.001 3.320 3.354 calculate_first_density_matrix 1 7.0 0.009 0.021 3.013 3.013 init_scf_loop 11 6.9 0.000 0.000 2.692 2.692 rs_pw_transfer 974 11.9 0.009 0.010 2.286 2.535 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.039 2.371 apply_single 119 13.6 0.000 0.000 2.038 2.371 jit_kernel_multiply 11 16.1 0.763 2.362 0.763 2.362 multiply_cannon_metrocomm3 18288 15.5 0.044 0.045 1.530 2.264 density_rs2pw 119 9.7 0.004 0.004 1.974 2.246 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.234 2.240 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.975 1.976 cp_dbcsr_syevd 50 12.0 0.003 0.004 1.971 1.971 pw_transfer 1439 11.6 0.065 0.069 1.917 1.927 make_m2s 4572 13.5 0.044 0.045 1.772 1.907 calculate_dm_sparse 119 9.5 0.000 0.001 1.873 1.897 grid_integrate_task_list 119 12.3 1.799 1.879 1.799 1.879 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.836 1.842 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.824 1.836 make_images 4572 14.5 0.189 0.202 1.687 1.821 prepare_preconditioner 11 7.9 0.000 0.000 1.763 1.767 make_preconditioner 11 8.9 0.000 0.000 1.763 1.767 potential_pw2rs 119 12.3 0.007 0.008 1.747 1.756 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.722 1.734 cp_fm_diag_elpa_base 50 14.0 1.698 1.709 1.720 1.732 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.614 1.698 multiply_cannon_sync_h2d 18288 15.5 1.368 1.614 1.368 1.614 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.556 1.570 ot_diis_step 108 11.5 0.011 0.011 1.482 1.482 fft_wrap_pw1pw2_140 487 13.2 0.253 0.259 1.459 1.471 mp_sum_l 7207 12.9 1.148 1.461 1.148 1.461 acc_transpose_blocks 18288 15.5 0.075 0.078 1.336 1.442 grid_collocate_task_list 119 9.7 1.215 1.361 1.215 1.361 fft3d_ps 1201 14.6 0.534 0.551 1.311 1.324 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.241 1.255 create_qs_kind_set 1 2.0 0.000 0.001 1.211 1.215 read_qs_kind 2 3.0 0.015 0.018 1.211 1.215 parser_read_line 2807 4.0 0.001 0.001 1.196 1.207 parser_read_line_low 5 5.0 0.037 1.175 1.196 1.207 broadcast_input_information 5 6.0 0.002 0.002 1.158 1.206 mp_bcast_i_src 20 7.0 1.152 1.201 1.152 1.201 wfi_extrapolate 11 7.9 0.001 0.001 1.156 1.156 multiply_cannon_metrocomm1 18288 15.5 0.028 0.030 0.404 1.036 acc_transpose_blocks_kernels 18288 16.5 0.210 0.220 0.898 0.992 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 0.947 0.959 make_images_data 4572 15.5 0.044 0.048 0.801 0.947 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.871 0.896 hybrid_alltoall_any 4725 16.4 0.055 0.112 0.687 0.863 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.810 0.812 parallel_gemm_fm 81 9.0 0.000 0.000 0.794 0.798 parallel_gemm_fm_cosma 81 10.0 0.794 0.797 0.794 0.797 mp_alltoall_d11v 2130 13.8 0.654 0.784 0.654 0.784 mp_waitany 9880 13.7 0.505 0.777 0.505 0.777 jit_kernel_transpose 5 15.6 0.688 0.772 0.688 0.772 rs_pw_transfer_RS2PW_140 130 11.5 0.119 0.121 0.498 0.755 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.658 0.744 cp_fm_cholesky_invert 11 10.9 0.738 0.742 0.738 0.742 make_basis_sm 11 9.8 0.000 0.000 0.709 0.711 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=34.708000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=497.636364, yerr=2.012359 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ef71e334ee8f071d2456394b624a331795da3ecc_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 557.076480E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63496. MP_Allreduce 10074 349. MP_Sync 54 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.033 0.063 37.549 37.550 qs_mol_dyn_low 1 2.0 0.003 0.003 37.217 37.224 qs_forces 11 3.9 0.007 0.038 35.850 35.851 qs_energies 11 4.9 0.008 0.035 34.007 34.015 scf_env_do_scf 11 5.9 0.001 0.001 28.433 28.435 scf_env_do_scf_inner_loop 108 6.5 0.003 0.023 24.977 24.978 velocity_verlet 10 3.0 0.002 0.003 19.036 19.092 dbcsr_multiply_generic 2286 12.5 0.098 0.102 18.436 18.529 qs_scf_new_mos 108 7.5 0.001 0.001 16.419 16.479 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.418 16.478 ot_scf_mini 108 9.5 0.002 0.003 15.477 15.526 multiply_cannon 2286 13.5 0.229 0.272 14.864 15.340 multiply_cannon_loop 2286 14.5 0.931 0.962 13.915 14.425 ot_mini 108 10.5 0.001 0.001 9.289 9.351 multiply_cannon_multrec 27432 15.5 2.330 3.004 8.668 9.222 dbcsr_mm_accdrv_process 47916 15.9 5.366 7.501 6.244 7.764 qs_ot_get_derivative 108 11.5 0.001 0.001 7.475 7.525 rebuild_ks_matrix 119 8.3 0.000 0.000 6.713 6.765 qs_ks_build_kohn_sham_matrix 119 9.3 0.017 0.025 6.713 6.764 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.910 5.959 init_scf_run 11 5.9 0.000 0.001 4.157 4.157 scf_env_initial_rho_setup 11 6.9 0.000 0.001 4.157 4.157 sum_up_and_integrate 119 10.3 0.035 0.038 3.934 3.955 integrate_v_rspace 119 11.3 0.003 0.003 3.899 3.921 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.682 3.748 calculate_rho_elec 119 8.7 0.040 0.047 3.682 3.748 qs_ot_get_p 119 10.4 0.001 0.002 3.588 3.659 init_scf_loop 11 6.9 0.005 0.025 3.436 3.438 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.845 3.235 mp_waitall_1 137007 16.6 2.190 2.831 2.190 2.831 calculate_first_density_matrix 1 7.0 0.035 0.139 2.679 2.781 prepare_preconditioner 11 7.9 0.000 0.000 2.554 2.567 make_preconditioner 11 8.9 0.000 0.002 2.554 2.566 calculate_dm_sparse 119 9.5 0.000 0.000 2.432 2.491 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.158 2.487 make_m2s 4572 13.5 0.054 0.056 2.188 2.310 qs_ot_p2m_diag 50 11.0 0.016 0.023 2.273 2.282 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.029 2.277 apply_single 119 13.6 0.000 0.000 2.029 2.276 rs_pw_transfer 974 11.9 0.009 0.009 2.067 2.261 make_images 4572 14.5 0.269 0.329 2.081 2.202 density_rs2pw 119 9.7 0.004 0.004 1.983 2.175 pw_transfer 1439 11.6 0.065 0.069 2.110 2.155 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.018 2.066 qs_ot_get_derivative_diag 49 12.0 0.001 0.003 1.962 1.989 cp_dbcsr_syevd 50 12.0 0.004 0.004 1.952 1.953 grid_integrate_task_list 119 12.3 1.825 1.902 1.825 1.902 jit_kernel_multiply 10 16.0 0.819 1.825 0.819 1.825 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.790 1.792 ot_diis_step 108 11.5 0.012 0.012 1.773 1.774 fft_wrap_pw1pw2_140 487 13.2 0.287 0.299 1.653 1.704 potential_pw2rs 119 12.3 0.009 0.009 1.682 1.699 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.683 1.692 cp_fm_diag_elpa_base 50 14.0 1.647 1.666 1.680 1.690 acc_transpose_blocks 27432 15.5 0.113 0.115 1.469 1.666 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.616 1.630 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.610 1.624 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.952 1.601 fft3d_ps 1201 14.6 0.557 0.610 1.456 1.493 mp_sum_dm 438 4.9 1.344 1.412 1.344 1.412 mp_sum_l 7207 12.9 0.987 1.406 0.987 1.406 md_write_output 11 3.9 0.057 1.342 0.058 1.354 md_output 10 3.0 0.000 0.000 0.061 1.354 update_particle_set 20 4.0 0.000 0.000 1.293 1.349 grid_collocate_task_list 119 9.7 1.222 1.340 1.222 1.340 wfi_extrapolate 11 7.9 0.001 0.001 1.321 1.321 cp_fm_upper_to_full 72 14.2 0.819 1.174 0.819 1.174 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.100 1.117 qs_energies_init_hamiltonians 11 5.9 0.002 0.004 1.099 1.100 multiply_cannon_sync_h2d 27432 15.5 0.998 1.063 0.998 1.063 dbcsr_complete_redistribute 329 12.2 0.119 0.145 0.779 1.058 acc_transpose_blocks_kernels 27432 16.5 0.267 0.273 0.847 1.037 mp_alltoall_d11v 2130 13.8 0.859 0.993 0.859 0.993 make_images_data 4572 15.5 0.045 0.049 0.852 0.984 multiply_cannon_metrocomm1 27432 15.5 0.034 0.035 0.381 0.973 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.912 0.916 hybrid_alltoall_any 4725 16.4 0.061 0.151 0.746 0.904 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.796 0.873 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.575 0.842 mp_alltoall_z22v 1201 16.6 0.783 0.819 0.783 0.819 jit_kernel_transpose 5 15.6 0.580 0.774 0.580 0.774 cp_fm_cholesky_invert 11 10.9 0.752 0.755 0.752 0.755 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=37.550000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=527.818182, yerr=4.858323 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ef71e334ee8f071d2456394b624a331795da3ecc_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 597.671936E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63495. MP_Allreduce 10074 348. MP_Sync 54 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.033 27.426 27.428 qs_mol_dyn_low 1 2.0 0.003 0.003 27.259 27.265 qs_forces 11 3.9 0.002 0.003 27.189 27.190 qs_energies 11 4.9 0.001 0.002 25.475 25.477 scf_env_do_scf 11 5.9 0.000 0.001 20.763 20.763 scf_env_do_scf_inner_loop 108 6.5 0.011 0.025 18.323 18.324 velocity_verlet 10 3.0 0.016 0.017 14.202 14.205 dbcsr_multiply_generic 2286 12.5 0.091 0.096 11.762 11.833 qs_scf_new_mos 108 7.5 0.001 0.001 10.440 10.475 qs_scf_loop_do_ot 108 8.5 0.001 0.001 10.439 10.475 ot_scf_mini 108 9.5 0.002 0.002 9.778 9.815 multiply_cannon 2286 13.5 0.227 0.232 9.362 9.796 multiply_cannon_loop 2286 14.5 0.327 0.338 8.475 8.658 multiply_cannon_multrec 9144 15.5 1.572 1.814 5.718 5.968 rebuild_ks_matrix 119 8.3 0.000 0.000 5.931 5.955 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 5.931 5.954 ot_mini 108 10.5 0.001 0.001 5.454 5.497 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.280 5.300 qs_ot_get_derivative 108 11.5 0.001 0.001 4.183 4.222 dbcsr_mm_accdrv_process 12550 15.8 3.040 3.763 4.045 4.127 sum_up_and_integrate 119 10.3 0.037 0.041 3.638 3.644 integrate_v_rspace 119 11.3 0.003 0.003 3.600 3.607 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.455 3.463 calculate_rho_elec 119 8.7 0.060 0.061 3.455 3.462 init_scf_run 11 5.9 0.000 0.001 3.188 3.188 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.188 3.188 qs_ot_get_p 119 10.4 0.001 0.001 2.583 2.625 init_scf_loop 11 6.9 0.000 0.000 2.421 2.422 calculate_first_density_matrix 1 7.0 0.001 0.001 2.065 2.067 pw_transfer 1439 11.6 0.066 0.069 2.011 2.021 grid_integrate_task_list 119 12.3 1.852 1.959 1.852 1.959 density_rs2pw 119 9.7 0.004 0.004 1.801 1.940 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.918 1.929 make_m2s 4572 13.5 0.034 0.035 1.652 1.819 mp_waitall_1 115863 16.7 1.295 1.805 1.295 1.805 rs_pw_transfer 974 11.9 0.008 0.008 1.641 1.769 make_images 4572 14.5 0.266 0.299 1.563 1.728 calculate_dm_sparse 119 9.5 0.000 0.000 1.699 1.717 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.697 1.700 prepare_preconditioner 11 7.9 0.000 0.000 1.683 1.688 make_preconditioner 11 8.9 0.000 0.000 1.683 1.688 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.576 1.602 fft_wrap_pw1pw2_140 487 13.2 0.364 0.373 1.584 1.597 jit_kernel_multiply 9 15.8 0.968 1.559 0.968 1.559 cp_dbcsr_syevd 50 12.0 0.004 0.004 1.507 1.507 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.473 1.501 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.495 1.497 grid_collocate_task_list 119 9.7 1.270 1.386 1.270 1.386 potential_pw2rs 119 12.3 0.010 0.011 1.373 1.375 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.345 1.353 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.262 1.270 cp_fm_diag_elpa_base 50 14.0 1.235 1.251 1.260 1.268 qs_energies_init_hamiltonians 11 5.9 0.002 0.002 1.261 1.262 fft3d_ps 1201 14.6 0.558 0.567 1.251 1.258 ot_diis_step 108 11.5 0.013 0.013 1.257 1.258 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.138 1.162 apply_single 119 13.6 0.000 0.000 1.138 1.162 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.129 1.135 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.095 1.111 wfi_extrapolate 11 7.9 0.001 0.001 1.072 1.072 hybrid_alltoall_any 4725 16.4 0.062 0.174 0.720 0.975 make_images_data 4572 15.5 0.039 0.042 0.769 0.969 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.874 0.925 acc_transpose_blocks 9144 15.5 0.038 0.039 0.884 0.890 cp_fm_cholesky_invert 11 10.9 0.842 0.845 0.842 0.845 mp_alltoall_d11v 2130 13.8 0.694 0.816 0.694 0.816 multiply_cannon_sync_h2d 9144 15.5 0.710 0.800 0.710 0.800 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.772 0.774 qs_env_update_s_mstruct 11 6.9 0.042 0.050 0.706 0.762 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.700 0.709 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.322 0.666 acc_transpose_blocks_kernels 9144 16.5 0.117 0.121 0.645 0.647 mp_allgather_i34 2286 14.5 0.221 0.613 0.221 0.613 mp_alltoall_z22v 1201 16.6 0.568 0.597 0.568 0.597 mp_waitany 5200 13.7 0.451 0.576 0.451 0.576 qs_create_task_list 11 7.9 0.000 0.001 0.541 0.568 generate_qs_task_list 11 8.9 0.188 0.210 0.540 0.567 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=27.428000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=567.454545, yerr=2.709030 