=== 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: 09e2302ee41b295db331f66d9974babba7191ee1 ################# 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/09e2302ee41b295db331f66d9974babba7191ee1_performance_tests/01 job id: 45405522 --- 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/09e2302ee41b295db331f66d9974babba7191ee1_performance_tests/02 job id: 45405524 --- 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/09e2302ee41b295db331f66d9974babba7191ee1_performance_tests/03 job id: 45405525 --- 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/09e2302ee41b295db331f66d9974babba7191ee1_performance_tests/04 job id: 45405526 --- 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/09e2302ee41b295db331f66d9974babba7191ee1_performance_tests/05 job id: 45405527 --- 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/09e2302ee41b295db331f66d9974babba7191ee1_performance_tests/06 job id: 45405528 --- 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/09e2302ee41b295db331f66d9974babba7191ee1_performance_tests/07 job id: 45405530 --- 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/09e2302ee41b295db331f66d9974babba7191ee1_performance_tests/08 job id: 45405531 --- 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/09e2302ee41b295db331f66d9974babba7191ee1_performance_tests/09 job id: 45405533 --- 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/09e2302ee41b295db331f66d9974babba7191ee1_performance_tests/10 job id: 45405535 --- 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/09e2302ee41b295db331f66d9974babba7191ee1_performance_tests/11 job id: 45405536 --- 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/09e2302ee41b295db331f66d9974babba7191ee1_performance_tests/12 job id: 45405537 --- 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/09e2302ee41b295db331f66d9974babba7191ee1_performance_tests/13 job id: 45405539 --- 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/09e2302ee41b295db331f66d9974babba7191ee1_performance_tests/14 job id: 45405540 --- 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/09e2302ee41b295db331f66d9974babba7191ee1_performance_tests/15 job id: 45405542 --- 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/09e2302ee41b295db331f66d9974babba7191ee1_performance_tests/16 job id: 45405543 --- 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/09e2302ee41b295db331f66d9974babba7191ee1_performance_tests/17 job id: 45405544 --- 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/09e2302ee41b295db331f66d9974babba7191ee1_performance_tests/18 job id: 45405546 --- 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/09e2302ee41b295db331f66d9974babba7191ee1_performance_tests/19 job id: 45405547 --- 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/09e2302ee41b295db331f66d9974babba7191ee1_performance_tests/20 job id: 45405548 --- 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/09e2302ee41b295db331f66d9974babba7191ee1_performance_tests/21 job id: 45405550 --- 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/09e2302ee41b295db331f66d9974babba7191ee1_performance_tests/22 job id: 45405551 --- 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/09e2302ee41b295db331f66d9974babba7191ee1_performance_tests/23 job id: 45405553 --- 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/09e2302ee41b295db331f66d9974babba7191ee1_performance_tests/24 job id: 45405554 --- 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/09e2302ee41b295db331f66d9974babba7191ee1_performance_tests/25 job id: 45405555 --- 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/09e2302ee41b295db331f66d9974babba7191ee1_performance_tests/26 job id: 45405558 --- 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/09e2302ee41b295db331f66d9974babba7191ee1_performance_tests/27 job id: 45405559 --- 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/09e2302ee41b295db331f66d9974babba7191ee1_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 0.022 0.037 141.314 141.315 farming_run 1 2.0 140.766 140.768 141.274 141.277 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.460834E+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 0.012 0.027 118.123 118.124 qs_energies 1 2.0 0.000 0.000 117.869 117.872 mp2_main 1 3.0 0.000 0.000 115.255 115.258 mp2_gpw_main 1 4.0 0.019 0.026 113.839 113.841 mp2_ri_gpw_compute_in 1 5.0 0.173 0.176 93.989 94.602 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 55.489 56.101 mp2_eri_3c_integrate_gpw 272 7.0 0.154 0.172 41.810 47.371 get_2c_integrals 1 6.0 0.000 0.000 37.469 38.326 integrate_v_rspace 273 8.0 0.435 0.447 25.182 30.392 pw_transfer 6555 10.6 0.373 0.391 27.523 28.054 fft_wrap_pw1pw2 5465 11.4 0.045 0.047 26.175 26.690 grid_integrate_task_list 273 9.0 20.985 26.676 20.985 26.676 fft_wrap_pw1pw2_100 2178 12.4 1.221 1.447 23.645 24.155 rpa_ri_compute_en 1 5.0 0.001 0.001 19.733 19.906 compute_2c_integrals 1 7.0 0.003 0.003 19.875 19.879 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.964 19.439 mp2_eri_2c_integrate_gpw 1 9.0 2.375 2.436 18.961 19.435 cp_fm_cholesky_decompose 12 8.2 17.792 18.751 17.792 18.751 cholesky_decomp 1 7.0 0.000 0.000 16.452 17.320 fft3d_s 5443 13.4 16.176 16.677 16.197 16.698 ao_to_mo_and_store_B_mult_1 272 7.0 10.845 15.571 10.845 15.571 calculate_wavefunction 272 8.0 5.420 5.581 12.553 13.136 rpa_num_int 1 6.0 0.002 0.024 11.173 11.173 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.870 11.147 calc_mat_Q 8 8.0 0.000 0.000 9.550 9.824 contract_S_to_Q 8 9.0 0.000 0.000 8.973 9.244 parallel_gemm_fm 14 9.1 0.000 0.000 8.544 8.839 parallel_gemm_fm_cosma 14 10.1 8.544 8.839 8.544 8.839 calc_potential_gpw 544 9.5 0.006 0.006 8.314 8.700 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.269 8.595 potential_pw2rs 545 10.0 0.106 0.108 7.693 8.341 create_integ_mat 1 6.0 0.021 0.028 7.852 7.852 collocate_single_gaussian 272 10.0 0.039 0.042 7.497 7.743 array2fm 1 7.0 0.000 0.000 6.519 6.940 pw_scatter_s 2720 13.7 4.440 4.652 4.440 4.652 pw_gather_s 2722 13.2 3.877 4.271 3.877 4.271 array2fm_buffer_send 1 8.0 2.898 3.073 2.898 3.073 pw_poisson_solve 545 10.5 1.134 1.172 2.189 2.419 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=113.841485, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2733.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/09e2302ee41b295db331f66d9974babba7191ee1_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 0.028 0.037 405.031 405.033 farming_run 1 2.0 404.252 404.255 404.988 404.990 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827141120 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788822 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.226527E+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 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 0.015 0.047 212.587 212.588 qs_energies 1 2.0 0.000 0.000 212.343 212.349 scf_env_do_scf 1 3.0 0.000 0.000 107.761 107.761 qs_ks_update_qs_env 5 5.0 0.000 0.000 106.554 106.562 rebuild_ks_matrix 4 6.0 0.000 0.000 106.552 106.561 qs_ks_build_kohn_sham_matrix 4 7.0 0.060 0.069 106.552 106.561 hfx_ks_matrix 4 8.0 0.001 0.001 106.113 106.117 integrate_four_center 4 9.0 0.144 0.458 106.112 106.116 mp2_main 1 3.0 0.000 0.000 104.240 104.246 mp2_gpw_main 1 4.0 0.032 0.046 102.901 102.909 integrate_four_center_main 4 10.0 0.107 0.568 97.007 100.647 integrate_four_center_bin 264 11.0 96.900 100.632 96.900 100.632 init_scf_loop 1 4.0 0.000 0.000 93.348 93.349 mp2_ri_gpw_compute_in 1 5.0 0.064 0.066 75.202 76.176 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.003 54.454 55.425 mp2_eri_3c_integrate_gpw 91 7.0 0.145 0.161 42.371 47.176 integrate_v_rspace 95 8.0 0.399 0.566 28.792 33.489 pw_transfer 2240 10.6 0.145 0.178 29.970 30.414 ao_to_mo_and_store_B_mult_1 91 7.0 10.398 29.645 10.398 29.645 fft_wrap_pw1pw2 1868 11.4 0.018 0.022 28.967 29.467 mp2_ri_gpw_compute_en 1 5.0 0.056 0.063 27.526 29.293 grid_integrate_task_list 95 9.0 24.036 28.920 24.036 28.920 fft_wrap_pw1pw2_100 730 12.4 1.294 1.480 26.644 27.165 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.838 1.897 25.914 25.924 get_2c_integrals 1 6.0 0.000 0.000 20.527 20.678 compute_2c_integrals 1 7.0 0.003 0.004 19.487 19.495 compute_2c_integrals_loop_lm 1 8.0 0.001 0.002 18.918 19.303 mp2_eri_2c_integrate_gpw 1 9.0 1.730 1.907 18.917 19.301 fft3d_s 1823 13.4 18.414 18.927 18.427 18.940 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.410 14.411 calculate_wavefunction 91 8.0 2.018 2.054 9.742 9.974 mp2_ri_gpw_compute_en_comm 22 7.0 0.494 0.510 8.942 9.441 mp2_ri_gpw_compute_en_expansio 172 7.0 0.554 0.585 8.747 9.318 potential_pw2rs 186 10.0 0.033 0.035 8.688 9.271 local_gemm 172 8.0 8.192 8.733 8.192 8.733 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.277 8.631 calc_potential_gpw 182 9.5 0.002 0.002 7.940 8.184 collocate_single_gaussian 91 10.0 0.017 0.024 7.904 8.160 mp_sendrecv_dm3 2068 8.0 6.986 7.472 6.986 7.472 mp_sync 37 10.5 3.840 7.350 3.840 7.350 mp2_ri_gpw_compute_en_ener 172 7.0 6.343 6.429 6.343 6.429 pw_gather_s 912 13.2 4.910 5.436 4.910 5.436 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=102.896844, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1514.