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ef71e334ee8f071d2456394b624a331795da3ecc_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 753.950720E+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_SendRecv 2499 189067. MP_ISendRecv 2499 189067. MP_Wait 6399 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.074 0.103 43.246 43.248 qs_mol_dyn_low 1 2.0 0.003 0.003 42.826 42.831 qs_forces 11 3.9 0.005 0.008 42.350 42.356 qs_energies 11 4.9 0.001 0.001 40.361 40.366 scf_env_do_scf 11 5.9 0.001 0.001 33.805 33.805 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 25.914 25.916 velocity_verlet 10 3.0 0.002 0.002 23.369 23.389 dbcsr_multiply_generic 2286 12.5 0.100 0.103 17.970 18.113 qs_scf_new_mos 108 7.5 0.001 0.001 16.237 16.340 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.236 16.340 ot_scf_mini 108 9.5 0.002 0.002 15.155 15.256 multiply_cannon 2286 13.5 0.302 0.308 14.026 15.112 multiply_cannon_loop 2286 14.5 0.342 0.348 12.777 13.852 ot_mini 108 10.5 0.001 0.001 9.303 9.420 multiply_cannon_multrec 9144 15.5 3.546 4.996 8.760 9.041 init_scf_loop 11 6.9 0.000 0.000 7.864 7.865 rebuild_ks_matrix 119 8.3 0.000 0.000 7.266 7.407 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.266 7.407 qs_ot_get_derivative 108 11.5 0.001 0.001 7.274 7.378 prepare_preconditioner 11 7.9 0.000 0.000 6.867 6.879 make_preconditioner 11 8.9 0.000 0.000 6.867 6.879 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.405 6.759 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.568 6.696 dbcsr_mm_accdrv_process 12550 15.8 4.111 6.054 5.088 6.649 cp_fm_upper_to_full 72 14.2 3.206 4.614 3.206 4.614 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.213 4.223 calculate_rho_elec 119 8.7 0.118 0.121 4.213 4.222 init_scf_run 11 5.9 0.000 0.001 4.136 4.136 scf_env_initial_rho_setup 11 6.9 0.000 0.000 4.136 4.136 sum_up_and_integrate 119 10.3 0.065 0.067 3.973 3.979 integrate_v_rspace 119 11.3 0.004 0.004 3.908 3.915 mp_waitall_1 94719 16.7 2.518 3.752 2.518 3.752 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.785 3.345 qs_ot_get_p 119 10.4 0.001 0.001 3.135 3.276 dbcsr_complete_redistribute 329 12.2 0.288 0.293 1.983 2.814 calculate_first_density_matrix 1 7.0 0.002 0.007 2.695 2.702 multiply_cannon_metrocomm3 9144 15.5 0.019 0.020 1.462 2.662 pw_transfer 1439 11.6 0.069 0.069 2.644 2.648 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.336 2.632 apply_single 119 13.6 0.000 0.000 2.336 2.632 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.545 2.550 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.684 2.510 make_m2s 4572 13.5 0.038 0.039 2.259 2.427 mp_alltoall_i22 627 13.8 1.516 2.345 1.516 2.345 make_images 4572 14.5 0.350 0.381 2.137 2.304 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.456 2.278 calculate_dm_sparse 119 9.5 0.000 0.000 2.167 2.181 density_rs2pw 119 9.7 0.004 0.004 2.140 2.155 fft_wrap_pw1pw2_140 487 13.2 0.617 0.623 2.144 2.152 grid_integrate_task_list 119 12.3 2.052 2.064 2.052 2.064 ot_diis_step 108 11.5 0.014 0.015 2.000 2.001 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.999 2.000 mp_sum_l 7207 12.9 1.151 1.898 1.151 1.898 qs_ot_p2m_diag 50 11.0 0.043 0.044 1.896 1.897 qs_energies_init_hamiltonians 11 5.9 0.002 0.002 1.789 1.798 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.673 1.714 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.653 1.706 cp_dbcsr_syevd 50 12.0 0.004 0.004 1.624 1.624 fft3d_ps 1201 14.6 0.598 0.612 1.562 1.567 grid_collocate_task_list 119 9.7 1.486 1.497 1.486 1.497 jit_kernel_multiply 8 15.4 0.950 1.461 0.950 1.461 rs_pw_transfer 974 11.9 0.009 0.009 1.429 1.453 cp_fm_cholesky_invert 11 10.9 1.440 1.443 1.440 1.443 hybrid_alltoall_any 4725 16.4 0.087 0.146 1.163 1.405 potential_pw2rs 119 12.3 0.014 0.015 1.370 1.372 wfi_extrapolate 11 7.9 0.001 0.001 1.371 1.371 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.343 1.343 cp_fm_diag_elpa_base 50 14.0 1.197 1.250 1.342 1.342 make_images_data 4572 15.5 0.043 0.046 1.100 1.310 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.233 1.238 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.129 1.148 qs_env_update_s_mstruct 11 6.9 0.004 0.011 1.118 1.145 acc_transpose_blocks 9144 15.5 0.038 0.039 1.015 1.128 mp_alltoall_d11v 2130 13.8 1.065 1.086 1.065 1.086 multiply_cannon_sync_h2d 9144 15.5 1.044 1.048 1.044 1.048 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.993 1.028 qs_create_task_list 11 7.9 0.011 0.020 0.953 0.966 generate_qs_task_list 11 8.9 0.367 0.387 0.942 0.964 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.909 0.922 acc_transpose_blocks_kernels 9144 16.5 0.118 0.118 0.766 0.878 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=43.248000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=710.000000, yerr=11.135529 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ef71e334ee8f071d2456394b624a331795da3ecc_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 505.266176E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66218. MP_Allreduce 9696 492. MP_Sync 52 MP_Alltoall 1938 513086. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.036 0.058 82.882 82.883 qs_mol_dyn_low 1 2.0 0.003 0.003 81.477 81.487 qs_forces 11 3.9 0.007 0.008 81.411 81.412 qs_energies 11 4.9 0.001 0.002 78.540 78.556 scf_env_do_scf 11 5.9 0.000 0.001 68.785 68.788 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 63.363 63.363 dbcsr_multiply_generic 2055 12.4 0.112 0.115 50.323 50.500 qs_scf_new_mos 99 7.5 0.000 0.001 46.153 46.321 qs_scf_loop_do_ot 99 8.5 0.001 0.001 46.153 46.320 ot_scf_mini 99 9.5 0.002 0.002 43.851 43.990 multiply_cannon 2055 13.4 0.189 0.195 42.335 43.022 multiply_cannon_loop 2055 14.4 1.495 1.537 41.411 42.109 velocity_verlet 10 3.0 0.001 0.002 41.965 41.965 ot_mini 99 10.5 0.001 0.001 26.146 26.298 qs_ot_get_derivative 99 11.5 0.001 0.001 19.418 19.557 multiply_cannon_multrec 49320 15.4 12.575 13.210 17.549 18.422 rebuild_ks_matrix 110 8.3 0.000 0.000 14.532 14.660 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.532 14.660 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.741 12.860 mp_waitall_1 220248 16.4 10.496 11.979 10.496 11.979 multiply_cannon_sync_h2d 49320 15.4 10.432 10.964 10.432 10.964 qs_ot_get_p 110 10.4 0.001 0.001 9.378 9.549 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.680 8.230 multiply_cannon_metrocomm3 49320 15.4 0.080 0.083 6.478 7.951 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.225 7.712 apply_single 110 13.6 0.000 0.000 7.225 7.712 init_scf_run 11 5.9 0.000 0.001 7.349 7.349 scf_env_initial_rho_setup 11 6.9 0.000 0.001 7.349 7.349 sum_up_and_integrate 110 10.3 0.036 0.043 7.061 7.076 integrate_v_rspace 110 11.3 0.003 0.004 7.025 7.048 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.473 6.623 calculate_rho_elec 110 8.6 0.021 0.026 6.472 6.623 ot_diis_step 99 11.5 0.005 0.006 6.538 6.538 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.402 6.422 cp_dbcsr_syevd 48 12.0 0.003 0.003 5.567 5.567 init_scf_loop 11 6.9 0.000 0.000 5.397 5.397 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.249 5.333 dbcsr_mm_accdrv_process 87628 16.1 1.902 1.981 4.848 5.258 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.014 5.037 cp_fm_diag_elpa_base 48 14.0 5.001 5.024 5.012 5.035 mp_sum_l 6514 12.8 3.695 4.533 3.695 4.533 rs_pw_transfer 902 11.9 0.012 0.013 3.642 4.065 wfi_extrapolate 11 7.9 0.001 0.001 3.995 3.995 make_m2s 4110 13.4 0.060 0.067 3.828 3.934 make_images 4110 14.4 0.176 0.189 3.733 3.844 density_rs2pw 110 9.6 0.004 0.005 3.316 3.776 calculate_dm_sparse 110 9.5 0.001 0.001 3.683 3.774 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.606 3.610 grid_integrate_task_list 110 12.3 3.234 3.397 3.234 3.397 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.278 3.330 prepare_preconditioner 11 7.9 0.000 0.000 3.255 3.276 make_preconditioner 11 8.9 0.000 0.000 3.255 3.276 calculate_first_density_matrix 1 7.0 0.005 0.013 3.261 3.271 multiply_cannon_metrocomm1 49320 15.4 0.062 0.063 2.214 3.259 pw_transfer 1331 11.6 0.055 0.062 3.081 3.157 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.038 3.093 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.040 3.081 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 2.993 3.072 jit_kernel_multiply 13 15.9 2.672 2.860 2.672 2.860 fft_wrap_pw1pw2_140 451 13.1 0.452 0.502 2.562 2.640 potential_pw2rs 110 12.3 0.006 0.007 2.579 2.596 mp_alltoall_d11v 2046 13.8 2.039 2.475 2.039 2.475 acc_transpose_blocks 49320 15.4 0.211 0.218 2.147 2.356 mp_waitany 14300 13.8 1.838 2.290 1.838 2.290 grid_collocate_task_list 110 9.6 2.087 2.285 2.087 2.285 fft3d_ps 1111 14.6 0.793 0.884 2.188 2.236 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.921 1.935 make_images_data 4110 15.4 0.042 0.045 1.767 1.897 mp_sum_d 3889 11.9 1.279 1.873 1.279 1.873 hybrid_alltoall_any 4261 16.3 0.082 0.482 1.541 1.831 parallel_gemm_fm 81 9.0 0.000 0.000 1.795 1.799 parallel_gemm_fm_cosma 81 10.0 1.794 1.799 1.794 1.799 cp_fm_cholesky_invert 11 10.9 1.754 1.758 1.754 1.758 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.657 1.693 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=82.883000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=478.272727, yerr=3.670296 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ef71e334ee8f071d2456394b624a331795da3ecc_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 596.058112E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1972800 MPI messages size (bytes): total size 1.077520E+12 min size 0.000000E+00 max size 4.537280E+06 average size 546.188250E+03 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 222984 1826684928 8192 < size <= 32768 520356 13399818240 32768 < size <= 131072 372336 35386294272 131072 < size <= 4194304 787758 788321309808 4194304 < size <= 16777216 54450 238588003280 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66437. MP_Allreduce 9695 570. MP_Sync 52 MP_Alltoall 1717 2549971. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.120 0.218 71.174 71.175 qs_mol_dyn_low 1 2.0 0.003 0.004 70.143 70.170 qs_forces 11 3.9 0.004 0.008 69.859 69.860 qs_energies 11 4.9 0.013 0.062 66.510 66.514 scf_env_do_scf 11 5.9 0.001 0.002 57.227 57.231 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 49.638 49.639 dbcsr_multiply_generic 2055 12.4 0.114 0.118 38.032 38.277 velocity_verlet 10 3.0 0.002 0.008 36.210 36.215 qs_scf_new_mos 99 7.5 0.001 0.001 33.374 33.511 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.374 33.510 multiply_cannon 2055 13.4 0.222 0.245 31.540 32.670 ot_scf_mini 99 9.5 0.003 0.004 31.726 31.868 multiply_cannon_loop 2055 14.4 0.917 0.941 30.238 31.017 ot_mini 99 10.5 0.001 0.001 18.421 18.563 multiply_cannon_multrec 24660 15.4 7.689 9.902 14.226 16.313 rebuild_ks_matrix 110 8.3 0.000 0.000 13.794 13.941 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.021 13.794 13.940 qs_ot_get_derivative 99 11.5 0.001 0.001 12.633 12.776 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.143 12.273 mp_waitall_1 176588 16.5 7.687 10.503 7.687 10.503 multiply_cannon_sync_h2d 24660 15.4 7.087 8.228 7.087 8.228 multiply_cannon_metrocomm3 24660 15.4 0.068 0.069 5.270 7.960 init_scf_loop 11 6.9 0.001 0.005 7.556 7.557 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.487 7.179 apply_single 110 13.6 0.000 0.001 6.487 7.178 qs_ot_get_p 110 10.4 0.001 0.001 6.693 6.870 dbcsr_mm_accdrv_process 52282 16.1 4.891 6.112 6.375 6.805 init_scf_run 11 5.9 0.000 0.001 6.636 6.637 scf_env_initial_rho_setup 11 6.9 0.000 0.003 6.636 6.637 sum_up_and_integrate 110 10.3 0.053 0.059 6.617 6.631 integrate_v_rspace 110 11.3 0.003 0.003 6.564 6.580 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.132 6.142 calculate_rho_elec 110 8.6 0.040 0.048 6.131 6.141 ot_diis_step 99 11.5 0.010 0.010 5.742 5.742 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.813 5.520 prepare_preconditioner 11 7.9 0.000 0.000 5.501 5.520 make_preconditioner 11 8.9 0.000 0.001 5.501 5.520 make_full_inverse_cholesky 11 9.9 0.000 0.001 5.060 5.226 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.834 4.854 make_m2s 4110 13.4 0.057 0.059 4.116 4.609 make_images 4110 14.4 0.398 0.442 4.008 4.499 cp_dbcsr_syevd 48 12.0 0.012 0.072 4.411 4.411 cp_fm_diag_elpa 48 13.0 0.012 0.093 3.838 3.859 density_rs2pw 110 9.6 0.005 0.006 3.251 3.805 pw_transfer 1331 11.6 0.067 0.076 3.661 3.800 cp_fm_diag_elpa_base 48 14.0 3.645 3.660 3.711 3.728 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.554 3.696 rs_pw_transfer 902 11.9 0.012 0.015 2.949 3.520 wfi_extrapolate 11 7.9 0.001 0.001 3.508 3.508 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.409 3.411 grid_integrate_task_list 110 12.3 3.144 3.380 3.144 3.380 qs_ot_get_derivative_diag 47 12.0 0.003 0.014 3.142 3.215 fft_wrap_pw1pw2_140 451 13.1 0.525 0.570 3.053 3.191 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.056 3.119 calculate_dm_sparse 110 9.5 0.001 0.001 3.007 3.038 calculate_first_density_matrix 1 7.0 0.000 0.003 3.033 3.036 jit_kernel_multiply 12 16.2 1.135 2.867 1.135 2.867 make_images_data 4110 15.4 0.046 0.050 2.307 2.795 hybrid_alltoall_any 4261 16.3 0.102 0.443 2.024 2.766 fft3d_ps 1111 14.6 1.113 1.334 2.523 2.687 cp_fm_cholesky_invert 11 10.9 2.543 2.549 2.543 2.549 grid_collocate_task_list 110 9.6 2.057 2.464 2.057 2.464 mp_sum_l 6514 12.8 1.827 2.450 1.827 2.450 potential_pw2rs 110 12.3 0.008 0.009 2.407 2.418 mp_alltoall_d11v 2046 13.8 1.761 2.118 1.761 2.118 qs_energies_init_hamiltonians 11 5.9 0.020 0.029 2.009 2.013 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.930 1.958 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.808 1.823 mp_waitany 10164 13.8 1.229 1.754 1.229 1.754 mp_allgather_i34 2055 14.4 0.616 1.609 0.616 1.609 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.590 1.605 rs_pw_transfer_RS2PW_140 121 11.5 0.209 0.222 0.973 1.537 acc_transpose_blocks 24660 15.4 0.110 0.112 1.517 1.534 multiply_cannon_metrocomm4 22605 15.4 0.077 0.080 0.774 1.497 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.379 1.492 dbcsr_complete_redistribute 325 12.2 0.244 0.321 1.159 1.424 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=71.175000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=556.363636, yerr=8.488203 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ef71e334ee8f071d2456394b624a331795da3ecc_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 660.795392E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66428. MP_Allreduce 9694 567. MP_Sync 52 MP_Alltoall 1496 4511006. MP_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.057 0.133 62.487 62.489 qs_mol_dyn_low 1 2.0 0.003 0.003 61.876 61.900 qs_forces 11 3.9 0.003 0.003 61.637 61.638 qs_energies 11 4.9 0.009 0.062 58.449 58.452 scf_env_do_scf 11 5.9 0.001 0.001 48.710 48.711 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 40.214 40.214 velocity_verlet 10 3.0 0.002 0.002 32.647 32.654 dbcsr_multiply_generic 2055 12.4 0.107 0.112 29.426 29.694 qs_scf_new_mos 99 7.5 0.001 0.001 25.102 25.195 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.101 25.194 multiply_cannon 2055 13.4 0.214 0.227 23.288 24.481 ot_scf_mini 99 9.5 0.002 0.003 23.887 24.000 multiply_cannon_loop 2055 14.4 0.611 0.624 22.107 23.037 ot_mini 99 10.5 0.001 0.001 13.712 13.829 rebuild_ks_matrix 110 8.3 0.000 0.000 12.400 12.533 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.399 12.533 multiply_cannon_multrec 16440 15.4 3.972 5.123 10.227 11.471 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.917 11.040 mp_waitall_1 139946 16.5 7.531 10.259 7.531 10.259 qs_ot_get_derivative 99 11.5 0.001 0.001 9.311 9.427 init_scf_loop 11 6.9 0.001 0.004 8.463 8.463 multiply_cannon_metrocomm3 16440 15.4 0.042 0.044 4.696 7.546 dbcsr_mm_accdrv_process 34862 16.1 5.134 6.183 6.108 6.960 init_scf_run 11 5.9 0.000 0.001 6.719 6.719 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.719 6.719 prepare_preconditioner 11 7.9 0.000 0.000 6.699 6.715 make_preconditioner 11 8.9 0.000 0.001 6.699 6.715 sum_up_and_integrate 110 10.3 0.061 0.061 6.564 6.581 integrate_v_rspace 110 11.3 0.003 0.003 6.503 6.521 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.022 6.389 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.298 6.345 calculate_rho_elec 110 8.6 0.059 0.060 6.297 6.345 qs_ot_get_p 110 10.4 0.001 0.001 5.405 5.545 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.914 5.331 apply_single 110 13.6 0.000 0.000 4.913 5.331 make_m2s 4110 13.4 0.050 0.051 4.103 4.479 ot_diis_step 99 11.5 0.010 0.011 4.373 4.374 make_images 4110 14.4 0.390 0.508 3.989 4.365 multiply_cannon_sync_h2d 16440 15.4 3.727 4.332 3.727 4.332 density_rs2pw 110 9.6 0.004 0.005 3.115 4.320 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.464 4.071 rs_pw_transfer 902 11.9 0.011 0.011 2.728 3.854 qs_ot_p2m_diag 48 11.0 0.056 0.159 3.818 3.822 pw_transfer 1331 11.6 0.065 0.073 3.667 3.677 calculate_first_density_matrix 1 7.0 0.040 0.316 3.396 3.669 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.560 3.573 cp_dbcsr_syevd 48 12.0 0.010 0.030 3.470 3.485 calculate_dm_sparse 110 9.5 0.001 0.001 3.383 3.414 grid_integrate_task_list 110 12.3 3.188 3.393 3.188 3.393 fft_wrap_pw1pw2_140 451 13.1 0.638 0.647 3.073 3.086 wfi_extrapolate 11 7.9 0.001 0.001 2.964 2.964 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.826 2.841 make_images_data 4110 15.4 0.043 0.048 2.361 2.832 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.824 2.825 cp_fm_diag_elpa_base 48 14.0 2.740 2.772 2.804 2.813 hybrid_alltoall_any 4261 16.3 0.105 0.374 2.109 2.763 cp_fm_cholesky_invert 11 10.9 2.618 2.624 2.618 2.624 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.492 2.537 qs_energies_init_hamiltonians 11 5.9 0.105 0.174 2.494 2.496 grid_collocate_task_list 110 9.6 2.080 2.474 2.080 2.474 mp_waitany 17072 13.8 1.227 2.421 1.227 2.421 fft3d_ps 1111 14.6 1.096 1.107 2.382 2.397 multiply_cannon_metrocomm4 14385 15.4 0.046 0.050 0.861 2.382 mp_alltoall_d11v 2046 13.8 2.044 2.375 2.044 2.375 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 2.351 2.363 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.281 2.345 potential_pw2rs 110 12.3 0.011 0.011 2.282 2.288 mp_irecv_dv 48980 15.7 0.791 2.255 0.791 2.255 jit_kernel_multiply 10 16.6 0.586 2.244 0.586 2.244 rs_pw_transfer_RS2PW_140 121 11.5 0.179 0.183 0.953 2.076 mp_sum_l 6514 12.8 1.435 2.019 1.435 2.019 dbcsr_complete_redistribute 325 12.2 0.329 0.369 1.454 1.944 cp_fm_upper_to_full 70 14.2 1.409 1.785 1.409 1.785 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.676 1.690 multiply_cannon_metrocomm1 16440 15.4 0.029 0.029 0.564 1.588 cp_fm_cholesky_decompose 22 10.9 1.539 1.557 1.539 1.557 qs_env_update_s_mstruct 11 6.9 0.000 0.002 1.418 1.550 mp_allgather_i34 2055 14.4 0.506 1.524 0.506 1.524 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.370 1.451 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.984 1.450 acc_transpose_blocks 16440 15.4 0.073 0.075 1.244 1.291 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=62.489000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=626.181818, yerr=8.144359 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ef71e334ee8f071d2456394b624a331795da3ecc_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 727.113728E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66426. MP_Allreduce 9694 608. MP_Sync 52 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.037 0.090 68.666 68.668 qs_mol_dyn_low 1 2.0 0.003 0.003 67.590 67.673 qs_forces 11 3.9 0.003 0.003 67.289 67.290 qs_energies 11 4.9 0.002 0.008 63.833 63.837 scf_env_do_scf 11 5.9 0.001 0.001 54.544 54.547 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 42.497 42.498 velocity_verlet 10 3.0 0.003 0.012 37.719 37.722 dbcsr_multiply_generic 2055 12.4 0.113 0.118 30.346 30.581 qs_scf_new_mos 99 7.5 0.001 0.001 27.202 27.317 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.201 27.317 ot_scf_mini 99 9.5 0.003 0.003 25.573 25.696 multiply_cannon 2055 13.4 0.247 0.296 23.330 24.481 multiply_cannon_loop 2055 14.4 0.879 0.901 21.771 22.455 multiply_cannon_multrec 24660 15.4 4.190 6.721 13.151 14.471 ot_mini 99 10.5 0.001 0.001 14.283 14.431 rebuild_ks_matrix 110 8.3 0.000 0.000 12.394 12.500 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 12.393 12.499 init_scf_loop 11 6.9 0.001 0.004 12.005 12.006 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.942 11.036 qs_ot_get_derivative 99 11.5 0.001 0.001 10.106 10.235 prepare_preconditioner 11 7.9 0.000 0.000 10.195 10.209 make_preconditioner 11 8.9 0.000 0.002 10.195 10.209 dbcsr_mm_accdrv_process 52304 16.0 7.154 8.434 8.814 9.930 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.402 9.881 mp_waitall_1 121746 16.5 4.427 6.758 4.427 6.758 sum_up_and_integrate 110 10.3 0.067 0.071 6.667 6.680 integrate_v_rspace 110 11.3 0.003 0.003 6.599 6.612 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.500 6.549 calculate_rho_elec 110 8.6 0.078 0.081 6.500 6.549 init_scf_run 11 5.9 0.000 0.001 6.512 6.512 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.511 6.512 qs_ot_get_p 110 10.4 0.001 0.001 6.271 6.425 make_m2s 4110 13.4 0.060 0.062 5.363 5.783 make_images 4110 14.4 0.571 0.698 5.222 5.636 cp_fm_upper_to_full 70 14.2 3.380 4.911 3.380 4.911 qs_ot_p2m_diag 48 11.0 0.055 0.065 4.380 4.395 ot_diis_step 99 11.5 0.011 0.011 4.141 4.141 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.029 4.131 apply_single 110 13.6 0.000 0.000 4.029 4.131 cp_dbcsr_syevd 48 12.0 0.003 0.004 3.898 3.899 dbcsr_complete_redistribute 325 12.2 0.412 0.456 2.728 3.874 pw_transfer 1331 11.6 0.066 0.075 3.841 3.873 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.735 3.771 density_rs2pw 110 9.6 0.004 0.005 3.131 3.646 calculate_dm_sparse 110 9.5 0.001 0.001 3.546 3.575 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.451 3.506 grid_integrate_task_list 110 12.3 3.257 3.433 3.257 3.433 multiply_cannon_sync_h2d 24660 15.4 3.184 3.405 3.184 3.405 calculate_first_density_matrix 1 7.0 0.039 0.310 3.124 3.392 multiply_cannon_metrocomm3 24660 15.4 0.036 0.037 1.474 3.367 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.151 3.288 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.236 3.244 cp_fm_diag_elpa_base 48 14.0 3.082 3.142 3.233 3.242 fft_wrap_pw1pw2_140 451 13.1 0.665 0.689 3.161 3.200 qs_ot_get_derivative_diag 47 12.0 0.001 0.003 3.079 3.148 hybrid_alltoall_any 4261 16.3 0.120 0.458 2.359 3.125 make_images_data 4110 15.4 0.046 0.050 2.652 3.080 rs_pw_transfer 902 11.9 0.010 0.011 2.533 3.067 wfi_extrapolate 11 7.9 0.001 0.001 3.025 3.025 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.783 2.905 cp_fm_cholesky_invert 11 10.9 2.865 2.874 2.865 2.874 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.859 2.861 mp_alltoall_i22 605 13.7 1.679 2.855 1.679 2.855 mp_alltoall_d11v 2046 13.8 2.165 2.705 2.165 2.705 jit_kernel_multiply 10 15.7 1.331 2.552 1.331 2.552 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.500 2.543 fft3d_ps 1111 14.6 1.085 1.117 2.500 2.515 grid_collocate_task_list 110 9.6 2.183 2.422 2.183 2.422 potential_pw2rs 110 12.3 0.013 0.013 2.288 2.303 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.267 2.268 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 2.162 2.178 cp_fm_cholesky_decompose 22 10.9 1.737 1.782 1.737 1.782 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.719 1.751 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.608 1.715 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.696 1.708 mp_sum_l 6514 12.8 1.018 1.655 1.018 1.655 acc_transpose_blocks 24660 15.4 0.107 0.110 1.610 1.647 mp_allgather_i34 2055 14.4 0.661 1.639 0.661 1.639 mp_waitany 13376 13.8 1.088 1.612 1.088 1.612 multiply_cannon_metrocomm4 20550 15.4 0.062 0.066 0.865 1.582 mp_irecv_dv 62702 16.1 0.762 1.415 0.762 1.415 rs_gather_matrices 110 12.3 0.267 0.295 0.968 1.387 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=68.668000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=689.000000, yerr=7.965836 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ef71e334ee8f071d2456394b624a331795da3ecc_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 835.534848E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66424. MP_Allreduce 9694 649. MP_Sync 52 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.051 0.075 57.096 57.109 qs_mol_dyn_low 1 2.0 0.003 0.003 56.645 56.650 qs_forces 11 3.9 0.003 0.003 56.557 56.558 qs_energies 11 4.9 0.005 0.007 52.856 52.863 scf_env_do_scf 11 5.9 0.000 0.001 43.519 43.519 scf_env_do_scf_inner_loop 99 6.5 0.022 0.034 35.614 35.614 velocity_verlet 10 3.0 0.002 0.002 31.543 31.546 dbcsr_multiply_generic 2055 12.4 0.104 0.108 23.441 23.592 qs_scf_new_mos 99 7.5 0.001 0.001 20.436 20.502 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.436 20.502 ot_scf_mini 99 9.5 0.002 0.002 19.193 19.247 multiply_cannon 2055 13.4 0.252 0.270 17.954 19.174 multiply_cannon_loop 2055 14.4 0.317 0.329 16.600 17.012 rebuild_ks_matrix 110 8.3 0.000 0.000 11.949 11.997 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 