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/09e2302ee41b295db331f66d9974babba7191ee1_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 452.280320E+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 589622. 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 0.019 0.048 57.157 57.161 qs_mol_dyn_low 1 2.0 0.003 0.006 56.714 56.735 qs_forces 11 3.9 0.006 0.028 56.563 56.564 qs_energies 11 4.9 0.047 0.064 54.987 55.005 scf_env_do_scf 11 5.9 0.001 0.002 48.295 48.296 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 46.130 46.131 qs_scf_new_mos 108 7.5 0.000 0.001 35.436 35.721 qs_scf_loop_do_ot 108 8.5 0.000 0.001 35.435 35.721 dbcsr_multiply_generic 2286 12.5 0.092 0.096 34.873 35.298 ot_scf_mini 108 9.5 0.002 0.002 33.710 33.879 multiply_cannon 2286 13.5 0.187 0.196 26.628 28.142 multiply_cannon_loop 2286 14.5 1.471 1.560 25.748 27.301 velocity_verlet 10 3.0 0.002 0.007 27.231 27.232 ot_mini 108 10.5 0.001 0.001 20.165 20.424 qs_ot_get_derivative 108 11.5 0.001 0.002 17.182 17.382 mp_waitall_1 245248 16.5 9.197 15.516 9.197 15.516 multiply_cannon_metrocomm3 54864 15.5 0.068 0.074 6.251 13.352 multiply_cannon_multrec 54864 15.5 4.208 6.557 7.605 11.202 qs_ot_get_p 119 10.4 0.001 0.001 8.751 9.018 rebuild_ks_matrix 119 8.3 0.000 0.000 8.145 8.300 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.013 8.145 8.300 mp_sum_l 7207 12.9 5.874 7.670 5.874 7.670 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.148 7.288 multiply_cannon_sync_h2d 54864 15.5 5.895 7.009 5.895 7.009 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 6.023 6.138 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.672 6.128 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.527 5.576 init_scf_run 11 5.9 0.004 0.029 5.201 5.204 scf_env_initial_rho_setup 11 6.9 0.001 0.008 5.197 5.204 dbcsr_mm_accdrv_process 76910 16.1 1.158 1.850 3.319 4.793 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.593 4.770 calculate_rho_elec 119 8.7 0.012 0.017 4.592 4.770 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.705 4.706 sum_up_and_integrate 119 10.3 0.012 0.015 4.684 4.695 integrate_v_rspace 119 11.3 0.002 0.004 4.672 4.684 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.398 4.402 cp_fm_redistribute_end 50 14.0 2.251 4.375 2.257 4.379 cp_fm_diag_elpa_base 50 14.0 2.113 4.249 2.117 4.256 multiply_cannon_metrocomm1 54864 15.5 0.052 0.057 2.084 3.266 rs_pw_transfer 974 11.9 0.012 0.013 3.071 3.174 calculate_dm_sparse 119 9.5 0.000 0.001 2.945 3.091 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.866 3.056 apply_single 119 13.6 0.000 0.000 2.866 3.056 jit_kernel_multiply 13 15.8 2.096 2.853 2.096 2.853 calculate_first_density_matrix 1 7.0 0.006 0.011 2.816 2.827 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.748 2.752 ot_diis_step 108 11.5 0.006 0.006 2.680 2.680 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.584 2.673 density_rs2pw 119 9.7 0.004 0.005 2.510 2.606 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.339 2.377 wfi_extrapolate 11 7.9 0.001 0.001 2.319 2.319 acc_transpose_blocks 54864 15.5 0.228 0.266 1.733 2.206 make_m2s 4572 13.5 0.054 0.056 2.108 2.197 grid_integrate_task_list 119 12.3 2.031 2.164 2.031 2.164 init_scf_loop 11 6.9 0.001 0.005 2.146 2.148 make_images 4572 14.5 0.133 0.139 2.026 2.113 mp_sum_d 4135 12.0 1.449 2.073 1.449 2.073 potential_pw2rs 119 12.3 0.004 0.005 1.999 2.023 pw_transfer 1439 11.6 0.053 0.059 1.936 2.018 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.859 1.946 mp_alltoall_d11v 2130 13.8 1.578 1.824 1.578 1.824 fft3d_ps 1201 14.6 0.370 0.473 1.630 1.715 fft_wrap_pw1pw2_140 487 13.2 0.082 0.097 1.447 1.535 mp_waitany 12084 13.8 1.351 1.460 1.351 1.460 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.387 1.406 grid_collocate_task_list 119 9.7 1.290 1.352 1.290 1.352 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=57.161000, 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/09e2302ee41b295db331f66d9974babba7191ee1_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 488.402944E+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 1030452. 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 0.024 0.053 41.262 41.264 qs_mol_dyn_low 1 2.0 0.003 0.004 40.944 40.954 qs_forces 11 3.9 0.002 0.003 40.760 40.762 qs_energies 11 4.9 0.002 0.005 39.047 39.051 scf_env_do_scf 11 5.9 0.001 0.003 33.929 33.930 scf_env_do_scf_inner_loop 108 6.5 0.002 0.008 31.311 31.311 dbcsr_multiply_generic 2286 12.5 0.097 0.100 22.119 22.483 qs_scf_new_mos 108 7.5 0.001 0.001 21.399 21.653 qs_scf_loop_do_ot 108 8.5 0.001 0.001 21.399 21.652 ot_scf_mini 108 9.5 0.002 0.003 20.517 20.694 velocity_verlet 10 3.0 0.002 0.002 19.492 19.496 multiply_cannon 2286 13.5 0.209 0.220 16.638 18.143 multiply_cannon_loop 2286 14.5 0.896 0.974 15.374 16.735 ot_mini 108 10.5 0.001 0.001 12.570 12.809 mp_waitall_1 200699 16.5 6.137 11.294 6.137 11.294 qs_ot_get_derivative 108 11.5 0.001 0.002 10.054 10.230 multiply_cannon_metrocomm3 27432 15.5 0.066 0.070 4.251 9.771 multiply_cannon_multrec 27432 15.5 1.979 4.485 5.892 8.644 rebuild_ks_matrix 119 8.3 0.000 0.000 7.473 7.613 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.472 7.613 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.615 6.743 dbcsr_mm_accdrv_process 47894 16.0 3.038 5.415 3.842 5.710 qs_ot_get_p 119 10.4 0.001 0.001 4.924 5.168 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.779 4.621 sum_up_and_integrate 119 10.3 0.024 0.027 4.432 4.437 integrate_v_rspace 119 11.3 0.002 0.003 4.407 4.414 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.284 4.320 calculate_rho_elec 119 8.7 0.021 0.024 4.283 4.320 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.080 4.169 apply_single 119 13.6 0.000 0.000 3.080 4.169 mp_sum_l 7207 12.9 2.127 4.096 2.127 4.096 init_scf_run 11 5.9 0.000 0.001 3.831 3.831 scf_env_initial_rho_setup 11 6.9 0.000 0.001 3.830 3.831 qs_ot_p2m_diag 50 11.0 0.009 0.013 3.403 3.422 rs_pw_transfer 974 11.9 0.010 0.011 2.925 3.359 make_m2s 4572 13.5 0.052 0.053 2.910 3.158 make_images 4572 14.5 0.200 0.236 2.824 3.069 density_rs2pw 119 9.7 0.004 0.004 2.603 3.051 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.960 2.961 multiply_cannon_sync_h2d 27432 15.5 2.188 2.829 2.188 2.829 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.510 2.606 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.599 2.600 cp_fm_redistribute_end 50 14.0 1.314 2.565 1.319 2.570 init_scf_loop 11 6.9 0.000 0.002 2.547 2.548 cp_fm_diag_elpa_base 50 14.0 1.215 2.457 1.244 2.504 ot_diis_step 108 11.5 0.011 0.011 2.463 2.464 pw_transfer 1439 11.6 0.065 0.072 2.382 2.415 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.290 2.325 calculate_first_density_matrix 1 7.0 0.001 0.004 2.320 2.322 potential_pw2rs 119 12.3 0.006 0.006 2.139 2.150 calculate_dm_sparse 119 9.5 0.000 0.001 2.017 2.092 fft3d_ps 1201 14.6 0.515 0.568 1.990 2.023 grid_integrate_task_list 119 12.3 1.843 1.969 1.843 1.969 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.885 1.888 make_images_data 4572 15.5 0.044 0.050 1.295 1.815 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.695 1.735 hybrid_alltoall_any 4725 16.4 0.051 0.112 1.125 1.733 fft_wrap_pw1pw2_140 487 13.2 0.080 0.088 1.625 1.661 prepare_preconditioner 11 7.9 0.000 0.001 1.620 1.645 make_preconditioner 11 8.9 0.000 0.000 1.620 1.645 jit_kernel_multiply 8 16.0 0.751 1.641 0.751 1.641 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.519 1.574 mp_allgather_i34 2286 14.5 0.705 1.521 0.705 1.521 acc_transpose_blocks 27432 15.5 0.108 0.113 1.200 1.512 wfi_extrapolate 11 7.9 0.001 0.001 1.452 1.452 mp_alltoall_d11v 2130 13.8 1.264 1.392 1.264 1.392 grid_collocate_task_list 119 9.7 1.231 1.361 1.231 1.361 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.244 1.252 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.177 1.225 make_images_sizes 4572 15.5 0.005 0.005 0.800 1.123 mp_alltoall_i44 4572 16.5 0.795 1.118 0.795 1.118 mp_sum_d 4135 12.0 0.649 1.112 0.649 1.112 mp_alltoall_z22v 1201 16.6 0.960 1.087 0.960 1.087 mp_waitany 5720 13.7 0.574 1.048 0.574 1.048 rs_pw_transfer_RS2PW_140 130 11.5 0.144 0.189 0.601 1.022 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 1.017 1.018 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.918 0.932 rs_pw_transfer_PW2RS_50 119 14.3 0.589 0.608 0.832 0.880 acc_transpose_blocks_kernels 27432 16.5 0.181 0.270 0.663 0.879 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=41.264000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=464.818182, yerr=1.402477 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/09e2302ee41b295db331f66d9974babba7191ee1_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 524.480512E+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 1607811. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.025 0.080 34.295 34.297 qs_mol_dyn_low 1 2.0 0.003 0.005 33.939 33.948 qs_forces 11 3.9 0.002 0.003 33.597 33.598 qs_energies 11 4.9 0.010 0.072 32.044 32.046 scf_env_do_scf 11 5.9 0.001 0.021 26.983 26.984 scf_env_do_scf_inner_loop 108 6.5 0.002 0.008 23.824 23.824 velocity_verlet 10 3.0 0.006 0.014 16.768 16.778 dbcsr_multiply_generic 2286 12.5 0.094 0.099 16.585 16.673 qs_scf_new_mos 108 7.5 0.001 0.001 15.297 15.310 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.297 15.309 ot_scf_mini 108 9.5 0.002 0.003 14.576 14.590 multiply_cannon 2286 13.5 0.198 0.214 13.154 13.887 multiply_cannon_loop 2286 14.5 0.633 0.669 12.330 13.094 ot_mini 108 10.5 0.001 0.001 8.985 9.003 qs_ot_get_derivative 108 11.5 0.001 0.002 7.487 7.501 multiply_cannon_multrec 18288 15.5 1.924 2.929 6.765 7.115 rebuild_ks_matrix 119 8.3 0.000 0.000 6.797 6.812 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.015 6.796 6.812 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.052 6.067 dbcsr_mm_accdrv_process 38222 16.0 4.062 5.410 4.756 5.560 sum_up_and_integrate 119 10.3 0.030 0.031 4.312 4.318 integrate_v_rspace 119 11.3 0.003 0.003 4.281 4.291 mp_waitall_1 158411 16.6 3.015 4.146 3.015 4.146 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.730 3.740 calculate_rho_elec 119 8.7 0.031 0.032 3.730 3.740 init_scf_run 11 5.9 0.000 0.001 3.694 3.695 scf_env_initial_rho_setup 11 6.9 0.000 0.001 3.694 3.694 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.940 3.547 qs_ot_get_p 119 10.4 0.001 0.001 3.409 3.429 init_scf_loop 11 6.9 0.001 0.005 3.121 3.123 rs_pw_transfer 974 11.9 0.009 0.010 2.682 2.908 density_rs2pw 119 9.7 0.004 0.005 2.242 2.486 multiply_cannon_metrocomm3 18288 15.5 0.044 0.045 1.432 2.410 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.019 2.353 apply_single 119 13.6 0.000 0.000 2.019 2.353 qs_ot_p2m_diag 50 11.0 0.012 0.012 2.287 2.292 calculate_first_density_matrix 1 7.0 0.001 0.003 2.281 2.283 pw_transfer 1439 11.6 0.066 0.072 2.154 2.171 make_m2s 4572 13.5 0.044 0.045 1.950 2.117 fft_wrap_pw1pw2 1201 12.6 0.008 0.009 2.060 2.077 potential_pw2rs 119 12.3 0.007 0.009 2.064 2.074 prepare_preconditioner 11 7.9 0.000 0.000 2.063 2.065 make_preconditioner 11 8.9 0.000 0.001 2.063 2.065 make_images 4572 14.5 0.189 0.203 1.866 2.031 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.023 2.023 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.913 1.999 grid_integrate_task_list 119 12.3 1.805 1.900 1.805 1.900 calculate_dm_sparse 119 9.5 0.000 0.001 1.806 1.820 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.763 1.774 cp_fm_diag_elpa_base 50 14.0 1.739 1.752 1.761 1.772 fft3d_ps 1201 14.6 0.528 0.543 1.739 1.757 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.731 1.733 fft_wrap_pw1pw2_140 487 13.2 0.091 0.096 1.574 1.589 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.573 1.584 multiply_cannon_sync_h2d 18288 15.5 1.370 1.556 1.370 1.556 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.542 1.549 jit_kernel_multiply 8 16.3 0.641 1.518 0.641 1.518 mp_sum_l 7207 12.9 1.140 1.497 1.140 1.497 ot_diis_step 108 11.5 0.011 0.011 1.473 1.473 grid_collocate_task_list 119 9.7 1.213 1.369 1.213 1.369 wfi_extrapolate 11 7.9 0.001 0.001 1.341 1.342 acc_transpose_blocks 18288 15.5 0.075 0.077 1.260 1.286 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.176 1.180 make_images_data 4572 15.5 0.044 0.047 0.884 1.084 qs_energies_init_hamiltonians 11 5.9 0.002 0.006 0.998 0.999 hybrid_alltoall_any 4725 16.4 0.055 0.112 0.765 0.996 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.885 0.911 mp_alltoall_d11v 2130 13.8 0.769 0.896 0.769 0.896 mp_alltoall_z22v 1201 16.6 0.790 0.893 0.790 0.893 acc_transpose_blocks_kernels 18288 16.5 0.211 0.219 0.815 0.829 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.818 0.820 mp_waitany 9880 13.7 0.533 0.785 0.533 0.785 rs_pw_transfer_RS2PW_140 130 11.5 0.119 0.128 0.523 0.753 cp_fm_cholesky_invert 11 10.9 0.733 0.737 0.733 0.737 cp_fm_cholesky_decompose 22 10.9 0.720 0.722 0.720 0.722 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.651 0.714 rs_pw_transfer_PW2RS_50 119 14.3 0.404 0.411 0.647 0.703 dbcsr_complete_redistribute 329 12.2 0.103 0.135 0.593 0.699 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=34.297000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=498.272727, yerr=3.193485 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/09e2302ee41b295db331f66d9974babba7191ee1_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 560.836608E+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 0.034 0.071 36.510 36.511 qs_mol_dyn_low 1 2.0 0.003 0.004 36.225 36.233 qs_forces 11 3.9 0.004 0.013 36.141 36.142 qs_energies 11 4.9 0.003 0.007 34.390 34.394 scf_env_do_scf 11 5.9 0.001 0.002 29.375 29.377 scf_env_do_scf_inner_loop 108 6.5 0.003 0.009 25.895 25.895 velocity_verlet 10 3.0 0.002 0.002 18.512 18.515 dbcsr_multiply_generic 2286 12.5 0.097 0.100 18.313 18.428 qs_scf_new_mos 108 7.5 0.001 0.001 16.777 16.830 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.776 16.830 ot_scf_mini 108 9.5 0.002 0.003 15.804 15.858 multiply_cannon 2286 13.5 0.229 0.262 14.435 14.976 multiply_cannon_loop 2286 14.5 0.937 0.966 13.443 13.907 ot_mini 108 10.5 0.001 0.001 9.346 9.416 multiply_cannon_multrec 27432 15.5 2.322 2.992 8.589 8.950 qs_ot_get_derivative 108 11.5 0.001 0.003 7.531 7.586 dbcsr_mm_accdrv_process 47916 15.9 5.464 7.307 6.172 7.385 rebuild_ks_matrix 119 8.3 0.000 0.000 6.894 6.946 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.015 6.894 6.946 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.146 6.191 sum_up_and_integrate 119 10.3 0.035 0.037 4.068 4.077 integrate_v_rspace 119 11.3 0.003 0.003 4.033 4.042 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.907 3.941 calculate_rho_elec 119 8.7 0.040 0.046 3.907 3.941 qs_ot_get_p 119 10.4 0.001 0.001 3.754 3.835 init_scf_run 11 5.9 0.000 0.001 3.625 3.626 scf_env_initial_rho_setup 11 6.9 0.000 0.001 3.625 3.626 init_scf_loop 11 6.9 0.001 0.007 3.459 3.460 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.720 3.205 mp_waitall_1 137007 16.6 2.053 2.645 2.053 2.645 rs_pw_transfer 974 11.9 0.009 0.010 2.419 2.612 prepare_preconditioner 11 7.9 0.000 0.001 2.599 2.607 make_preconditioner 11 8.9 0.000 0.001 2.599 2.607 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.134 2.588 apply_single 119 13.6 0.000 0.000 2.134 2.588 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.194 2.529 make_m2s 4572 13.5 0.054 0.056 2.363 2.489 density_rs2pw 119 9.7 0.004 0.004 2.296 2.468 make_images 4572 14.5 0.270 0.333 2.257 2.381 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.351 2.361 pw_transfer 1439 11.6 0.066 0.073 2.251 2.289 calculate_first_density_matrix 1 7.0 0.001 0.004 2.252 2.254 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.157 2.200 calculate_dm_sparse 119 9.5 0.000 0.001 2.060 2.117 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.011 2.013 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.955 1.984 grid_integrate_task_list 119 12.3 1.840 1.921 1.840 1.921 fft3d_ps 1201 14.6 0.567 0.619 1.831 1.869 potential_pw2rs 119 12.3 0.009 0.010 1.800 1.803 fft_wrap_pw1pw2_140 487 13.2 0.089 0.097 1.758 1.803 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.787 1.788 ot_diis_step 108 11.5 0.012 0.012 1.772 1.773 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.707 1.718 cp_fm_diag_elpa_base 50 14.0 1.670 1.689 1.705 1.716 jit_kernel_multiply 9 16.2 0.645 1.655 0.645 1.655 mp_sum_l 7207 12.9 1.110 1.652 1.110 1.652 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.617 1.629 acc_transpose_blocks 27432 15.5 0.112 