11.949 11.997 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.612 10.655 ot_mini 99 10.5 0.001 0.001 10.532 10.580 multiply_cannon_multrec 8220 15.4 3.257 5.109 8.026 9.357 init_scf_loop 11 6.9 0.000 0.000 7.851 7.852 mp_waitall_1 103326 16.6 6.080 7.697 6.080 7.697 qs_ot_get_derivative 99 11.5 0.001 0.001 6.684 6.738 sum_up_and_integrate 110 10.3 0.080 0.081 6.561 6.571 integrate_v_rspace 110 11.3 0.003 0.003 6.481 6.493 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.318 6.338 calculate_rho_elec 110 8.6 0.114 0.115 6.318 6.337 prepare_preconditioner 11 7.9 0.000 0.000 6.170 6.175 make_preconditioner 11 8.9 0.000 0.000 6.170 6.175 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.744 5.817 dbcsr_mm_accdrv_process 17442 15.9 3.161 4.475 4.638 5.677 init_scf_run 11 5.9 0.000 0.001 5.669 5.669 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.668 5.669 qs_ot_get_p 110 10.4 0.001 0.001 4.772 4.828 multiply_cannon_metrocomm3 8220 15.4 0.017 0.018 3.086 4.433 make_m2s 4110 13.4 0.039 0.039 4.181 4.426 make_images 4110 14.4 0.633 0.686 4.051 4.295 pw_transfer 1331 11.6 0.066 0.072 3.977 3.993 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.870 3.889 ot_diis_step 99 11.5 0.012 0.012 3.826 3.826 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.725 3.789 apply_single 110 13.6 0.000 0.000 3.725 3.789 grid_integrate_task_list 110 12.3 3.364 3.532 3.364 3.532 density_rs2pw 110 9.6 0.004 0.005 3.062 3.439 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.399 3.403 fft_wrap_pw1pw2_140 451 13.1 0.834 0.846 3.340 3.360 qs_energies_init_hamiltonians 11 5.9 0.016 0.063 3.130 3.132 multiply_cannon_sync_h2d 8220 15.4 2.894 3.121 2.894 3.121 cp_dbcsr_syevd 48 12.0 0.004 0.004 3.092 3.092 cp_fm_cholesky_invert 11 10.9 3.016 3.020 3.016 3.020 calculate_dm_sparse 110 9.5 0.001 0.001 2.863 2.904 calculate_first_density_matrix 1 7.0 0.003 0.004 2.804 2.806 make_images_data 4110 15.4 0.038 0.043 2.374 2.775 wfi_extrapolate 11 7.9 0.001 0.001 2.765 2.765 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.687 2.688 hybrid_alltoall_any 4261 16.3 0.199 0.862 2.306 2.649 rs_pw_transfer 902 11.9 0.010 0.011 2.206 2.598 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.578 2.586 cp_fm_diag_elpa_base 48 14.0 2.523 2.548 2.576 2.584 grid_collocate_task_list 110 9.6 2.276 2.536 2.276 2.536 fft3d_ps 1111 14.6 1.137 1.179 2.415 2.445 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.333 2.353 jit_kernel_multiply 10 15.8 1.169 2.223 1.169 2.223 potential_pw2rs 110 12.3 0.015 0.016 2.093 2.098 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.023 2.049 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.896 2.024 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.786 1.999 mp_alltoall_d11v 2046 13.8 1.672 1.910 1.672 1.910 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.816 1.843 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.821 1.832 qs_create_task_list 11 7.9 0.381 0.385 1.609 1.715 cp_fm_cholesky_decompose 22 10.9 1.686 1.707 1.686 1.707 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.668 1.673 mp_allgather_i34 2055 14.4 0.536 1.604 0.536 1.604 dbcsr_complete_redistribute 325 12.2 0.560 0.577 1.471 1.557 mp_waitany 9240 13.8 1.068 1.486 1.068 1.486 multiply_cannon_metrocomm4 6165 15.4 0.018 0.019 0.483 1.406 mp_irecv_dv 24056 15.7 0.458 1.363 0.458 1.363 generate_qs_task_list 11 8.9 0.375 0.443 1.228 1.335 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.166 1.195 rs_pw_transfer_RS2PW_140 121 11.5 0.162 0.166 0.799 1.193 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=57.109000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=785.818182, yerr=12.268349 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ef71e334ee8f071d2456394b624a331795da3ecc_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 1.328161E+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_SendRecv 2310 360267. MP_ISendRecv 2310 360267. MP_Wait 5214 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.024 0.049 90.499 90.500 qs_mol_dyn_low 1 2.0 0.003 0.003 89.848 89.853 qs_forces 11 3.9 0.003 0.003 89.741 89.745 qs_energies 11 4.9 0.001 0.002 85.544 85.548 scf_env_do_scf 11 5.9 0.001 0.001 74.563 74.564 velocity_verlet 10 3.0 0.002 0.002 56.916 56.921 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 45.170 45.173 dbcsr_multiply_generic 2055 12.4 0.118 0.123 29.761 29.831 init_scf_loop 11 6.9 0.000 0.000 29.319 29.323 prepare_preconditioner 11 7.9 0.000 0.000 27.060 27.071 make_preconditioner 11 8.9 0.000 0.000 27.060 27.071 qs_scf_new_mos 99 7.5 0.001 0.001 26.734 26.769 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.734 26.769 make_full_inverse_cholesky 11 9.9 0.000 0.000 21.219 26.523 ot_scf_mini 99 9.5 0.002 0.002 24.976 25.000 multiply_cannon 2055 13.4 0.349 0.391 22.574 23.258 multiply_cannon_loop 2055 14.4 0.340 0.342 20.776 21.127 cp_fm_upper_to_full 70 14.2 12.803 18.361 12.803 18.361 rebuild_ks_matrix 110 8.3 0.001 0.001 14.319 14.366 qs_ks_build_kohn_sham_matrix 110 9.3 0.017 0.020 14.318 14.366 ot_mini 99 10.5 0.001 0.001 13.913 13.938 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.977 13.021 dbcsr_complete_redistribute 325 12.2 1.022 1.045 7.471 10.684 multiply_cannon_multrec 8220 15.4 4.664 4.881 10.143 10.634 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.414 9.624 qs_ot_get_derivative 99 11.5 0.001 0.001 9.317 9.342 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.826 9.012 mp_waitall_1 84994 16.7 8.060 8.813 8.060 8.813 mp_alltoall_i22 605 13.7 5.451 8.653 5.451 8.653 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.809 7.843 calculate_rho_elec 110 8.6 0.224 0.225 7.808 7.843 sum_up_and_integrate 110 10.3 0.150 0.152 7.406 7.419 integrate_v_rspace 110 11.3 0.004 0.004 7.256 7.268 init_scf_run 11 5.9 0.000 0.001 6.610 6.610 scf_env_initial_rho_setup 11 6.9 0.000 0.000 6.610 6.610 make_m2s 4110 13.4 0.043 0.044 5.444 5.957 dbcsr_mm_accdrv_process 11614 15.7 3.267 3.665 5.337 5.948 qs_ot_get_p 110 10.4 0.001 0.001 5.828 5.853 make_images 4110 14.4 0.874 0.910 5.255 5.768 cp_fm_cholesky_invert 11 10.9 5.675 5.679 5.675 5.679 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 4.974 5.635 pw_transfer 1331 11.6 0.075 0.075 5.148 5.152 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.678 5.133 apply_single 110 13.6 0.000 0.000 4.678 5.132 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 5.032 5.035 ot_diis_step 99 11.5 0.017 0.021 4.571 4.572 fft_wrap_pw1pw2_140 451 13.1 1.335 1.343 4.356 4.361 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.256 4.263 multiply_cannon_sync_h2d 8220 15.4 3.951 3.954 3.951 3.954 calculate_dm_sparse 110 9.5 0.001 0.001 3.824 3.845 density_rs2pw 110 9.6 0.004 0.004 3.811 3.844 hybrid_alltoall_any 4261 16.3 0.256 0.553 3.049 3.817 cp_dbcsr_syevd 48 12.0 0.004 0.004 3.802 3.803 qs_energies_init_hamiltonians 11 5.9 0.002 0.004 3.768 3.771 grid_integrate_task_list 110 12.3 3.676 3.768 3.676 3.768 make_images_data 4110 15.4 0.042 0.044 3.059 3.714 wfi_extrapolate 11 7.9 0.001 0.001 3.370 3.370 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.859 3.301 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.189 3.189 cp_fm_diag_elpa_base 48 14.0 2.643 2.841 3.186 3.186 calculate_first_density_matrix 1 7.0 0.001 0.001 3.116 3.118 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.040 3.044 fft3d_ps 1111 14.6 1.305 1.317 2.961 2.967 grid_collocate_task_list 110 9.6 2.644 2.683 2.644 2.683 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.539 2.553 potential_pw2rs 110 12.3 0.022 0.022 2.454 2.460 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.428 2.441 rs_pw_transfer 902 11.9 0.011 0.011 2.341 2.382 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 2.363 2.377 qs_env_update_s_mstruct 11 6.9 0.011 0.023 2.267 2.342 cp_fm_cholesky_decompose 22 10.9 2.200 2.222 2.200 2.222 mp_alltoall_d11v 2046 13.8 2.061 2.189 2.061 2.189 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.102 2.170 jit_kernel_multiply 10 15.3 1.870 2.073 1.870 2.073 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.948 1.952 qs_create_task_list 11 7.9 0.000 0.001 1.898 1.943 generate_qs_task_list 11 8.9 0.734 0.788 1.898 1.943 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.805 1.838 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=90.500000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1199.636364, yerr=55.279904 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ef71e334ee8f071d2456394b624a331795da3ecc_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 630.345728E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175954870160 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 57910. MP_Allreduce 10979 803. MP_Sync 87 MP_Alltoall 2226 849138. MP_SendRecv 24320 18752. MP_ISendRecv 24320 18752. MP_Wait 42476 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.029 0.061 210.611 210.612 qs_mol_dyn_low 1 2.0 0.004 0.012 210.186 210.199 qs_forces 11 3.9 0.006 0.011 209.668 209.670 qs_energies 11 4.9 0.034 0.260 204.073 204.090 scf_env_do_scf 11 5.9 0.001 0.001 186.121 186.125 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 165.159 165.161 dbcsr_multiply_generic 2507 12.6 0.179 0.183 125.148 125.764 qs_scf_new_mos 117 7.6 0.001 0.001 124.826 125.102 qs_scf_loop_do_ot 117 8.6 0.001 0.001 124.825 125.101 velocity_verlet 10 3.0 0.048 0.379 124.953 124.960 ot_scf_mini 117 9.6 0.003 0.004 118.227 118.508 multiply_cannon 2507 13.6 0.241 0.251 101.603 103.078 multiply_cannon_loop 2507 14.6 2.046 2.097 99.368 100.967 ot_mini 117 10.6 0.001 0.002 65.946 66.168 multiply_cannon_multrec 60168 15.6 33.449 35.473 41.891 44.121 qs_ot_get_derivative 117 11.6 0.001 0.002 41.277 41.532 rebuild_ks_matrix 128 8.3 0.001 0.001 33.927 34.145 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.020 33.926 34.144 mp_waitall_1 267128 16.5 28.517 31.811 28.517 31.811 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.476 30.687 qs_ot_get_p 128 10.4 0.001 0.001 30.277 30.530 multiply_cannon_sync_h2d 60168 15.6 27.555 29.338 27.555 29.338 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.325 24.873 apply_single 128 13.6 0.001 0.001 24.325 24.873 ot_diis_step 117 11.6 0.011 0.036 24.448 24.470 qs_ot_p2m_diag 83 11.4 0.079 0.092 23.694 23.745 cp_dbcsr_syevd 83 12.4 0.005 0.005 21.043 21.044 init_scf_loop 11 6.9 0.000 0.000 20.882 20.883 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.738 19.986 cp_fm_diag_elpa 83 13.4 0.000 0.000 18.033 18.059 cp_fm_diag_elpa_base 83 14.4 17.964 17.995 18.029 18.054 multiply_cannon_metrocomm3 60168 15.6 0.114 0.120 15.751 17.978 prepare_preconditioner 11 7.9 0.000 0.000 16.246 16.305 make_preconditioner 11 8.9 0.000 0.000 16.246 16.305 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.465 15.636 sum_up_and_integrate 128 10.3 0.089 0.105 14.387 14.403 make_m2s 5014 13.6 0.103 0.114 13.897 14.339 integrate_v_rspace 128 11.3 0.004 0.004 14.297 14.315 qs_rho_update_rho_low 128 7.7 0.022 0.173 14.114 14.225 calculate_rho_elec 128 8.7 0.046 0.065 14.092 14.224 make_images 5014 14.6 0.398 0.415 13.716 14.166 init_scf_run 11 5.9 0.000 0.001 13.008 13.008 scf_env_initial_rho_setup 11 6.9 0.014 0.116 13.008 13.008 density_rs2pw 128 9.7 0.006 0.007 7.198 10.490 mp_sum_l 7870 13.0 8.379 9.679 8.379 9.679 cp_fm_cholesky_invert 11 10.9 9.320 9.328 9.320 9.328 rs_pw_transfer 1046 11.9 0.017 0.019 5.926 9.224 wfi_extrapolate 11 7.9 0.001 0.003 9.162 9.162 calculate_dm_sparse 128 9.5 0.001 0.001 8.540 8.642 dbcsr_mm_accdrv_process 124484 16.2 3.125 3.236 8.004 8.488 multiply_cannon_metrocomm1 60168 15.6 0.089 0.092 6.141 8.244 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.974 8.113 pw_transfer 1547 11.6 0.074 0.090 7.899 8.079 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.848 7.917 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 7.696 7.878 make_images_data 5014 15.6 0.066 0.072 6.837 7.787 grid_integrate_task_list 128 12.3 7.052 7.500 7.052 7.500 hybrid_alltoall_any 5200 16.5 0.292 2.257 5.965 7.108 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.920 6.931 fft_wrap_pw1pw2_140 523 13.2 1.272 1.314 6.689 6.884 mp_waitany 16020 13.9 2.764 6.149 2.764 6.149 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 6.004 6.098 mp_alltoall_d11v 2415 14.1 4.591 5.836 4.591 5.836 fft3d_ps 1291 14.7 2.147 2.784 5.465 5.781 grid_collocate_task_list 128 9.7 4.551 5.755 4.551 5.755 rs_pw_transfer_RS2PW_140 139 11.5 0.281 0.299 2.168 5.490 cp_fm_cholesky_decompose 22 10.9 4.722 4.737 4.722 4.737 potential_pw2rs 128 12.3 0.009 0.011 4.714 4.732 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=210.612000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=596.272727, yerr=6.903107 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ef71e334ee8f071d2456394b624a331795da3ecc_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430457200640 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986252263424 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613077360640 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239167967232 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239167967232 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.233007E+12 0.0% 0.0% 100.0% flops max/rank 2.200017E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383584 0.0% 0.0% 100.0% number of processed stacks 6024768 0.0% 0.0% 100.0% average stack size 0.0 0.0 1129.7 marketing flops 145.651870E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 828.518400E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2427840 MPI messages size (bytes): total size 4.132587E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.702166E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 71532 2339635200 32768 < size <= 131072 729952 56049532928 131072 < size <= 4194304 1387568 1410045313024 4194304 < size <= 16777216 155760 1473828432560 16777216 < size 68112 1190343475200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 57641. MP_Allreduce 11105 963. MP_Sync 88 MP_Alltoall 1983 5620942. MP_SendRecv 12126 47072. MP_ISendRecv 12126 47072. MP_Wait 26114 MP_ISend 11836 212447. MP_IRecv 11836 212447. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.076 0.106 192.346 192.348 qs_mol_dyn_low 1 2.0 0.003 0.004 191.349 191.434 qs_forces 11 3.9 0.006 0.009 191.259 191.260 qs_energies 11 4.9 0.006 0.026 184.419 184.436 scf_env_do_scf 11 5.9 0.001 0.001 167.417 167.427 scf_env_do_scf_inner_loop 118 6.6 0.004 0.011 134.614 134.617 velocity_verlet 10 3.0 0.005 0.016 120.818 120.819 dbcsr_multiply_generic 2529 12.6 0.189 0.194 97.821 99.152 qs_scf_new_mos 118 7.6 0.001 0.001 94.177 94.627 qs_scf_loop_do_ot 118 8.6 0.001 0.001 94.176 94.627 ot_scf_mini 118 9.6 0.004 0.005 89.346 89.812 multiply_cannon 2529 13.6 0.479 0.532 77.896 82.469 multiply_cannon_loop 2529 14.6 1.253 1.293 74.896 77.466 ot_mini 118 10.6 0.001 0.001 49.917 50.398 mp_waitall_1 216598 16.6 24.518 37.436 24.518 37.436 multiply_cannon_multrec 30348 15.6 22.256 26.646 32.084 36.813 rebuild_ks_matrix 129 8.3 0.001 0.001 33.572 33.985 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.021 33.571 33.984 init_scf_loop 11 6.9 0.001 0.005 32.715 32.716 qs_ks_update_qs_env 129 7.6 0.001 0.001 30.211 30.599 prepare_preconditioner 11 7.9 0.000 0.000 28.323 28.377 make_preconditioner 11 8.9 0.000 0.001 28.323 28.377 multiply_cannon_metrocomm3 30348 15.6 0.092 0.098 15.720 28.316 qs_ot_get_derivative 118 11.6 0.001 0.002 27.785 28.243 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.001 27.541 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 22.164 23.239 apply_single 129 13.6 0.001 0.001 22.163 23.238 ot_diis_step 118 11.6 0.014 0.015 21.962 21.964 multiply_cannon_sync_h2d 30348 15.6 19.553 21.678 19.553 21.678 qs_ot_get_p 129 10.4 0.001 0.001 21.102 21.664 cp_fm_cholesky_invert 11 10.9 16.630 16.643 16.630 16.643 qs_ot_p2m_diag 84 11.4 0.190 0.219 16.346 16.387 make_m2s 5058 13.6 0.090 0.096 14.270 15.584 make_images 5058 14.6 1.158 1.352 14.060 15.374 cp_dbcsr_syevd 84 12.4 0.006 0.007 15.181 15.182 sum_up_and_integrate 129 10.3 0.117 0.134 14.860 14.890 integrate_v_rspace 129 11.3 0.004 0.004 14.743 14.779 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.440 14.542 calculate_rho_elec 129 8.7 0.088 0.105 14.439 14.541 init_scf_run 11 5.9 0.000 0.001 12.022 12.024 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.022 12.024 cp_fm_diag_elpa 84 13.4 0.000 0.001 11.977 12.011 cp_fm_diag_elpa_base 84 14.4 11.720 11.825 11.971 12.001 qs_ot_get_derivative_diag 78 12.4 0.002 0.004 11.004 11.389 multiply_cannon_metrocomm4 27819 15.6 0.101 0.116 3.771 10.537 make_images_data 5058 15.6 0.064 0.071 8.698 10.314 mp_irecv_dv 70084 16.3 3.571 10.147 3.571 10.147 density_rs2pw 129 9.7 0.006 0.007 7.552 9.984 dbcsr_mm_accdrv_process 62758 16.2 4.604 5.495 9.278 9.901 hybrid_alltoall_any 5245 16.5 0.345 1.532 7.397 9.373 pw_transfer 1559 11.6 0.087 0.094 8.860 8.915 fft_wrap_pw1pw2 1301 12.7 0.011 0.011 8.634 8.690 rs_pw_transfer 1054 12.0 0.015 0.017 5.955 8.449 wfi_extrapolate 11 7.9 0.001 0.001 8.375 8.375 grid_integrate_task_list 129 12.3 7.190 7.722 7.190 7.722 fft_wrap_pw1pw2_140 527 13.2 1.346 1.364 7.617 7.697 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.144 6.914 cp_fm_cholesky_decompose 22 10.9 6.811 6.897 6.811 6.897 calculate_dm_sparse 129 9.5 0.001 0.001 6.683 6.833 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.121 6.132 fft3d_ps 1301 14.7 2.836 3.002 5.921 5.974 grid_collocate_task_list 129 9.7 4.739 5.942 4.739 5.942 mp_alltoall_d11v 2429 14.1 4.402 5.470 4.402 5.470 qs_ot_get_orbitals 118 10.6 0.001 0.001 5.387 5.459 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 5.302 5.447 mp_sum_l 7936 13.1 3.861 5.281 3.861 5.281 mp_waitany 11836 13.9 2.639 5.194 2.639 5.194 potential_pw2rs 129 12.3 0.016 0.019 5.033 5.050 rs_pw_transfer_RS2PW_140 140 11.5 0.355 0.401 2.237 4.711 mp_allgather_i34 2529 14.6 1.567 4.679 1.567 4.679 mp_sum_d 4509 12.1 2.670 4.005 2.670 4.005 dbcsr_complete_redistribute 397 12.7 0.769 0.847 3.124 3.965 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=192.348000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=790.181818, yerr=1.849726 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ef71e334ee8f071d2456394b624a331795da3ecc_performance_tests/17/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 2.906045E+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 3951168 0.0% 0.0% 100.0% average stack size 0.0 0.0 1697.1 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 967.340032E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1033760 MPI messages size (bytes): total size 2.695213E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.607194E+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 279168 36591108096 131072 < size <= 4194304 654272 987691483136 4194304 < size <= 16777216 65184 925172769472 16777216 < size 28448 745747251200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4007 58141. MP_Allreduce 11015 1006. MP_Sync 86 MP_Alltoall 1700 9383497. MP_SendRecv 7874 75008. MP_ISendRecv 7874 75008. MP_Wait 21654 MP_ISend 11660 275234. MP_IRecv 11660 275234. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.060 0.151 176.508 176.510 qs_mol_dyn_low 1 2.0 0.003 0.003 175.837 175.851 qs_forces 11 3.9 0.004 0.005 175.735 175.743 qs_energies 11 4.9 0.011 0.048 169.096 169.106 scf_env_do_scf 11 5.9 0.001 0.001 152.290 152.291 scf_env_do_scf_inner_loop 116 6.6 0.004 0.008 117.169 117.170 velocity_verlet 10 3.0 0.002 0.002 112.117 112.119 dbcsr_multiply_generic 2485 12.5 0.177 0.181 81.510 82.580 qs_scf_new_mos 116 7.6 0.001 0.001 79.183 79.593 qs_scf_loop_do_ot 116 8.6 0.001 0.001 79.182 79.592 ot_scf_mini 116 9.6 0.004 0.005 75.092 75.524 multiply_cannon 2485 13.5 0.495 0.516 61.819 66.991 multiply_cannon_loop 2485 14.5 0.841 0.874 58.990 61.368 ot_mini 116 10.6 0.001 0.001 42.262 42.701 mp_waitall_1 169034 16.6 25.340 35.261 25.340 35.261 init_scf_loop 11 6.9 0.001 0.003 35.025 35.026 rebuild_ks_matrix 127 8.3 0.001 0.001 30.614 31.080 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.018 30.613 31.079 prepare_preconditioner 11 7.9 0.000 0.001 30.987 31.037 make_preconditioner 11 8.9 0.000 0.002 30.987 31.037 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.593 30.021 qs_ks_update_qs_env 127 7.6 0.001 0.001 27.521 27.944 multiply_cannon_multrec 19880 15.5 13.240 16.421 22.506 25.597 multiply_cannon_metrocomm3 19880 15.5 0.058 0.062 15.477 25.390 qs_ot_get_derivative 116 11.6 0.001 0.002 22.495 22.938 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.943 20.857 apply_single 127 13.6 0.001 0.001 19.943 20.857 ot_diis_step 116 11.6 0.018 0.018 19.664 19.664 qs_ot_get_p 127 10.4 0.001 0.001 18.350 18.876 multiply_cannon_sync_h2d 19880 15.5 14.155 16.007 14.155 16.007 make_m2s 4970 13.5 0.081 0.088 14.835 15.595 make_images 4970 14.5 1.161 1.264 14.603 15.366 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.494 14.600 calculate_rho_elec 127 8.7 0.129 0.144 14.494 14.599 sum_up_and_integrate 127 10.3 0.131 0.144 14.565 14.597 cp_fm_cholesky_invert 11 10.9 14.468 14.477 14.468 14.477 integrate_v_rspace 127 11.3 0.004 0.004 14.434 14.463 qs_ot_p2m_diag 82 11.4 0.262 0.269 14.234 14.247 cp_dbcsr_syevd 82 12.4 0.006 0.006 13.227 13.227 init_scf_run 11 5.9 0.000 0.001 11.271 11.271 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.271 11.271 make_images_data 4970 15.5 0.060 0.069 9.305 10.464 cp_fm_diag_elpa 82 13.4 0.000 0.000 10.111 10.127 cp_fm_diag_elpa_base 82 14.4 9.706 9.856 10.107 10.124 hybrid_alltoall_any 5155 16.4 0.431 1.969 8.009 9.659 multiply_cannon_metrocomm4 17395 15.5 0.064 0.076 3.482 9.469 mp_irecv_dv 49801 16.2 3.357 9.211 3.357 9.211 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 8.705 9.013 pw_transfer 1535 11.6 0.085 0.099 8.829 8.928 density_rs2pw 127 9.7 0.006 0.007 7.147 8.901 dbcsr_mm_accdrv_process 41158 16.2 4.920 5.819 8.731 8.853 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.607 8.707 fft_wrap_pw1pw2_140 519 13.2 1.402 1.426 7.636 7.754 grid_integrate_task_list 127 12.3 7.280 7.731 7.280 7.731 wfi_extrapolate 11 7.9 0.001 0.001 7.404 7.404 cp_fm_cholesky_decompose 22 10.9 7.326 7.389 7.326 7.389 rs_pw_transfer 1038 11.9 0.014 0.015 5.303 7.329 cp_fm_upper_to_full 104 14.8 5.738 7.276 5.738 7.276 dbcsr_complete_redistribute 393 12.7 1.166 1.195 4.604 6.342 calculate_dm_sparse 127 9.5 0.001 0.001 6.125 6.197 fft3d_ps 1281 14.7 2.721 2.950 5.753 5.824 grid_collocate_task_list 127 9.7 4.877 5.755 4.877 5.755 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.534 5.538 mp_alltoall_d11v 2401 14.1 4.570 5.413 4.570 5.413 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.517 5.188 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.427 5.166 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.657 4.790 potential_pw2rs 127 12.3 0.021 0.023 4.700 4.713 mp_sum_l 7804 13.0 3.128 4.659 3.128 4.659 mp_waitany 11660 13.9 2.492 4.617 2.492 4.617 mp_allgather_i34 2485 14.5 1.365 4.537 1.365 4.537 qs_energies_init_hamiltonians 11 5.9 0.030 0.068 4.213 4.214 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.374 4.076 rs_pw_transfer_RS2PW_140 138 11.5 0.329 0.351 2.051 4.073 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.990 4.027 mp_alltoall_i22 712 14.1 1.925 3.800 1.925 3.800 calculate_first_density_matrix 1 7.0 0.076 0.302 3.442 3.665 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.585 3.629 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=176.510000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=910.272727, yerr=20.933188 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ef71e334ee8f071d2456394b624a331795da3ecc_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 4.353791E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 5977344 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.139638E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1143192 MPI messages size (bytes): total size 2.023815E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.770320E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 319024 36042702848 131072 < size <= 4194304 715736 785529176064 4194304 < size <= 16777216 70320 665379475120 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 58210. MP_Allreduce 11002 1090. MP_Sync 87 MP_Alltoall 1712 12503107. MP_SendRecv 5888 75008. MP_ISendRecv 5888 75008. MP_Wait 22442 MP_ISend 14952 244818. MP_IRecv 14952 244818. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.094 0.112 187.932 187.944 qs_mol_dyn_low 1 2.0 0.003 0.003 187.310 187.329 qs_forces 11 3.9 0.005 0.007 187.175 187.186 qs_energies 11 4.9 0.001 0.002 179.983 179.994 scf_env_do_scf 11 5.9 0.001 0.001 162.308 162.316 velocity_verlet 10 3.0 0.002 0.004 124.012 124.014 scf_env_do_scf_inner_loop 117 6.6 0.003 0.010 116.232 116.234 dbcsr_multiply_generic 2507 12.6 0.186 0.192 80.159 80.868 qs_scf_new_mos 117 7.6 0.001 0.001 78.789 79.089 qs_scf_loop_do_ot 117 8.6 0.001 0.001 78.788 79.089 ot_scf_mini 117 9.6 0.003 0.003 74.327 74.602 multiply_cannon 2507 13.6 0.560 0.602 56.178 58.784 multiply_cannon_loop 2507 14.6 1.171 1.201 52.310 54.027 init_scf_loop 11 6.9 0.000 0.000 45.953 45.954 ot_mini 117 10.6 0.001 0.001 42.278 42.549 prepare_preconditioner 11 7.9 0.000 0.000 41.831 41.855 make_preconditioner 11 8.9 0.000 0.000 41.831 41.855 make_full_inverse_cholesky 11 9.9 0.000 0.000 35.497 40.519 multiply_cannon_multrec 30084 15.6 14.077 19.166 26.852 31.722 rebuild_ks_matrix 128 8.3 0.001 0.001 29.860 30.082 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.020 29.859 30.081 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.905 27.107 mp_waitall_1 147882 16.7 17.191 26.423 17.191 26.423 qs_ot_get_derivative 117 11.6 0.001 0.002 22.800 23.075 make_m2s 5014 13.6 0.096 0.101 19.687 20.789 make_images 5014 14.6 1.915 2.215 19.382 20.487 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 18.896 19.398 apply_single 128 13.6 0.001 0.001 18.895 19.398 ot_diis_step 117 11.6 0.017 0.018 19.354 19.356 qs_ot_get_p 128 10.4 0.001 0.001 18.419 18.716 cp_fm_upper_to_full 105 14.8 11.442 16.879 11.442 16.879 cp_fm_cholesky_invert 11 10.9 15.881 15.890 15.881 15.890 multiply_cannon_metrocomm3 30084 15.6 0.047 0.049 6.634 15.779 sum_up_and_integrate 128 10.3 0.140 0.154 14.739 14.766 integrate_v_rspace 128 11.3 0.004 0.004 14.598 14.629 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.543 14.580 calculate_rho_elec 128 8.7 0.173 0.189 14.542 14.579 qs_ot_p2m_diag 83 11.4 0.343 0.391 14.308 14.360 cp_dbcsr_syevd 83 12.4 0.006 0.007 13.039 13.041 dbcsr_mm_accdrv_process 62264 16.2 7.774 9.151 12.349 12.938 dbcsr_complete_redistribute 395 12.7 1.497 1.624 8.873 12.661 multiply_cannon_sync_h2d 30084 15.6 11.707 12.463 11.707 12.463 make_images_data 5014 15.6 0.063 0.067 10.558 12.294 init_scf_run 11 5.9 0.000 0.001 11.830 11.831 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.829 11.831 copy_fm_to_dbcsr 209 11.7 0.002 0.002 7.515 11.298 hybrid_alltoall_any 5200 16.5 0.525 2.215 9.477 11.252 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.311 10.016 cp_fm_diag_elpa 83 13.4 0.000 0.001 10.000 10.013 cp_fm_diag_elpa_base 83 14.4 9.041 9.366 9.994 10.006 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.322 9.529 mp_alltoall_i22 716 14.1 5.600 9.494 5.600 9.494 pw_transfer 1547 11.6 0.086 0.101 9.156 9.236 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.931 9.016 density_rs2pw 128 9.7 0.006 0.007 6.976 8.435 fft_wrap_pw1pw2_140 523 13.2 1.560 1.589 7.954 8.038 grid_integrate_task_list 128 12.3 7.517 7.945 7.517 7.945 wfi_extrapolate 11 7.9 0.001 0.001 7.636 7.636 cp_fm_cholesky_decompose 22 10.9 7.467 7.544 7.467 7.544 multiply_cannon_metrocomm4 25070 15.6 0.084 0.096 2.804 7.200 calculate_dm_sparse 128 9.5 0.001 0.001 6.879 6.987 mp_irecv_dv 76098 16.2 2.650 6.923 2.650 6.923 rs_pw_transfer 1046 11.9 0.013 0.014 4.641 6.155 fft3d_ps 1291 14.7 2.838 2.896 5.856 5.918 grid_collocate_task_list 128 9.7 5.075 5.870 5.075 5.870 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.762 5.811 mp_alltoall_d11v 2415 14.1 4.793 5.437 4.793 5.437 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.709 4.792 potential_pw2rs 128 12.3 0.023 0.023 4.511 4.523 qs_energies_init_hamiltonians 11 5.9 0.002 0.003 4.455 4.459 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.344 4.429 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.147 4.203 calculate_first_density_matrix 1 7.0 0.000 0.001 3.994 3.996 mp_waitany 14952 13.9 2.246 3.779 2.246 3.779 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=187.944000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1072.545455, yerr=19.270154 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ef71e334ee8f071d2456394b624a331795da3ecc_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456039424 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613072052224 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176077312 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176077312 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.233020E+12 0.0% 0.0% 100.0% flops max/rank 5.910120E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383904 0.0% 0.0% 100.0% number of processed stacks 1976928 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.9 marketing flops 145.650931E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 1.509155E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 242784 MPI messages size (bytes): total size 1.341806E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.526748E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 115008 60297314304 4194304 < size <= 16777216 105840 554906419200 16777216 < size 20352 726592466352 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 9010 51. MP_Alltoall 9724 794507. MP_ISend 40420 2100460. MP_IRecv 40420 2099564. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4039 57686. MP_Allreduce 11094 1171. MP_Sync 88 MP_Alltoall 1724 18848034. MP_SendRecv 3870 122880. MP_ISendRecv 3870 122880. MP_Wait 16244 MP_ISend 10760 423501. MP_IRecv 10760 423501. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.028 0.071 176.338 176.344 qs_mol_dyn_low 1 2.0 0.003 0.003 175.564 175.575 qs_forces 11 3.9 0.016 0.021 174.996 175.000 qs_energies 11 4.9 0.005 0.010 167.416 167.427 scf_env_do_scf 11 5.9 0.001 0.001 149.210 149.212 velocity_verlet 10 3.0 0.046 0.047 115.376 115.405 scf_env_do_scf_inner_loop 118 6.6 0.003 0.010 113.383 113.384 dbcsr_multiply_generic 2529 12.6 0.181 0.193 75.565 76.083 qs_scf_new_mos 118 7.6 0.001 0.001 74.466 74.587 qs_scf_loop_do_ot 118 8.6 0.001 0.001 74.465 74.586 ot_scf_mini 118 9.6 0.003 0.004 70.006 70.115 multiply_cannon 2529 13.6 0.590 0.638 55.671 60.055 multiply_cannon_loop 2529 14.6 0.444 0.455 51.057 52.074 ot_mini 118 10.6 0.001 0.001 39.825 39.923 init_scf_loop 11 6.9 0.000 0.000 35.681 35.682 mp_waitall_1 126876 16.7 26.300 33.765 26.300 33.765 prepare_preconditioner 11 7.9 0.000 0.000 31.739 31.761 make_preconditioner 11 8.9 0.000 0.000 31.739 31.761 rebuild_ks_matrix 129 8.3 0.001 0.001 30.090 30.200 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.019 30.089 30.199 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.632 29.913 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.370 27.475 multiply_cannon_multrec 10116 15.6 10.515 14.384 18.368 21.612 ot_diis_step 118 11.6 0.020 0.021 20.056 20.056 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.573 19.904 apply_single 129 13.6 0.001 0.001 19.572 19.904 qs_ot_get_derivative 118 11.6 0.002 0.002 19.700 19.803 multiply_cannon_metrocomm3 10116 15.6 0.023 0.024 12.346 19.323 make_m2s 5058 13.6 0.067 0.071 16.297 18.669 make_images 5058 14.6 2.300 2.807 15.989 18.352 cp_fm_cholesky_invert 11 10.9 18.103 18.111 18.103 18.111 qs_ot_get_p 129 10.4 0.001 0.001 17.201 17.317 qs_rho_update_rho_low 129 7.7 0.001 0.001 15.440 15.476 calculate_rho_elec 129 8.7 0.258 0.269 15.439 15.475 sum_up_and_integrate 129 10.3 0.183 0.194 15.099 15.144 integrate_v_rspace 129 11.3 0.004 0.004 14.915 14.969 qs_ot_p2m_diag 84 11.4 0.502 0.508 13.480 13.497 make_images_data 5058 15.6 0.052 0.060 9.993 12.511 multiply_cannon_sync_h2d 10116 15.6 11.719 12.350 11.719 12.350 cp_dbcsr_syevd 84 12.4 0.383 0.384 12.334 12.336 hybrid_alltoall_any 5245 16.5 0.830 3.668 9.866 12.277 init_scf_run 11 5.9 0.000 0.001 11.338 11.338 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.338 11.338 pw_transfer 1559 11.6 0.087 0.098 9.515 9.538 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 9.290 9.320 cp_fm_diag_elpa 84 13.4 0.000 0.000 8.947 8.959 cp_fm_diag_elpa_base 84 14.4 8.687 8.776 8.929 8.941 grid_integrate_task_list 129 12.3 7.791 8.312 7.791 8.312 dbcsr_mm_accdrv_process 20934 16.1 2.699 3.444 7.477 8.309 fft_wrap_pw1pw2_140 527 13.2 1.929 1.966 8.225 8.253 cp_fm_cholesky_decompose 22 10.9 7.974 8.090 7.974 8.090 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 7.976 8.054 multiply_cannon_metrocomm1 10116 15.6 0.029 0.029 4.733 7.810 density_rs2pw 129 9.7 0.006 0.006 6.915 7.778 wfi_extrapolate 11 7.9 0.001 0.001 7.406 7.406 mp_allgather_i34 2529 14.6 2.809 7.334 2.809 7.334 calculate_dm_sparse 129 9.5 0.001 0.001 6.714 6.801 grid_collocate_task_list 129 9.7 5.439 6.113 5.439 6.113 mp_alltoall_d11v 2429 14.1 5.041 6.106 5.041 6.106 fft3d_ps 1301 14.7 2.789 2.870 5.737 5.769 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.533 5.542 dbcsr_complete_redistribute 397 12.7 2.125 2.166 5.081 5.467 qs_energies_init_hamiltonians 11 5.9 0.021 0.036 5.338 5.345 rs_pw_transfer 1054 12.0 0.013 0.013 4.139 5.036 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.480 4.501 potential_pw2rs 129 12.3 0.027 0.028 4.426 4.440 multiply_cannon_metrocomm4 7587 15.6 0.025 0.028 1.871 4.233 mp_irecv_dv 29102 15.9 1.833 4.154 1.833 4.154 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.583 3.897 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.707 3.741 calculate_first_density_matrix 1 7.0 0.001 0.001 3.720 3.722 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.628 3.638 copy_fm_to_dbcsr 210 11.7 0.001 0.002 3.318 3.623 copy_dbcsr_to_fm 187 11.8 0.004 0.004 3.494 3.587 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=176.344000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1401.636364, yerr=47.772706 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ef71e334ee8f071d2456394b624a331795da3ecc_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1399806074880 0.0% 0.0% 100.0% flops 32 x 32 x 32 1906965479424 0.0% 0.0% 100.0% flops 22 x 9 x 32 1943685697536 0.0% 0.0% 100.0% flops 9 x 22 x 32 1949309632512 0.0% 0.0% 100.0% flops 22 x 22 x 32 2694948675584 0.0% 0.0% 100.0% flops 32 x 32 x 9 4338990710784 0.0% 0.0% 100.0% flops 32 x 32 x 22 5303210868736 0.0% 0.0% 100.0% flops 9 x 32 x 32 5347334946816 0.0% 0.0% 100.0% flops 22 x 32 x 32 6535631601664 0.0% 0.0% 100.0% flops 9 x 32 x 9 11360521838592 0.0% 0.0% 100.0% flops 22 x 32 x 9 14909195169792 0.0% 0.0% 100.0% flops 9 x 32 x 22 14909195169792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19481731219456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.080527E+12 0.0% 0.0% 100.0% flops max/rank 11.516854E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6655092288 0.0% 0.0% 100.0% number of processed stacks 1931936 0.0% 0.0% 100.0% average stack size 0.0 0.0 3444.8 marketing flops 142.438295E+12 ------------------------------------------------------------------------------- # multiplications 2464 max memory usage/rank 3.106001E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 98560 MPI messages size (bytes): total size 1.118682E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.350266E+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 44368 34435235840 4194304 < size <= 16777216 43616 373376942080 16777216 < size 9960 710866914816 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4018 58913. MP_Allreduce 11042 1514. MP_Sync 87 MP_Alltoall 1688 36934192. MP_SendRecv 1764 218624. MP_ISendRecv 1764 218624. MP_Wait 9654 MP_ISend 6312 1080635. MP_IRecv 6312 1080635. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.024 0.043 297.467 297.469 qs_mol_dyn_low 1 2.0 0.003 0.003 296.522 296.556 qs_forces 11 3.9 0.005 0.005 296.357 296.360 qs_energies 11 4.9 0.002 0.002 287.003 287.016 scf_env_do_scf 11 5.9 0.001 0.001 264.345 264.366 velocity_verlet 10 3.0 0.008 0.009 212.777 212.793 scf_env_do_scf_inner_loop 115 6.6 0.013 0.082 139.157 139.158 init_scf_loop 11 6.9 0.000 0.000 124.934 124.936 prepare_preconditioner 11 7.9 0.000 0.000 119.905 119.925 make_preconditioner 11 8.9 0.000 0.000 119.905 119.925 make_full_inverse_cholesky 11 9.9 0.000 0.000 96.158 117.061 qs_scf_new_mos 115 7.6 0.001 0.001 89.188 89.307 qs_scf_loop_do_ot 115 8.6 0.001 0.001 89.187 89.306 ot_scf_mini 115 9.6 0.004 0.004 84.522 84.563 dbcsr_multiply_generic 2464 12.5 0.208 0.220 82.697 83.222 cp_fm_upper_to_full 105 14.8 53.654 76.387 53.654 76.387 multiply_cannon 2464 13.5 0.692 0.738 58.536 59.113 multiply_cannon_loop 2464 14.5 0.461 0.471 54.920 56.236 ot_mini 115 10.6 0.001 0.001 43.924 43.966 dbcsr_complete_redistribute 395 12.7 3.996 4.037 29.343 42.044 copy_fm_to_dbcsr 209 11.7 0.002 0.002 25.974 38.692 rebuild_ks_matrix 126 8.3 0.001 0.001 37.762 37.804 qs_ks_build_kohn_sham_matrix 126 9.3 0.018 0.018 37.761 37.804 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 23.702 36.355 qs_ks_update_qs_env 126 7.6 0.001 0.001 34.764 34.811 mp_alltoall_i22 712 14.1 21.511 34.326 21.511 34.326 cp_fm_cholesky_invert 11 10.9 33.442 33.448 33.442 33.448 mp_waitall_1 101902 16.8 27.920 31.850 27.920 31.850 qs_ot_get_p 126 10.4 0.001 0.001 25.443 25.452 qs_ot_get_derivative 115 11.6 0.002 0.002 24.199 24.243 qs_ot_p2m_diag 83 11.4 0.880 0.883 21.502 21.528 make_m2s 4928 13.5 0.076 0.080 19.662 20.630 qs_rho_update_rho_low 126 7.7 0.001 0.001 20.187 20.211 calculate_rho_elec 126 8.7 0.474 0.475 20.187 20.210 make_images 4928 14.5 3.695 3.917 19.189 20.161 cp_dbcsr_syevd 83 12.4 0.006 0.007 19.787 19.789 multiply_cannon_metrocomm3 9856 15.5 0.023 0.023 18.556 19.759 ot_diis_step 115 11.6 0.022 0.022 19.699 19.700 sum_up_and_integrate 126 10.3 0.316 0.319 19.443 19.525 apply_preconditioner_dbcsr 126 12.6 0.000 0.000 19.006 19.209 apply_single 126 13.6 0.001 0.001 19.006 19.209 integrate_v_rspace 126 11.3 0.004 0.005 19.127 19.208 multiply_cannon_multrec 9856 15.5 10.366 12.164 17.944 18.241 cp_fm_diag_elpa 83 13.4 0.000 0.000 16.564 16.564 cp_fm_diag_elpa_base 83 14.4 12.066 13.650 16.560 16.560 multiply_cannon_sync_h2d 9856 15.5 15.417 15.423 15.417 15.423 init_scf_run 11 5.9 0.000 0.001 12.680 12.680 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.680 12.680 hybrid_alltoall_any 5114 16.4 1.278 2.990 10.711 12.606 make_images_data 4928 15.5 0.058 0.064 10.554 12.212 pw_transfer 1523 11.6 0.091 0.092 11.244 11.250 fft_wrap_pw1pw2 1271 12.7 0.011 0.011 11.015 11.020 fft_wrap_pw1pw2_140 515 13.2 3.024 3.035 9.791 9.799 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.372 9.409 dbcsr_mm_accdrv_process 20422 16.0 3.712 5.491 7.344 9.347 mp_alltoall_d11v 2399 14.1 8.389 9.243 8.389 9.243 wfi_extrapolate 11 7.9 0.001 0.001 9.004 9.004 cp_fm_cholesky_decompose 22 10.9 8.934 8.964 8.934 8.964 grid_integrate_task_list 126 12.3 8.452 8.661 8.452 8.661 density_rs2pw 126 9.7 0.005 0.006 7.877 8.010 qs_energies_init_hamiltonians 11 5.9 0.002 0.002 8.001 8.002 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.840 6.948 calculate_dm_sparse 126 9.5 0.001 0.001 6.523 6.621 grid_collocate_task_list 126 9.7 6.263 6.303 6.263 6.303 rs_scatter_matrices 137 9.7 3.630 4.574 6.035 6.245 fft3d_ps 1271 14.7 2.796 2.801 6.167 6.187 copy_dbcsr_to_fm 186 11.8 0.004 0.004 6.040 6.139 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=297.469000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2795.818182, yerr=165.834429 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ef71e334ee8f071d2456394b624a331795da3ecc_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 1.262662E+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 8077588. MP_SendRecv 285 19200. MP_ISendRecv 285 19200. MP_Wait 1017 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.032 85.811 85.811 qs_energies 1 2.0 0.000 0.000 85.316 85.384 ls_scf 1 3.0 0.000 0.000 84.166 84.233 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.966 73.165 multiply_cannon 111 7.7 0.018 0.021 56.317 57.465 multiply_cannon_loop 111 8.7 0.208 0.221 52.916 54.368 ls_scf_main 1 4.0 0.000 0.000 52.205 52.211 density_matrix_trs4 2 5.0 0.002 0.003 46.705 46.795 ls_scf_init_scf 1 4.0 0.028 0.046 28.912 28.922 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.631 27.687 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.230 25.249 mp_waitall_1 11031 10.9 22.424 24.833 22.424 24.833 multiply_cannon_multrec 2664 9.7 8.188 8.888 15.623 17.358 multiply_cannon_sync_h2d 2664 9.7 13.720 15.252 13.720 15.252 make_m2s 222 7.7 0.008 0.010 12.978 13.486 make_images 222 8.7 0.098 0.108 12.956 13.467 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.570 11.879 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.588 8.392 make_images_data 222 9.7 0.004 0.005 7.601 8.152 dbcsr_mm_accdrv_process 4760 10.4 0.513 0.628 7.052 8.073 hybrid_alltoall_any 227 10.6 0.216 1.854 6.490 7.701 dbcsr_mm_accdrv_process_sort 4760 11.4 6.208 7.089 6.208 7.089 calculate_norms 4752 9.8 5.543 6.215 5.543 6.215 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.254 5.397 mp_sum_l 807 5.4 3.133 5.052 3.133 5.052 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.345 3.662 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.081 3.644 mp_irecv_dv 6231 10.9 2.064 3.618 2.064 3.618 make_images_sizes 222 9.7 0.000 0.000 0.756 3.556 mp_alltoall_i44 222 10.7 0.756 3.556 0.756 3.556 arnoldi_extremal 4 6.8 0.000 0.000 3.214 3.241 arnoldi_normal_ev 4 7.8 0.007 0.024 3.214 3.241 ls_scf_post 1 4.0 0.000 0.000 3.048 3.115 build_subspace 16 8.4 0.009 0.012 3.097 3.100 ls_scf_store_result 1 5.0 0.000 0.000 2.846 2.891 dbcsr_special_finalize 555 9.7 0.005 0.006 2.269 2.745 dbcsr_merge_single_wm 555 10.7 0.450 0.573 2.261 2.737 make_images_pack 222 9.7 2.207 2.619 2.209 2.621 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.313 2.560 dbcsr_matrix_vector_mult_local 304 10.0 2.068 2.467 2.070 2.469 dbcsr_sort_data 658 11.4 2.070 2.466 2.070 2.466 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.271 2.364 buffer_matrices_ensure_size 222 8.7 1.753 2.102 1.753 2.102 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.784 1.794 rebuild_ks_matrix 3 7.3 0.000 0.000 1.775 1.785 qs_ks_build_kohn_sham_matrix 3 8.3 0.006 0.015 1.774 1.785 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.811000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1133.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ef71e334ee8f071d2456394b624a331795da3ecc_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 2.085151E+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_SendRecv 141 57600. MP_ISendRecv 141 57600. MP_Wait 687 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.064 0.070 91.852 91.853 qs_energies 1 2.0 0.000 0.000 91.366 91.370 ls_scf 1 3.0 0.000 0.000 89.730 89.733 dbcsr_multiply_generic 111 6.7 0.015 0.016 75.663 75.979 multiply_cannon 111 7.7 0.030 0.045 53.694 57.285 ls_scf_main 1 4.0 0.000 0.000 54.846 54.879 multiply_cannon_loop 111 8.7 0.115 0.123 50.406 53.184 density_matrix_trs4 2 5.0 0.002 0.003 49.122 49.336 ls_scf_init_scf 1 4.0 0.000 0.001 31.345 31.351 mp_waitall_1 9105 10.9 21.588 30.310 21.588 30.310 ls_scf_init_matrix_S 1 5.0 0.000 0.000 30.092 30.195 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.380 27.390 multiply_cannon_multrec 1332 9.7 13.055 16.826 22.360 27.182 multiply_cannon_metrocomm3 1332 9.7 0.006 0.007 12.122 21.126 make_m2s 222 7.7 0.006 0.007 15.326 16.068 make_images 222 8.7 1.562 1.959 15.296 16.040 dbcsr_mm_accdrv_process 4041 10.4 0.360 0.698 8.904 10.413 make_images_data 222 9.7 0.004 0.004 8.848 9.942 dbcsr_mm_accdrv_process_sort 4041 11.4 8.367 9.829 8.367 9.829 hybrid_alltoall_any 227 10.6 0.519 2.472 8.267 9.230 mp_sum_l 807 5.4 5.346 8.162 5.346 8.162 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.231 7.781 mp_irecv_dv 3311 11.0 3.210 7.725 3.210 7.725 calculate_norms 2376 9.8 5.991 6.777 5.991 6.777 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.123 6.651 multiply_cannon_sync_h2d 1332 9.7 4.786 5.784 4.786 5.784 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.375 5.586 arnoldi_extremal 4 6.8 0.000 0.000 4.691 4.706 arnoldi_normal_ev 4 7.8 0.001 0.005 4.690 4.706 build_subspace 16 8.4 0.014 0.021 4.439 4.442 ls_scf_post 1 4.0 0.000 0.000 3.539 3.542 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.147 3.391 ls_scf_store_result 1 5.0 0.000 0.000 3.269 3.364 dbcsr_matrix_vector_mult_local 304 10.0 2.761 3.250 2.763 3.252 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.247 2.869 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.581 2.669 mp_allgather_i34 111 8.7 0.982 2.392 0.982 2.392 make_images_pack 222 9.7 2.018 2.376 2.020 2.378 dbcsr_sort_data 436 11.2 1.832 2.061 1.832 2.061 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.940 1.947 rebuild_ks_matrix 3 7.3 0.000 0.000 1.927 1.935 qs_ks_build_kohn_sham_matrix 3 8.3 0.034 0.048 1.927 1.935 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.853000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1763.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ef71e334ee8f071d2456394b624a331795da3ecc_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 2.760483E+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_SendRecv 93 57600. MP_ISendRecv 93 57600. MP_Wait 639 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.031 0.047 94.545 94.546 qs_energies 1 2.0 0.000 0.000 94.007 94.013 ls_scf 1 3.0 0.000 0.000 92.598 92.604 dbcsr_multiply_generic 111 6.7 0.016 0.016 77.284 77.587 ls_scf_main 1 4.0 0.000 0.000 57.729 57.734 multiply_cannon 111 7.7 0.041 0.066 53.218 56.858 multiply_cannon_loop 111 8.7 0.099 0.105 49.637 53.913 density_matrix_trs4 2 5.0 0.002 0.003 51.749 51.961 mp_waitall_1 7281 11.0 24.196 34.305 24.196 34.305 ls_scf_init_scf 1 4.0 0.013 0.039 31.324 31.327 ls_scf_init_matrix_S 1 5.0 0.000 0.000 30.116 30.203 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.570 27.580 multiply_cannon_multrec 888 9.7 12.521 15.241 21.076 24.373 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.147 23.424 make_m2s 222 7.7 0.006 0.007 17.134 18.361 make_images 222 8.7 1.970 2.294 17.096 18.323 hybrid_alltoall_any 227 10.6 0.622 2.858 9.444 10.938 make_images_data 222 9.7 0.003 0.004 9.745 10.797 dbcsr_mm_accdrv_process 3754 10.4 0.263 0.480 8.080 9.312 mp_sum_l 807 5.4 5.370 9.187 5.370 9.187 dbcsr_mm_accdrv_process_sort 3754 11.4 7.651 8.837 7.651 8.837 multiply_cannon_sync_h2d 888 9.7 6.020 8.017 6.020 8.017 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.488 7.314 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.184 7.273 mp_irecv_dv 2335 11.1 2.472 7.271 2.472 7.271 multiply_cannon_metrocomm1 888 9.7 0.002 0.003 3.930 7.037 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.000 5.216 arnoldi_extremal 4 6.8 0.000 0.000 5.072 5.082 arnoldi_normal_ev 4 7.8 0.003 0.007 5.072 5.082 build_subspace 16 8.4 0.014 0.020 4.758 4.767 calculate_norms 1584 9.8 4.282 4.644 4.282 4.644 mp_allgather_i34 111 8.7 1.427 3.879 1.427 3.879 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.447 3.770 dbcsr_matrix_vector_mult_local 304 10.0 3.028 3.592 3.030 3.594 ls_scf_post 1 4.0 0.000 0.000 3.544 3.551 ls_scf_store_result 1 5.0 0.000 0.000 3.296 3.374 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.859 2.973 make_images_sizes 222 9.7 0.000 0.000 1.082 2.327 mp_alltoall_i44 222 10.7 1.082 2.327 1.082 2.327 dbcsr_sort_data 325 11.1 1.882 2.119 1.882 2.119 make_images_pack 222 9.7 1.813 2.063 1.816 2.066 dbcsr_data_release 9322 10.9 1.301 1.966 1.301 1.966 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.936 1.937 rebuild_ks_matrix 3 7.3 0.000 0.000 1.918 1.919 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.917 1.919 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=94.546000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2174.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ef71e334ee8f071d2456394b624a331795da3ecc_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 3.338531E+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_SendRecv 69 86400. MP_ISendRecv 69 86400. MP_Wait 531 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.115 0.630 98.938 98.939 qs_energies 1 2.0 0.000 0.000 97.280 97.297 ls_scf 1 3.0 0.003 0.021 95.206 95.224 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.999 79.253 ls_scf_main 1 4.0 0.001 0.012 58.706 58.710 multiply_cannon 111 7.7 0.059 0.120 52.006 56.101 density_matrix_trs4 2 5.0 0.003 0.014 52.543 52.635 multiply_cannon_loop 111 8.7 0.114 0.125 47.035 49.615 ls_scf_init_scf 1 4.0 0.000 0.002 33.262 33.264 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.071 32.127 mp_waitall_1 6369 11.0 23.036 31.194 23.036 31.194 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.035 29.049 multiply_cannon_multrec 1332 9.7 14.159 16.746 22.078 25.238 make_m2s 222 7.7 0.007 0.007 21.129 22.470 make_images 222 8.7 3.133 3.588 21.078 22.422 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.474 17.688 make_images_data 222 9.7 0.004 0.004 11.775 13.404 hybrid_alltoall_any 227 10.6 0.798 3.741 11.195 12.912 dbcsr_mm_accdrv_process 3641 10.4 0.212 0.413 7.546 9.537 dbcsr_mm_accdrv_process_sort 3641 11.4 7.075 8.541 7.075 8.541 mp_sum_l 807 5.4 4.284 7.160 4.284 7.160 multiply_cannon_sync_h2d 1332 9.7 5.503 6.264 5.503 6.264 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.092 6.006 mp_irecv_dv 3229 10.9 2.069 5.928 2.069 5.928 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.045 5.581 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.618 5.513 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.124 5.286 arnoldi_extremal 4 6.8 0.000 0.000 5.237 5.251 arnoldi_normal_ev 4 7.8 0.001 0.005 5.237 5.251 build_subspace 16 8.4 0.014 0.021 4.889 4.900 calculate_norms 2376 9.8 4.195 4.497 4.195 4.497 mp_allgather_i34 111 8.7 2.098 4.231 2.098 4.231 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.586 3.886 dbcsr_matrix_vector_mult_local 304 10.0 3.199 3.686 3.201 3.688 dbcsr_sort_data 658 11.4 3.098 3.418 3.098 3.418 ls_scf_post 1 4.0 0.002 0.013 3.235 3.252 dbcsr_special_finalize 555 9.7 0.006 0.007 2.840 3.206 dbcsr_merge_single_wm 555 10.7 0.537 0.667 2.832 3.197 ls_scf_dm_to_ks 2 5.0 0.007 0.077 3.031 3.091 ls_scf_store_result 1 5.0 0.000 0.000 2.978 3.038 dbcsr_data_release 10477 10.7 1.605 2.421 1.605 2.421 qs_energies_init_hamiltonians 1 3.0 0.144 0.196 2.065 2.066 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=98.939000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2688.