0.114 1.462 1.476 grid_collocate_task_list 119 9.7 1.221 1.327 1.221 1.327 wfi_extrapolate 11 7.9 0.001 0.001 1.323 1.323 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.208 1.218 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.703 1.209 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.164 1.179 cp_fm_upper_to_full 72 14.2 0.823 1.176 0.823 1.176 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.123 1.124 make_images_data 4572 15.5 0.044 0.048 0.962 1.114 dbcsr_complete_redistribute 329 12.2 0.122 0.150 0.820 1.098 hybrid_alltoall_any 4725 16.4 0.062 0.151 0.818 1.042 multiply_cannon_sync_h2d 27432 15.5 0.983 1.042 0.983 1.042 mp_alltoall_d11v 2130 13.8 0.790 0.903 0.790 0.903 mp_alltoall_z22v 1201 16.6 0.847 0.886 0.847 0.886 copy_fm_to_dbcsr 176 11.2 0.001 0.002 0.603 0.870 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.792 0.869 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.843 0.848 acc_transpose_blocks_kernels 27432 16.5 0.268 0.276 0.829 0.838 cp_fm_cholesky_invert 11 10.9 0.766 0.769 0.766 0.769 mp_alltoall_i22 627 13.8 0.444 0.737 0.444 0.737 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=36.511000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=530.818182, yerr=5.781132 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/09e2302ee41b295db331f66d9974babba7191ee1_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 617.136128E+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 0.018 0.033 38.910 38.911 qs_mol_dyn_low 1 2.0 0.003 0.004 38.622 38.629 qs_forces 11 3.9 0.002 0.002 38.510 38.511 qs_energies 11 4.9 0.002 0.003 36.588 36.592 scf_env_do_scf 11 5.9 0.001 0.003 30.669 30.669 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 27.047 27.048 velocity_verlet 10 3.0 0.002 0.002 20.487 20.493 dbcsr_multiply_generic 2286 12.5 0.092 0.097 16.261 16.625 qs_scf_new_mos 108 7.5 0.001 0.001 15.753 15.807 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.752 15.806 ot_scf_mini 108 9.5 0.002 0.002 14.869 14.911 multiply_cannon 2286 13.5 0.234 0.244 11.685 12.248 multiply_cannon_loop 2286 14.5 0.331 0.344 10.144 11.170 ot_mini 108 10.5 0.001 0.001 8.207 8.302 rebuild_ks_matrix 119 8.3 0.000 0.000 8.012 8.134 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 8.012 8.133 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.180 7.297 qs_ot_get_derivative 108 11.5 0.003 0.012 6.612 6.660 multiply_cannon_multrec 9144 15.5 1.727 1.956 6.107 6.377 mp_waitall_1 115863 16.7 4.163 6.354 4.163 6.354 qs_rho_update_rho_low 119 7.7 0.001 0.001 5.409 5.449 calculate_rho_elec 119 8.7 0.060 0.061 5.409 5.449 sum_up_and_integrate 119 10.3 0.038 0.042 5.242 5.251 integrate_v_rspace 119 11.3 0.003 0.003 5.204 5.213 dbcsr_mm_accdrv_process 12550 15.8 3.381 4.292 4.278 4.353 init_scf_run 11 5.9 0.000 0.001 4.351 4.352 scf_env_initial_rho_setup 11 6.9 0.000 0.000 4.351 4.352 rs_pw_transfer 974 11.9 0.008 0.008 4.064 4.280 qs_ot_get_p 119 10.4 0.001 0.001 4.220 4.269 density_rs2pw 119 9.7 0.004 0.004 3.570 3.740 make_m2s 4572 13.5 0.034 0.036 3.003 3.626 init_scf_loop 11 6.9 0.000 0.000 3.586 3.587 make_images 4572 14.5 0.271 0.305 2.914 3.536 pw_transfer 1439 11.6 0.067 0.070 2.911 2.952 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.816 2.858 qs_ot_p2m_diag 50 11.0 0.022 0.023 2.803 2.818 potential_pw2rs 119 12.3 0.010 0.011 2.755 2.781 prepare_preconditioner 11 7.9 0.000 0.000 2.630 2.637 make_preconditioner 11 8.9 0.000 0.000 2.630 2.637 make_images_data 4572 15.5 0.038 0.042 1.415 2.533 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.471 2.525 fft3d_ps 1201 14.6 0.573 0.586 2.476 2.513 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.481 2.486 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.409 2.456 calculate_first_density_matrix 1 7.0 0.001 0.001 2.340 2.340 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.822 2.303 hybrid_alltoall_any 4725 16.4 0.063 0.176 1.263 2.230 fft_wrap_pw1pw2_140 487 13.2 0.086 0.091 2.166 2.209 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.057 2.066 calculate_dm_sparse 119 9.5 0.000 0.000 2.030 2.065 cp_fm_diag_elpa_base 50 14.0 2.028 2.047 2.055 2.065 grid_integrate_task_list 119 12.3 1.908 2.015 1.908 2.015 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.959 1.966 jit_kernel_multiply 10 15.7 0.859 1.956 0.859 1.956 wfi_extrapolate 11 7.9 0.001 0.001 1.943 1.943 mp_sum_l 7207 12.9 1.141 1.736 1.141 1.736 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.707 1.717 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.657 1.709 mp_allgather_i34 2286 14.5 0.862 1.622 0.862 1.622 multiply_cannon_metrocomm1 9144 15.5 0.022 0.023 0.927 1.592 cp_fm_cholesky_invert 11 10.9 1.551 1.558 1.551 1.558 ot_diis_step 108 11.5 0.013 0.013 1.517 1.517 grid_collocate_task_list 119 9.7 1.325 1.448 1.325 1.448 mp_alltoall_z22v 1201 16.6 1.353 1.431 1.353 1.431 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.338 1.407 apply_single 119 13.6 0.000 0.000 1.338 1.407 mp_alltoall_d11v 2130 13.8 1.248 1.398 1.248 1.398 mp_sendrecv_dv 5355 12.7 1.294 1.323 1.294 1.323 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.250 1.294 make_images_sizes 4572 15.5 0.005 0.005 0.913 1.276 mp_alltoall_i44 4572 16.5 0.908 1.271 0.908 1.271 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.260 1.268 qs_energies_init_hamiltonians 11 5.9 0.008 0.009 1.253 1.254 yz_to_x 606 15.1 0.260 0.272 0.980 1.021 rs_pw_transfer_RS2PW_50 119 11.7 0.202 0.210 0.898 1.016 acc_transpose_blocks 9144 15.5 0.037 0.038 0.981 0.988 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.956 0.960 mp_waitany 5200 13.7 0.726 0.956 0.726 0.956 rs_pw_transfer_PW2RS_50 119 14.3 0.217 0.226 0.821 0.955 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.868 0.922 x_to_yz 595 16.2 0.278 0.290 0.911 0.916 dbcsr_complete_redistribute 329 12.2 0.158 0.166 0.811 0.864 multiply_cannon_sync_h2d 9144 15.5 0.712 0.794 0.712 0.794 rs_pw_transfer_RS2PW_140 130 11.5 0.100 0.102 0.616 0.786 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=38.911000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=581.272727, yerr=8.812088 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/09e2302ee41b295db331f66d9974babba7191ee1_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 759.042048E+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 0.023 0.044 52.432 52.434 qs_mol_dyn_low 1 2.0 0.003 0.004 52.198 52.204 qs_forces 11 3.9 0.002 0.002 52.071 52.072 qs_energies 11 4.9 0.001 0.001 49.581 49.586 scf_env_do_scf 11 5.9 0.002 0.004 43.332 43.333 scf_env_do_scf_inner_loop 108 6.5 0.003 0.008 33.993 33.994 velocity_verlet 10 3.0 0.002 0.002 29.848 29.862 dbcsr_multiply_generic 2286 12.5 0.100 0.102 21.623 22.284 qs_scf_new_mos 108 7.5 0.001 0.001 21.371 21.486 qs_scf_loop_do_ot 108 8.5 0.001 0.001 21.370 21.486 ot_scf_mini 108 9.5 0.002 0.002 20.055 20.165 multiply_cannon 2286 13.5 0.309 0.318 15.535 17.720 multiply_cannon_loop 2286 14.5 0.348 0.354 13.824 15.934 ot_mini 108 10.5 0.001 0.001 11.649 11.850 rebuild_ks_matrix 119 8.3 0.000 0.000 9.475 9.734 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 9.474 9.734 init_scf_loop 11 6.9 0.000 0.000 9.306 9.307 qs_ot_get_derivative 108 11.5 0.001 0.002 9.013 9.128 multiply_cannon_multrec 9144 15.5 3.378 4.726 8.582 8.686 qs_ks_update_qs_env 119 7.6 0.001 0.001 8.319 8.553 prepare_preconditioner 11 7.9 0.000 0.000 8.106 8.127 make_preconditioner 11 8.9 0.000 0.000 8.106 8.127 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.601 7.967 mp_waitall_1 94719 16.7 5.085 7.433 5.085 7.433 dbcsr_mm_accdrv_process 12550 15.8 4.187 5.650 5.078 6.372 sum_up_and_integrate 119 10.3 0.065 0.067 5.672 5.675 integrate_v_rspace 119 11.3 0.003 0.003 5.606 5.610 qs_rho_update_rho_low 119 7.7 0.001 0.001 5.598 5.605 calculate_rho_elec 119 8.7 0.118 0.121 5.598 5.605 qs_ot_get_p 119 10.4 0.001 0.001 4.885 5.175 multiply_cannon_metrocomm3 9144 15.5 0.019 0.020 2.683 4.831 cp_fm_upper_to_full 72 14.2 3.204 4.620 3.204 4.620 init_scf_run 11 5.9 0.000 0.001 4.105 4.105 scf_env_initial_rho_setup 11 6.9 0.000 0.001 4.105 4.105 pw_transfer 1439 11.6 0.069 0.070 4.084 4.101 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 3.984 4.002 make_m2s 4572 13.5 0.038 0.039 3.488 3.897 make_images 4572 14.5 0.357 0.393 3.368 3.777 fft3d_ps 1201 14.6 0.602 0.612 3.608 3.624 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.110 3.562 mp_sum_l 7207 12.9 2.059 3.515 2.059 3.515 density_rs2pw 119 9.7 0.004 0.004 3.318 3.351 fft_wrap_pw1pw2_140 487 13.2 0.096 0.098 3.191 3.215 dbcsr_complete_redistribute 329 12.2 0.285 0.287 2.245 3.101 rs_pw_transfer 974 11.9 0.009 0.010 2.907 2.948 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.554 2.796 apply_single 119 13.6 0.000 0.000 2.554 2.796 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.776 2.789 potential_pw2rs 119 12.3 0.014 0.014 2.724 2.736 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.795 2.652 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.525 2.592 ot_diis_step 108 11.5 0.014 0.015 2.517 2.518 mp_alltoall_i22 627 13.8 1.546 2.505 1.546 2.505 calculate_dm_sparse 119 9.5 0.000 0.000 2.401 2.407 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.403 2.405 make_images_data 4572 15.5 0.043 0.045 1.696 2.367 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.499 2.336 hybrid_alltoall_any 4725 16.4 0.087 0.148 1.605 2.313 cp_fm_cholesky_invert 11 10.9 2.187 2.190 2.187 2.190 grid_integrate_task_list 119 12.3 2.156 2.183 2.156 2.183 calculate_first_density_matrix 1 7.0 0.001 0.001 2.175 2.178 mp_alltoall_z22v 1201 16.6 2.057 2.119 2.057 2.119 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.982 1.983 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.973 1.974 cp_fm_diag_elpa_base 50 14.0 1.812 1.866 1.971 1.971 wfi_extrapolate 11 7.9 0.001 0.001 1.845 1.845 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.823 1.825 mp_alltoall_d11v 2130 13.8 1.700 1.762 1.700 1.762 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.603 1.696 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.613 1.692 yz_to_x 606 15.1 0.457 0.468 1.566 1.594 grid_collocate_task_list 119 9.7 1.557 1.581 1.557 1.581 x_to_yz 595 16.2 0.479 0.489 1.427 1.433 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.369 1.394 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.246 1.257 make_images_sizes 4572 15.5 0.005 0.005 0.842 1.200 dbcsr_dot_sd 1205 11.9 0.326 0.329 0.824 1.198 mp_alltoall_i44 4572 16.5 0.838 1.195 0.838 1.195 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.158 1.175 mp_allgather_i34 2286 14.5 0.679 1.171 0.679 1.171 mp_sum_d 4133 12.0 0.723 1.171 0.723 1.171 cp_fm_cholesky_decompose 22 10.9 1.046 1.053 1.046 1.053 multiply_cannon_sync_h2d 9144 15.5 1.043 1.051 1.043 1.051 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=52.434000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=713.727273, yerr=14.554543 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/09e2302ee41b295db331f66d9974babba7191ee1_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 501.854208E+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 1379060. 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 0.026 0.075 96.541 96.543 qs_mol_dyn_low 1 2.0 0.006 0.029 96.130 96.140 qs_forces 11 3.9 0.004 0.011 96.015 96.017 qs_energies 11 4.9 0.002 0.003 92.944 92.954 scf_env_do_scf 11 5.9 0.001 0.001 82.967 82.970 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 75.678 75.687 dbcsr_multiply_generic 2055 12.4 0.105 0.108 54.943 55.257 qs_scf_new_mos 99 7.5 0.000 0.001 53.180 53.358 qs_scf_loop_do_ot 99 8.5 0.001 0.001 53.179 53.358 ot_scf_mini 99 9.5 0.002 0.002 50.708 50.868 velocity_verlet 10 3.0 0.001 0.002 50.275 50.278 multiply_cannon 2055 13.4 0.183 0.187 43.592 44.749 multiply_cannon_loop 2055 14.4 1.503 1.550 41.836 42.940 ot_mini 99 10.5 0.001 0.001 28.935 29.080 qs_ot_get_derivative 99 11.5 0.001 0.002 22.009 22.145 multiply_cannon_multrec 49320 15.4 12.197 12.964 17.073 17.816 rebuild_ks_matrix 110 8.3 0.000 0.001 17.367 17.515 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 17.367 17.515 mp_waitall_1 220248 16.4 14.105 15.807 14.105 15.807 qs_ks_update_qs_env 110 7.6 0.001 0.001 15.403 15.538 qs_ot_get_p 110 10.4 0.001 0.001 12.720 12.894 multiply_cannon_sync_h2d 49320 15.4 10.025 10.748 10.025 10.748 qs_rho_update_rho_low 110 7.6 0.001 0.001 9.471 9.604 calculate_rho_elec 110 8.6 0.021 0.026 9.471 9.603 sum_up_and_integrate 110 10.3 0.036 0.043 9.266 9.279 integrate_v_rspace 110 11.3 0.003 0.005 9.230 9.251 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 8.478 9.028 qs_ot_p2m_diag 48 11.0 0.012 0.019 8.255 8.299 multiply_cannon_metrocomm3 49320 15.4 0.078 0.080 6.569 7.864 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.343 7.860 apply_single 110 13.6 0.000 0.001 7.342 7.860 init_scf_run 11 5.9 0.000 0.001 7.751 7.751 scf_env_initial_rho_setup 11 6.9 0.000 0.001 7.750 7.751 init_scf_loop 11 6.9 0.000 0.000 7.248 7.255 cp_dbcsr_syevd 48 12.0 0.002 0.003 7.220 7.220 rs_pw_transfer 902 11.9 0.012 0.013 6.590 7.112 ot_diis_step 99 11.5 0.006 0.006 6.720 6.720 density_rs2pw 110 9.6 0.004 0.005 6.146 6.629 cp_fm_diag_elpa 48 13.0 0.000 0.000 6.455 6.481 cp_fm_diag_elpa_base 48 14.0 6.429 6.456 6.453 6.479 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.317 6.408 make_m2s 4110 13.4 0.060 0.064 6.160 6.377 make_images 4110 14.4 0.178 0.193 6.065 6.281 mp_sum_l 6514 12.8 4.767 5.927 4.767 5.927 pw_transfer 1331 11.6 0.056 0.069 5.462 5.615 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 5.373 5.530 multiply_cannon_metrocomm1 49320 15.4 0.060 0.062 3.477 5.511 dbcsr_mm_accdrv_process 87628 16.1 1.932 2.038 4.753 5.054 fft3d_ps 1111 14.6 0.778 0.886 4.906 5.054 wfi_extrapolate 11 7.9 0.001 0.001 4.860 4.860 prepare_preconditioner 11 7.9 0.000 0.000 4.763 4.780 make_preconditioner 11 8.9 0.000 0.000 4.763 4.780 potential_pw2rs 110 12.3 0.006 0.007 4.610 4.701 make_full_inverse_cholesky 11 9.9 0.000 0.000 4.520 4.565 fft_wrap_pw1pw2_140 451 13.1 0.167 0.189 4.162 4.320 calculate_dm_sparse 110 9.5 0.001 0.001 3.856 3.959 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.007 3.839 3.845 mp_alltoall_z22v 1111 16.6 3.139 3.497 3.139 3.497 grid_integrate_task_list 110 12.3 3.232 3.493 3.232 3.493 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.402 3.454 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.307 3.350 mp_alltoall_d11v 2046 13.8 2.750 3.208 2.750 3.208 make_images_data 4110 15.4 0.042 0.045 2.644 3.156 hybrid_alltoall_any 4261 16.3 0.081 0.468 2.138 2.859 mp_waitany 14300 13.8 2.238 2.770 2.238 2.770 calculate_first_density_matrix 1 7.0 0.001 0.006 2.732 2.740 cp_fm_cholesky_invert 11 10.9 2.687 2.691 2.687 2.691 make_images_sizes 4110 15.4 0.004 0.004 1.992 2.676 mp_alltoall_i44 4110 16.4 1.988 2.672 1.988 2.672 jit_kernel_multiply 13 15.9 2.541 2.576 2.541 2.576 grid_collocate_task_list 110 9.6 2.081 2.319 2.081 2.319 acc_transpose_blocks 49320 15.4 0.210 0.219 2.090 2.151 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.107 2.122 mp_sum_d 3889 11.9 1.497 2.100 1.497 2.100 x_to_yz 330 15.4 0.259 0.312 1.973 2.003 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=96.543000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.909091, yerr=2.574510 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/09e2302ee41b295db331f66d9974babba7191ee1_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 588.308480E+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 2462973. 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 0.018 0.038 87.297 87.298 qs_mol_dyn_low 1 2.0 0.003 0.005 86.066 87.012 qs_forces 11 3.9 0.003 0.003 85.975 85.977 qs_energies 11 4.9 0.001 0.002 82.065 82.083 scf_env_do_scf 11 5.9 0.001 0.001 72.282 72.285 scf_env_do_scf_inner_loop 99 6.5 0.002 0.008 61.875 61.876 velocity_verlet 10 3.0 0.001 0.002 45.466 45.468 dbcsr_multiply_generic 2055 12.4 0.115 0.119 43.726 44.855 qs_scf_new_mos 99 7.5 0.001 0.001 42.664 42.792 qs_scf_loop_do_ot 99 8.5 0.001 0.001 42.663 42.791 ot_scf_mini 99 9.5 0.003 0.003 40.650 40.761 multiply_cannon 2055 13.4 0.224 0.248 33.329 36.947 multiply_cannon_loop 2055 14.4 0.925 0.955 31.261 34.910 ot_mini 99 10.5 0.001 0.001 22.556 22.709 mp_waitall_1 176588 16.5 11.086 17.571 11.086 17.571 qs_ot_get_derivative 99 11.5 0.002 0.005 16.320 16.445 rebuild_ks_matrix 110 8.3 0.000 0.000 16.127 16.413 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 16.127 16.412 multiply_cannon_multrec 24660 15.4 7.352 9.407 13.570 15.429 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.963 14.228 multiply_cannon_metrocomm3 24660 15.4 0.067 0.069 6.744 13.855 qs_ot_get_p 110 10.4 0.001 0.001 10.527 10.754 init_scf_loop 11 6.9 0.000 0.000 10.350 10.352 sum_up_and_integrate 110 10.3 0.053 0.060 8.416 8.430 integrate_v_rspace 110 11.3 0.003 0.003 8.362 8.378 prepare_preconditioner 11 7.9 0.000 0.000 8.147 8.166 make_preconditioner 11 8.9 0.000 0.000 8.147 8.166 multiply_cannon_sync_h2d 24660 15.4 6.754 8.087 6.754 8.087 make_full_inverse_cholesky 11 9.9 0.000 0.000 7.574 7.782 qs_ot_p2m_diag 48 11.0 0.029 0.044 