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ef71e334ee8f071d2456394b624a331795da3ecc_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 4.683936E+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_SendRecv 45 115200. MP_ISendRecv 45 115200. MP_Wait 528 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.289 0.497 99.619 99.620 qs_energies 1 2.0 0.000 0.000 97.918 97.921 ls_scf 1 3.0 0.000 0.000 95.907 95.911 dbcsr_multiply_generic 111 6.7 0.017 0.018 77.403 77.626 ls_scf_main 1 4.0 0.000 0.000 61.690 61.691 multiply_cannon 111 7.7 0.082 0.153 54.966 59.951 density_matrix_trs4 2 5.0 0.002 0.003 54.512 54.601 multiply_cannon_loop 111 8.7 0.069 0.077 50.389 52.061 mp_waitall_1 5436 11.0 26.104 32.381 26.104 32.381 ls_scf_init_scf 1 4.0 0.000 0.001 30.555 30.556 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.296 29.338 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.145 27.157 multiply_cannon_multrec 444 9.7 13.959 16.277 20.965 23.309 make_m2s 222 7.7 0.004 0.005 17.629 20.196 make_images 222 8.7 3.712 4.415 17.566 20.135 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.201 16.249 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.855 16.060 make_images_data 222 9.7 0.003 0.004 9.846 12.377 hybrid_alltoall_any 227 10.6 0.786 3.761 9.635 12.190 dbcsr_mm_accdrv_process 3003 10.4 0.177 0.343 6.706 7.830 multiply_cannon_sync_h2d 444 9.7 6.505 7.813 6.505 7.813 dbcsr_mm_accdrv_process_sort 3003 11.4 6.391 7.490 6.391 7.490 mp_allgather_i34 111 8.7 2.812 6.965 2.812 6.965 arnoldi_extremal 4 6.8 0.000 0.000 5.752 5.763 arnoldi_normal_ev 4 7.8 0.002 0.005 5.752 5.763 build_subspace 16 8.4 0.015 0.020 5.341 5.354 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.649 4.801 mp_sum_l 807 5.4 2.896 4.676 2.896 4.676 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.127 4.287 dbcsr_matrix_vector_mult_local 304 10.0 3.646 4.091 3.648 4.093 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.640 3.862 mp_irecv_dv 1241 11.2 1.620 3.840 1.620 3.840 calculate_norms 792 9.8 3.590 3.757 3.590 3.757 ls_scf_post 1 4.0 0.000 0.000 3.662 3.664 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.519 3.610 make_images_sizes 222 9.7 0.000 0.000 1.153 3.512 mp_alltoall_i44 222 10.7 1.153 3.511 1.153 3.511 ls_scf_store_result 1 5.0 0.000 0.000 3.377 3.419 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.929 3.377 dbcsr_finalize 304 7.8 0.062 0.078 2.192 2.271 dbcsr_merge_all 275 8.9 0.472 0.519 2.042 2.096 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.061 2.063 rebuild_ks_matrix 3 7.3 0.000 0.000 2.029 2.031 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.029 2.031 qs_energies_init_hamiltonians 1 3.0 0.002 0.002 1.995 1.995 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=99.620000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3726.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ef71e334ee8f071d2456394b624a331795da3ecc_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 8.786743E+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_SendRecv 42 732600. MP_ISendRecv 42 732600. MP_Wait 267 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.079 0.101 108.788 108.789 qs_energies 1 2.0 0.000 0.000 107.311 107.331 ls_scf 1 3.0 0.000 0.000 103.893 103.913 dbcsr_multiply_generic 111 6.7 0.024 0.026 77.560 77.652 ls_scf_main 1 4.0 0.000 0.000 65.802 65.803 density_matrix_trs4 2 5.0 0.002 0.003 56.602 56.654 multiply_cannon 111 7.7 0.167 0.243 49.967 51.625 multiply_cannon_loop 111 8.7 0.067 0.070 46.632 47.575 ls_scf_init_scf 1 4.0 0.000 0.001 34.421 34.422 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.848 32.865 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 30.002 30.013 mp_waitall_1 4527 11.1 22.380 25.840 22.380 25.840 make_m2s 222 7.7 0.005 0.005 24.026 24.933 make_images 222 8.7 4.582 4.998 23.920 24.825 multiply_cannon_multrec 444 9.7 17.914 18.639 22.519 23.179 hybrid_alltoall_any 227 10.6 1.662 3.622 13.062 15.931 make_images_data 222 9.7 0.003 0.003 13.281 15.612 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.509 11.131 multiply_cannon_sync_h2d 444 9.7 8.815 8.868 8.815 8.868 arnoldi_extremal 4 6.8 0.000 0.000 7.263 7.275 arnoldi_normal_ev 4 7.8 0.003 0.009 7.263 7.275 build_subspace 16 8.4 0.026 0.036 6.702 6.711 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.353 5.479 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.032 5.283 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.174 5.264 dbcsr_matrix_vector_mult_local 304 10.0 4.930 5.187 4.933 5.190 dbcsr_mm_accdrv_process 1814 10.4 0.246 0.382 4.439 4.572 dbcsr_mm_accdrv_process_sort 1814 11.4 4.108 4.234 4.108 4.234 ls_scf_post 1 4.0 0.000 0.000 3.670 3.689 make_images_sizes 222 9.7 0.000 0.000 1.418 3.472 mp_alltoall_i44 222 10.7 1.418 3.472 1.418 3.472 ls_scf_store_result 1 5.0 0.000 0.000 3.418 3.428 qs_energies_init_hamiltonians 1 3.0 0.014 0.029 3.387 3.387 calculate_norms 792 9.8 3.272 3.320 3.272 3.320 dbcsr_finalize 304 7.8 0.082 0.089 3.082 3.135 mp_allgather_i34 111 8.7 1.000 3.104 1.000 3.104 dbcsr_merge_all 275 8.9 0.890 0.914 2.868 2.915 dbcsr_complete_redistribute 5 7.6 1.430 1.468 2.769 2.906 dbcsr_data_release 12724 10.6 2.320 2.834 2.320 2.834 matrix_ls_to_qs 2 6.0 0.000 0.000 2.433 2.583 dbcsr_sort_data 325 11.1 2.443 2.515 2.443 2.515 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.490 2.492 rebuild_ks_matrix 3 7.3 0.000 0.000 2.424 2.426 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.424 2.426 dbcsr_new_transposed 4 7.5 0.243 0.251 2.316 2.347 dbcsr_frobenius_norm 74 6.6 2.058 2.140 2.203 2.235 dbcsr_add_d 103 6.2 0.000 0.000 2.133 2.207 dbcsr_add_anytype 103 7.2 0.860 0.893 2.133 2.206 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=108.789000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6962.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ef71e334ee8f071d2456394b624a331795da3ecc_performance_tests/27/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 7009386627072 0.0% 0.0% 100.0% flops 9 x 9 x 32 7335108845568 0.0% 0.0% 100.0% flops 9 x 22 x 32 9866241589248 0.0% 0.0% 100.0% flops 22 x 9 x 32 9884108906496 0.0% 0.0% 100.0% flops 22 x 22 x 32 13354440523776 0.0% 0.0% 100.0% flops 32 x 32 x 9 20607185977344 0.0% 0.0% 100.0% flops 32 x 32 x 22 25186560638976 0.0% 0.0% 100.0% flops 9 x 32 x 32 28458319085568 0.0% 0.0% 100.0% flops 22 x 32 x 32 34782389993472 0.0% 0.0% 100.0% flops 9 x 32 x 9 42881542373376 0.0% 0.0% 100.0% flops 22 x 32 x 9 55680402235392 0.0% 0.0% 100.0% flops 9 x 32 x 22 55680402235392 0.0% 0.0% 100.0% flops 22 x 32 x 22 72328573419520 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 383.054662E+12 0.0% 0.0% 100.0% flops max/rank 733.641090E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 26899403712 0.0% 0.0% 100.0% number of processed stacks 118860288 0.0% 0.0% 100.0% average stack size 0.0 0.0 226.3 marketing flops 780.439111E+12 ------------------------------------------------------------------------------- # multiplications 1445 max memory usage/rank 590.913536E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 102097920 MPI messages size (bytes): total size 37.227590E+12 min size 0.000000E+00 max size 4.551360E+06 average size 364.626312E+03 MPI breakdown and total messages size (bytes): size <= 128 731472 0 128 < size <= 8192 11922720 97670922240 8192 < size <= 32768 24718992 614677610496 32768 < size <= 131072 20000256 1970081366016 131072 < size <= 4194304 42515668 24886801223040 4194304 < size <= 16777216 2208812 9656099886720 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4640 76559. MP_Allreduce 13152 2094. MP_Sync 1064 MP_Alltoall 2588 772243637. MP_SendRecv 168740 11136. MP_ISendRecv 92040 11136. MP_Wait 102830 MP_comm_split 40 MP_ISend 26090 85106. MP_IRecv 37890 59644. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.062 0.116 232.032 232.077 qs_mol_dyn_low 1 2.0 0.005 0.027 230.888 230.928 qs_forces 5 3.8 0.006 0.057 230.715 230.760 qs_energies 5 4.8 0.003 0.038 227.085 227.127 scf_env_do_scf 5 5.8 0.000 0.003 212.755 212.796 scf_env_do_scf_inner_loop 105 6.6 0.007 0.023 185.624 185.661 qs_scf_new_mos 105 7.6 0.000 0.001 144.026 144.206 qs_scf_loop_do_ot 105 8.6 0.001 0.001 144.026 144.206 ot_scf_mini 105 9.6 0.003 0.005 134.255 134.426 dbcsr_multiply_generic 1445 12.2 0.125 0.134 132.620 133.104 multiply_cannon 1445 13.2 0.276 0.288 113.195 115.423 multiply_cannon_loop 1445 14.2 2.369 2.497 111.403 112.908 velocity_verlet 4 3.0 0.004 0.014 107.438 107.461 ot_mini 105 10.6 0.004 0.204 59.604 59.845 multiply_cannon_multrec 69360 15.2 31.699 37.473 39.791 45.393 qs_ot_get_p 112 10.4 0.001 0.001 44.499 44.878 mp_waitall_1 488190 16.1 34.328 42.823 34.328 42.823 qs_ot_get_derivative 55 11.6 0.001 0.001 37.949 38.080 multiply_cannon_sync_h2d 69360 15.2 30.732 34.435 30.732 34.435 multiply_cannon_metrocomm3 69360 15.2 0.191 0.201 25.073 33.733 qs_ot_p2m_diag 40 11.0 0.020 0.031 33.371 33.446 rebuild_ks_matrix 110 8.4 0.000 0.000 30.201 30.385 qs_ks_build_kohn_sham_matrix 110 9.4 0.012 0.021 30.200 30.385 cp_dbcsr_syevd 40 12.0 0.002 0.003 30.176 30.182 qs_ks_update_qs_env 112 7.6 0.001 0.001 27.275 27.450 init_scf_loop 7 6.6 0.000 0.005 27.101 27.106 cp_fm_syevd 40 13.0 0.000 0.002 24.996 25.156 apply_preconditioner_dbcsr 62 12.6 0.000 0.001 22.917 23.203 apply_single 62 13.6 0.000 0.000 22.917 23.203 prepare_preconditioner 7 7.6 0.000 0.000 22.143 22.172 make_preconditioner 7 8.6 0.009 0.188 22.143 22.172 ot_new_cg_direction 55 11.6 0.001 0.002 20.877 20.881 cp_fm_redistribute_end 40 14.0 10.118 20.189 10.124 20.190 cp_fm_syevd_base 40 14.0 10.058 20.129 10.058 20.129 qs_rho_update_rho_low 110 7.6 0.001 0.001 17.850 18.243 calculate_rho_elec 110 8.6 0.030 0.033 17.849 18.242 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.800 15.016 make_full_inverse_cholesky 7 9.6 0.000 0.001 14.646 14.712 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 13.804 13.912 mp_sum_l 4684 12.4 12.476 13.320 12.476 13.320 pw_transfer 1645 12.4 0.083 0.105 13.074 13.271 rs_pw_transfer 690 11.5 0.011 0.013 11.914 13.225 density_rs2pw 110 9.6 0.006 0.011 11.825 13.206 fft_wrap_pw1pw2 1425 13.5 0.013 0.016 12.932 13.132 calculate_dm_sparse 110 9.5 0.000 0.001 11.949 12.079 init_scf_run 5 5.8 0.000 0.001 11.578 11.580 scf_env_initial_rho_setup 5 6.8 0.000 0.003 11.577 11.580 fft_wrap_pw1pw2_240 915 15.0 1.189 1.273 11.123 11.304 qs_vxc_create 110 10.4 0.002 0.011 10.763 10.804 cp_fm_cholesky_invert 7 10.6 10.070 10.078 10.070 10.078 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 9.971 10.038 check_diag 80 13.5 8.590 8.866 9.575 9.718 sum_up_and_integrate 60 10.3 0.028 0.038 9.420 9.450 integrate_v_rspace 60 11.3 0.002 0.002 9.392 9.421 fft3d_pb 915 16.0 2.385 2.621 9.186 9.419 dbcsr_mm_accdrv_process 154766 15.8 4.005 4.176 7.961 8.936 xc_rho_set_and_dset_create 110 12.4 0.076 0.097 7.474 7.717 multiply_cannon_metrocomm1 69360 15.2 0.091 0.097 4.631 7.247 make_full_single_inverse 7 9.6 0.001 0.009 7.212 7.240 xc_vxc_pw_create 60 11.3 0.039 0.050 7.193 7.235 make_m2s 2890 13.2 0.076 0.085 6.498 7.103 make_images 2890 14.2 0.241 0.259 6.394 6.999 calculate_first_density_matrix 1 7.0 0.001 0.004 6.712 6.727 mp_alltoall_z22v 2340 17.7 6.269 6.559 6.269 6.559 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.002 6.542 6.557 xc_pw_derive 510 13.4 0.005 0.006 6.474 6.531 acc_transpose_blocks 69360 15.2 0.361 0.378 6.004 6.495 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 6.082 6.162 mp_waitany 7680 13.5 4.437 5.752 4.437 5.752 potential_pw2rs 60 12.3 0.003 0.003 5.528 5.567 multiply_cannon_metrocomm4 67915 15.2 0.184 0.204 2.073 5.176 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=232.077000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=560.000000, yerr=3.633180 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: ef71e334ee8f071d2456394b624a331795da3ecc Summary: empty Status: OK