7.557 7.618 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.505 7.549 calculate_rho_elec 110 8.6 0.040 0.048 7.505 7.548 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.601 7.305 apply_single 110 13.6 0.000 0.001 6.600 7.305 init_scf_run 11 5.9 0.000 0.001 7.236 7.237 scf_env_initial_rho_setup 11 6.9 0.000 0.001 7.236 7.236 cp_dbcsr_syevd 48 12.0 0.003 0.003 6.915 6.917 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 5.791 6.493 make_m2s 4110 13.4 0.057 0.059 5.816 6.379 dbcsr_mm_accdrv_process 52282 16.1 4.754 5.600 6.062 6.374 make_images 4110 14.4 0.398 0.444 5.708 6.267 ot_diis_step 99 11.5 0.010 0.010 6.065 6.065 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.973 5.987 cp_fm_diag_elpa_base 48 14.0 5.893 5.940 5.971 5.984 pw_transfer 1331 11.6 0.066 0.074 5.118 5.304 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 5.011 5.199 mp_sum_l 6514 12.8 3.950 5.147 3.950 5.147 rs_pw_transfer 902 11.9 0.012 0.013 4.438 4.988 density_rs2pw 110 9.6 0.004 0.004 4.407 4.899 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 4.815 4.894 fft3d_ps 1111 14.6 1.107 1.326 4.364 4.544 wfi_extrapolate 11 7.9 0.001 0.001 4.511 4.511 potential_pw2rs 110 12.3 0.008 0.008 4.044 4.080 cp_fm_cholesky_invert 11 10.9 3.843 3.856 3.843 3.856 fft_wrap_pw1pw2_140 451 13.1 0.199 0.218 3.626 3.827 make_images_data 4110 15.4 0.046 0.049 3.177 3.820 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.732 3.738 hybrid_alltoall_any 4261 16.3 0.103 0.445 2.683 3.663 calculate_dm_sparse 110 9.5 0.001 0.001 3.339 3.381 grid_integrate_task_list 110 12.3 3.133 3.371 3.133 3.371 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.084 3.176 mp_alltoall_d11v 2046 13.8 2.583 3.039 2.583 3.039 mp_allgather_i34 2055 14.4 1.392 2.737 1.392 2.737 mp_alltoall_z22v 1111 16.6 2.514 2.651 2.514 2.651 calculate_first_density_matrix 1 7.0 0.001 0.001 2.616 2.620 qs_ot_get_orbitals 99 10.5 0.001 0.001 2.462 2.584 cp_fm_cholesky_decompose 22 10.9 2.483 2.490 2.483 2.490 grid_collocate_task_list 110 9.6 2.048 2.480 2.048 2.480 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.331 2.347 multiply_cannon_metrocomm4 22605 15.4 0.076 0.082 0.823 2.082 dbcsr_complete_redistribute 325 12.2 0.238 0.314 1.731 2.044 mp_waitany 10164 13.8 1.491 2.007 1.491 2.007 jit_kernel_multiply 10 16.3 0.956 1.967 0.956 1.967 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.878 1.879 mp_irecv_dv 57340 16.2 0.700 1.873 0.700 1.873 mp_sum_d 3889 11.9 1.415 1.845 1.415 1.845 multiply_cannon_metrocomm1 24660 15.4 0.035 0.036 0.606 1.760 make_images_sizes 4110 15.4 0.005 0.005 1.268 1.759 mp_alltoall_i44 4110 16.4 1.264 1.754 1.264 1.754 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=87.298000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=554.909091, yerr=6.841777 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/09e2302ee41b295db331f66d9974babba7191ee1_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 662.470656E+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 0.016 0.030 62.349 62.350 qs_mol_dyn_low 1 2.0 0.003 0.004 62.031 62.036 qs_forces 11 3.9 0.007 0.041 61.676 61.677 qs_energies 11 4.9 0.005 0.028 58.254 58.257 scf_env_do_scf 11 5.9 0.001 0.001 50.011 50.012 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 41.093 41.094 velocity_verlet 10 3.0 0.002 0.002 33.988 33.998 dbcsr_multiply_generic 2055 12.4 0.104 0.109 28.985 29.249 qs_scf_new_mos 99 7.5 0.001 0.001 25.392 25.478 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.391 25.478 ot_scf_mini 99 9.5 0.002 0.003 24.123 24.232 multiply_cannon 2055 13.4 0.214 0.226 22.624 23.873 multiply_cannon_loop 2055 14.4 0.618 0.632 21.418 22.388 ot_mini 99 10.5 0.001 0.001 13.861 13.967 rebuild_ks_matrix 110 8.3 0.000 0.000 12.734 12.857 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.733 12.856 multiply_cannon_multrec 16440 15.4 3.899 5.059 10.145 11.211 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.070 11.177 mp_waitall_1 139946 16.5 7.079 9.960 7.079 9.960 qs_ot_get_derivative 99 11.5 0.001 0.002 9.450 9.560 init_scf_loop 11 6.9 0.000 0.000 8.881 8.882 prepare_preconditioner 11 7.9 0.000 0.000 7.083 7.102 make_preconditioner 11 8.9 0.000 0.000 7.083 7.102 multiply_cannon_metrocomm3 16440 15.4 0.042 0.045 4.268 6.976 sum_up_and_integrate 110 10.3 0.061 0.062 6.828 6.843 integrate_v_rspace 110 11.3 0.003 0.003 6.767 6.781 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.341 6.712 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.495 6.504 calculate_rho_elec 110 8.6 0.059 0.059 6.495 6.504 dbcsr_mm_accdrv_process 34862 16.1 4.823 5.400 6.100 6.224 init_scf_run 11 5.9 0.000 0.001 5.793 5.793 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.793 5.793 qs_ot_get_p 110 10.4 0.001 0.001 5.448 5.579 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.970 5.383 apply_single 110 13.6 0.000 0.000 4.969 5.383 density_rs2pw 110 9.6 0.004 0.005 3.536 4.763 make_m2s 4110 13.4 0.050 0.051 4.241 4.596 make_images 4110 14.4 0.391 0.507 4.127 4.482 ot_diis_step 99 11.5 0.011 0.011 4.383 4.383 rs_pw_transfer 902 11.9 0.011 0.011 3.134 4.342 multiply_cannon_sync_h2d 16440 15.4 3.662 4.249 3.662 4.249 pw_transfer 1331 11.6 0.066 0.072 3.955 3.967 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.848 3.861 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.124 3.812 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.801 3.804 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.464 3.464 grid_integrate_task_list 110 12.3 3.164 3.382 3.164 3.382 fft_wrap_pw1pw2_140 451 13.1 0.212 0.216 3.262 3.282 fft3d_ps 1111 14.6 1.093 1.118 3.147 3.159 wfi_extrapolate 11 7.9 0.001 0.001 3.021 3.021 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.911 2.919 cp_fm_diag_elpa_base 48 14.0 2.841 2.875 2.908 2.916 make_images_data 4110 15.4 0.042 0.046 2.418 2.882 hybrid_alltoall_any 4261 16.3 0.105 0.371 2.174 2.864 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.862 2.863 cp_fm_cholesky_invert 11 10.9 2.840 2.846 2.840 2.846 calculate_dm_sparse 110 9.5 0.001 0.001 2.747 2.777 calculate_first_density_matrix 1 7.0 0.001 0.001 2.686 2.688 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.532 2.578 mp_waitany 17072 13.8 1.299 2.563 1.299 2.563 potential_pw2rs 110 12.3 0.011 0.012 2.537 2.552 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.444 2.498 grid_collocate_task_list 110 9.6 2.083 2.467 2.083 2.467 multiply_cannon_metrocomm4 14385 15.4 0.046 0.049 0.880 2.391 mp_alltoall_d11v 2046 13.8 1.924 2.303 1.924 2.303 mp_irecv_dv 48980 15.7 0.809 2.272 0.809 2.272 rs_pw_transfer_RS2PW_140 121 11.5 0.175 0.179 0.925 2.136 mp_sum_l 6514 12.8 1.507 2.081 1.507 2.081 dbcsr_complete_redistribute 325 12.2 0.325 0.364 1.499 1.981 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.973 1.974 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.867 1.884 cp_fm_upper_to_full 70 14.2 1.407 1.803 1.407 1.803 jit_kernel_multiply 8 16.6 0.891 1.753 0.891 1.753 cp_fm_cholesky_decompose 22 10.9 1.658 1.679 1.658 1.679 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.659 1.671 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.050 1.512 mp_allgather_i34 2055 14.4 0.527 1.497 0.527 1.497 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.368 1.468 rs_gather_matrices 110 12.3 0.232 0.261 0.985 1.348 mp_alltoall_z22v 1111 16.6 1.285 1.315 1.285 1.315 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.288 1.297 acc_transpose_blocks 16440 15.4 0.072 0.074 1.270 1.286 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=62.350000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=626.272727, yerr=8.180808 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/09e2302ee41b295db331f66d9974babba7191ee1_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 722.296832E+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 0.032 0.073 68.554 68.555 qs_mol_dyn_low 1 2.0 0.003 0.004 67.914 67.923 qs_forces 11 3.9 0.003 0.003 67.845 67.845 qs_energies 11 4.9 0.006 0.035 64.392 64.395 scf_env_do_scf 11 5.9 0.015 0.020 55.567 55.573 scf_env_do_scf_inner_loop 99 6.5 0.009 0.038 43.095 43.096 velocity_verlet 10 3.0 0.002 0.002 39.099 39.101 dbcsr_multiply_generic 2055 12.4 0.111 0.117 30.096 30.330 qs_scf_new_mos 99 7.5 0.001 0.001 27.377 27.494 qs_scf_loop_do_ot 99 8.5 0.001 0.007 27.376 27.493 ot_scf_mini 99 9.5 0.003 0.008 25.746 25.863 multiply_cannon 2055 13.4 0.247 0.270 22.554 23.457 multiply_cannon_loop 2055 14.4 0.883 0.906 21.034 21.609 ot_mini 99 10.5 0.001 0.001 14.254 14.376 multiply_cannon_multrec 24660 15.4 4.187 6.763 12.606 13.639 rebuild_ks_matrix 110 8.3 0.000 0.000 12.787 12.904 qs_ks_build_kohn_sham_matrix 110 9.3 0.015 0.038 12.787 12.904 init_scf_loop 11 6.9 0.001 0.005 12.417 12.434 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.335 11.438 prepare_preconditioner 11 7.9 0.000 0.001 10.430 10.445 make_preconditioner 11 8.9 0.000 0.002 10.430 10.445 qs_ot_get_derivative 99 11.5 0.001 0.001 10.070 10.186 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.512 10.009 dbcsr_mm_accdrv_process 52304 16.0 6.830 8.362 8.271 9.241 sum_up_and_integrate 110 10.3 0.067 0.072 6.893 6.907 integrate_v_rspace 110 11.3 0.003 0.003 6.826 6.838 mp_waitall_1 121746 16.5 4.826 6.646 4.826 6.646 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.455 6.466 calculate_rho_elec 110 8.6 0.078 0.081 6.455 6.466 qs_ot_get_p 110 10.4 0.001 0.001 6.302 6.440 make_m2s 4110 13.4 0.060 0.062 5.852 6.187 make_images 4110 14.4 0.577 0.698 5.711 6.042 init_scf_run 11 5.9 0.000 0.001 5.958 5.958 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.957 5.958 cp_fm_upper_to_full 70 14.2 3.373 4.915 3.373 4.915 qs_ot_p2m_diag 48 11.0 0.055 0.064 4.362 4.376 ot_diis_step 99 11.5 0.011 0.011 4.148 4.150 dbcsr_complete_redistribute 325 12.2 0.420 0.463 2.982 4.140 pw_transfer 1331 11.6 0.065 0.075 4.095 4.137 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.017 4.120 apply_single 110 13.6 0.000 0.000 4.017 4.119 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.989 4.034 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.873 3.874 density_rs2pw 110 9.6 0.004 0.004 3.244 3.751 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.366 3.510 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.426 3.484 make_images_data 4110 15.4 0.045 0.049 3.054 3.447 fft_wrap_pw1pw2_140 451 13.1 0.202 0.212 3.393 3.441 grid_integrate_task_list 110 12.3 3.271 3.438 3.271 3.438 wfi_extrapolate 11 7.9 0.001 0.001 3.376 3.376 multiply_cannon_sync_h2d 24660 15.4 3.169 3.355 3.169 3.355 hybrid_alltoall_any 4261 16.3 0.120 0.458 2.536 3.330 fft3d_ps 1111 14.6 1.097 1.132 3.280 3.311 multiply_cannon_metrocomm3 24660 15.4 0.035 0.036 1.390 3.231 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.101 3.164 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.138 3.149 cp_fm_diag_elpa_base 48 14.0 2.989 3.050 3.136 3.147 rs_pw_transfer 902 11.9 0.010 0.012 2.614 3.138 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.909 3.045 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.020 3.022 calculate_dm_sparse 110 9.5 0.001 0.001 2.917 2.952 mp_alltoall_i22 605 13.7 1.713 2.887 1.713 2.887 cp_fm_cholesky_invert 11 10.9 2.774 2.784 2.774 2.784 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.529 2.571 mp_alltoall_d11v 2046 13.8 2.152 2.498 2.152 2.498 calculate_first_density_matrix 1 7.0 0.001 0.002 2.467 2.470 grid_collocate_task_list 110 9.6 2.183 2.443 2.183 2.443 potential_pw2rs 110 12.3 0.012 0.013 2.422 2.434 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.323 2.324 cp_fm_cholesky_decompose 22 10.9 1.964 2.009 1.964 2.009 jit_kernel_multiply 10 15.8 1.110 1.814 1.110 1.814 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.707 1.741 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.698 1.710 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.598 1.699 mp_sum_l 6514 12.8 1.068 1.654 1.068 1.654 mp_waitany 13376 13.8 1.088 1.571 1.088 1.571 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.522 1.543 acc_transpose_blocks 24660 15.4 0.107 0.109 1.515 1.537 multiply_cannon_metrocomm4 20550 15.4 0.062 0.065 0.861 1.493 mp_alltoall_z22v 1111 16.6 1.320 1.431 1.320 1.431 mp_irecv_dv 62702 16.1 0.759 1.415 0.759 1.415 rs_gather_matrices 110 12.3 0.266 0.294 1.045 1.383 qs_env_update_s_mstruct 11 6.9 0.001 0.002 1.277 1.382 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=68.555000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=686.181818, yerr=7.221410 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/09e2302ee41b295db331f66d9974babba7191ee1_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 828.129280E+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 0.039 0.057 56.914 56.917 qs_mol_dyn_low 1 2.0 0.003 0.004 56.653 56.663 qs_forces 11 3.9 0.003 0.004 56.460 56.461 qs_energies 11 4.9 0.001 0.001 52.774 52.780 scf_env_do_scf 11 5.9 0.001 0.002 44.446 44.446 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 36.206 36.207 velocity_verlet 10 3.0 0.002 0.002 32.397 32.408 dbcsr_multiply_generic 2055 12.4 0.103 0.107 23.139 23.281 qs_scf_new_mos 99 7.5 0.001 0.001 20.659 20.707 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.659 20.706 ot_scf_mini 99 9.5 0.002 0.002 19.384 19.401 multiply_cannon 2055 13.4 0.248 0.262 17.492 18.982 multiply_cannon_loop 2055 14.4 0.321 0.333 16.104 16.468 rebuild_ks_matrix 110 8.3 0.000 0.000 12.206 12.233 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 12.205 12.233 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.873 10.897 ot_mini 99 10.5 0.001 0.001 10.525 10.538 multiply_cannon_multrec 8220 15.4 3.219 4.456 7.433 8.297 init_scf_loop 11 6.9 0.000 0.000 8.189 8.190 mp_waitall_1 103326 16.6 6.356 7.939 6.356 7.939 qs_ot_get_derivative 99 11.5 0.001 0.002 6.713 6.735 sum_up_and_integrate 110 10.3 0.080 0.081 6.699 6.709 integrate_v_rspace 110 11.3 0.003 0.003 6.618 6.629 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.582 6.598 calculate_rho_elec 110 8.6 0.116 0.116 6.581 6.598 prepare_preconditioner 11 7.9 0.000 0.000 6.485 6.490 make_preconditioner 11 8.9 0.000 0.000 6.485 6.490 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.079 6.153 init_scf_run 11 5.9 0.000 0.001 5.158 5.158 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.158 5.158 dbcsr_mm_accdrv_process 17442 15.9 2.808 3.598 4.083 5.015 qs_ot_get_p 110 10.4 0.001 0.001 4.862 4.875 make_m2s 4110 13.4 0.039 0.040 4.352 4.653 make_images 4110 14.4 0.637 0.685 4.223 4.525 pw_transfer 1331 11.6 0.066 0.072 4.299 4.308 multiply_cannon_metrocomm3 8220 15.4 0.017 0.018 2.953 4.245 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.191 4.204 ot_diis_step 99 11.5 0.012 0.012 3.788 3.788 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.711 3.733 apply_single 110 13.6 0.000 0.000 3.711 3.733 density_rs2pw 110 9.6 0.004 0.004 3.302 3.679 fft_wrap_pw1pw2_140 451 13.1 0.216 0.220 3.620 3.637 grid_integrate_task_list 110 12.3 3.348 3.535 3.348 3.535 fft3d_ps 1111 14.6 1.148 1.178 3.433 3.442 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.404 3.409 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.089 3.090 cp_fm_cholesky_invert 11 10.9 3.027 3.030 3.027 3.030 multiply_cannon_sync_h2d 8220 15.4 2.916 3.019 2.916 3.019 make_images_data 4110 15.4 0.038 0.042 2.501 2.967 hybrid_alltoall_any 4261 16.3 0.199 0.863 2.422 2.888 wfi_extrapolate 11 7.9 0.001 0.001 2.817 2.817 rs_pw_transfer 902 11.9 0.010 0.011 2.335 2.736 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.681 2.682 calculate_dm_sparse 110 9.5 0.001 0.001 2.513 2.556 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.546 2.554 cp_fm_diag_elpa_base 48 14.0 2.484 2.509 2.543 2.550 grid_collocate_task_list 110 9.6 2.277 2.538 2.277 2.538 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.492 2.494 calculate_first_density_matrix 1 7.0 0.001 0.001 2.229 2.230 potential_pw2rs 110 12.3 0.015 0.015 2.222 2.229 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.153 2.168 cp_fm_cholesky_decompose 22 10.9 2.025 2.047 2.025 2.047 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.770 1.986 mp_alltoall_d11v 2046 13.8 1.703 1.985 1.703 1.985 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.925 1.936 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.840 1.851 mp_allgather_i34 2055 14.4 0.568 1.704 0.568 1.704 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.668 1.670 qs_env_update_s_mstruct 11 6.9 0.001 0.002 1.522 1.647 dbcsr_complete_redistribute 325 12.2 0.557 0.578 1.442 1.538 mp_waitany 9240 13.8 1.100 1.528 1.100 1.528 multiply_cannon_metrocomm1 8220 15.4 0.021 0.021 1.010 1.506 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.439 1.452 qs_create_task_list 11 7.9 0.000 0.001 1.222 1.327 generate_qs_task_list 11 8.9 0.375 0.442 1.221 1.326 rs_gather_matrices 110 12.3 0.323 0.364 0.959 1.239 rs_pw_transfer_RS2PW_140 121 11.5 0.161 0.165 0.807 1.216 jit_kernel_multiply 7 15.8 0.969 1.205 0.969 1.205 mp_alltoall_z22v 1111 16.6 1.173 1.198 1.173 1.198 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.148 1.165 x_to_yz 550 16.1 0.573 0.582 1.148 1.161 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=56.917000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=780.909091, yerr=11.719871 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/09e2302ee41b295db331f66d9974babba7191ee1_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 1.352557E+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 0.019 0.041 88.086 88.087 qs_mol_dyn_low 1 2.0 0.003 0.004 87.782 87.793 qs_forces 11 3.9 0.003 0.004 87.687 87.689 qs_energies 11 4.9 0.003 0.004 83.408 83.411 scf_env_do_scf 11 5.9 0.001 0.001 73.384 73.385 velocity_verlet 10 3.0 0.017 0.034 56.316 56.322 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 44.826 44.827 init_scf_loop 11 6.9 0.000 0.000 28.480 28.482 dbcsr_multiply_generic 2055 12.4 0.117 0.121 28.324 28.405 prepare_preconditioner 11 7.9 0.000 0.000 26.408 26.414 make_preconditioner 11 8.9 0.000 0.000 26.408 26.414 qs_scf_new_mos 99 7.5 0.001 0.001 25.991 26.028 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.991 26.027 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.591 25.901 ot_scf_mini 99 9.5 0.002 0.002 24.254 24.273 multiply_cannon 2055 13.4 0.344 0.363 21.254 22.095 multiply_cannon_loop 2055 14.4 0.341 0.343 19.479 19.817 cp_fm_upper_to_full 70 14.2 12.686 18.251 12.686 18.251 rebuild_ks_matrix 110 8.3 0.001 0.001 14.567 14.591 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 14.566 14.591 ot_mini 99 10.5 0.001 0.001 13.524 13.546 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.114 13.138 dbcsr_complete_redistribute 325 12.2 1.021 1.040 7.397 10.586 multiply_cannon_multrec 8220 15.4 4.351 4.536 9.542 9.651 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.368 9.560 qs_ot_get_derivative 99 11.5 0.002 0.002 9.048 9.067 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.801 8.971 mp_alltoall_i22 605 13.7 5.435 8.615 5.435 8.615 mp_waitall_1 84994 16.7 7.551 8.335 7.551 8.335 qs_rho_update_rho_low 110 7.6 0.001 0.001 8.021 8.060 calculate_rho_elec 110 8.6 0.227 0.228 8.021 8.059 sum_up_and_integrate 110 10.3 0.150 0.152 7.684 7.699 integrate_v_rspace 110 11.3 0.004 0.004 7.534 7.549 make_m2s 4110 13.4 0.043 0.044 5.405 5.898 init_scf_run 11 5.9 0.000 0.001 5.745 5.745 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.745 5.745 make_images 4110 14.4 0.882 0.936 5.217 5.710 qs_ot_get_p 110 10.4 0.001 0.001 5.586 5.605 pw_transfer 1331 11.6 0.076 0.076 5.555 5.567 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 5.438 5.450 cp_fm_cholesky_invert 11 10.9 5.391 5.396 5.391 5.396 dbcsr_mm_accdrv_process 11614 15.7 3.196 3.367 5.048 5.310 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.595 5.064 apply_single 110 13.6 0.000 0.000 4.595 5.064 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 4.555 4.920 fft_wrap_pw1pw2_140 451 13.1 0.222 0.224 4.719 4.727 fft3d_ps 1111 14.6 1.297 1.306 4.631 4.643 ot_diis_step 99 11.5 0.015 0.015 4.461 4.462 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.087 4.096 density_rs2pw 110 9.6 0.004 0.004 4.026 4.061 multiply_cannon_sync_h2d 8220 15.4 3.944 3.953 3.944 3.953 grid_integrate_task_list 110 12.3 3.682 3.755 3.682 3.755 hybrid_alltoall_any 4261 16.3 0.257 0.554 2.950 3.719 qs_energies_init_hamiltonians 11 5.9 0.002 0.004 3.667 3.669 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.646 3.646 make_images_data 4110 15.4 0.041 0.044 2.975 3.626 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.874 3.317 wfi_extrapolate 11 7.9 0.001 0.001 3.289 3.289 calculate_dm_sparse 110 9.5 0.001 0.001 3.096 3.118 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.041 3.041 cp_fm_diag_elpa_base 48 14.0 2.496 2.706 3.040 3.040 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.837 2.841 potential_pw2rs 110 12.3 0.021 0.021 2.714 2.720 grid_collocate_task_list 110 9.6 2.646 2.689 2.646 2.689 rs_pw_transfer 902 11.9 0.011 0.011 2.477 2.507 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.480 2.491 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.354 2.364 calculate_first_density_matrix 1 7.0 0.002 0.004 2.334 2.335 qs_env_update_s_mstruct 11 6.9 0.001 0.001 2.197 2.265 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.086 2.175 mp_alltoall_d11v 2046 13.8 2.032 2.109 2.032 2.109 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.055 2.057 cp_fm_cholesky_decompose 22 10.9 2.001 2.019 2.001 2.019 qs_create_task_list 11 7.9 0.001 0.001 1.886 1.935 generate_qs_task_list 11 8.9 0.726 0.779 1.886 1.934 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.752 1.788 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=88.087000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1214.272727, yerr=63.028263 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/09e2302ee41b295db331f66d9974babba7191ee1_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 638.717952E+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 2520958. 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 0.028 0.033 236.060 236.061 qs_mol_dyn_low 1 2.0 0.003 0.003 235.667 235.681 qs_forces 11 3.9 0.005 0.006 235.565 235.568 qs_energies 11 4.9 0.001 0.002 229.470 229.491 scf_env_do_scf 11 5.9 0.001 0.002 211.652 211.656 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 185.024 185.027 velocity_verlet 10 3.0 0.002 0.003 142.011 142.013 qs_scf_new_mos 117 7.6 0.001 0.001 139.211 139.502 qs_scf_loop_do_ot 117 8.6 0.001 0.001 139.210 139.502 dbcsr_multiply_generic 2507 12.6 0.177 0.180 131.624 132.984 ot_scf_mini 117 9.6 0.003 0.003 132.281 132.538 multiply_cannon 2507 13.6 0.243 0.251 103.958 107.092 multiply_cannon_loop 2507 14.6 2.078 2.126 100.775 103.948 ot_mini 117 10.6 0.001 0.001 70.951 71.227 qs_ot_get_derivative 117 11.6 0.001 0.002 45.611 45.862 multiply_cannon_multrec 60168 15.6 32.973 35.322 41.157 43.751 qs_ot_get_p 128 10.4 0.001 0.001 38.421 38.793 rebuild_ks_matrix 128 8.3 0.001 0.001 36.905 37.517 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 36.904 37.516 mp_waitall_1 267128 16.5 34.445 37.187 34.445 37.187 qs_ks_update_qs_env 128 7.6 0.001 0.001 33.131 33.710 qs_ot_p2m_diag 83 11.4 0.079 0.091 30.281 30.367 multiply_cannon_sync_h2d 60168 15.6 27.014 30.086 27.014 30.086 cp_dbcsr_syevd 83 12.4 0.005 0.005 27.398 27.400 init_scf_loop 11 6.9 0.000 0.000 26.540 26.544 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.663 26.144 apply_single 128 13.6 0.001 0.001 24.663 26.144 ot_diis_step 117 11.6 0.008 0.008 25.033 25.034 cp_fm_diag_elpa 83 13.4 0.000 0.000 23.477 23.516 cp_fm_diag_elpa_base 83 14.4 23.334 23.405 23.472 23.511 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 21.726 21.916 prepare_preconditioner 11 7.9 0.000 0.000 21.584 21.639 make_preconditioner 11 8.9 0.000 0.000 21.584 21.639 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.687 20.925 multiply_cannon_metrocomm3 60168 15.6 0.113 0.118 16.219 18.971 make_m2s 5014 13.6 0.102 0.108 17.130 17.590 make_images 5014 14.6 0.404 0.423 16.951 17.420 qs_rho_update_rho_low 128 7.7 0.001 0.001 17.093 17.211 calculate_rho_elec 128 8.7 0.046 0.065 17.092 17.211 sum_up_and_integrate 128 10.3 0.089 0.108 16.913 16.928 integrate_v_rspace 128 11.3 0.004 0.005 16.824 16.842 init_scf_run 11 5.9 0.000 0.001 13.530 13.530 scf_env_initial_rho_setup 11 6.9 0.000 0.001 13.529 13.530 density_rs2pw 128 9.7 0.006 0.007 10.207 13.441 cp_fm_cholesky_invert 11 10.9 13.014 13.023 13.014 13.023 rs_pw_transfer 1046 11.9 0.017 0.020 9.728 12.964 multiply_cannon_metrocomm1 60168 15.6 0.087 0.092 8.445 11.027 mp_sum_l 7870 13.0 9.282 10.794 9.282 10.794 wfi_extrapolate 11 7.9 0.001 0.001 10.258 10.258 pw_transfer 1547 11.6 0.078 0.099 9.601 10.024 fft_wrap_pw1pw2 1291 12.7 0.011 0.012 9.394 9.820 make_images_data 5014 15.6 0.065 0.071 8.355 9.811 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 9.348 9.499 hybrid_alltoall_any 5200 16.5 0.290 2.245 6.969 8.963 calculate_dm_sparse 128 9.5 0.001 0.001 8.749 8.870 qs_ot_get_orbitals 117 10.6 0.001 0.001 8.514 8.582 fft3d_ps 1291 14.7 2.103 2.872 8.188 8.528 dbcsr_mm_accdrv_process 124484 16.2 3.183 3.297 7.754 8.255 fft_wrap_pw1pw2_140 523 13.2 0.434 0.511 7.500 7.968 grid_integrate_task_list 128 12.3 7.007 7.612 7.007 7.612 potential_pw2rs 128 12.3 0.009 0.011 7.093 7.233 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.890 6.897 mp_alltoall_d11v 2415 14.1 5.319 6.696 5.319 6.696 cp_fm_cholesky_decompose 22 10.9 6.122 6.137 6.122 6.137 mp_waitany 16020 13.9 2.891 6.026 2.891 6.026 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.837 6.019 grid_collocate_task_list 128 9.7 4.557 5.804 4.557 5.804 make_images_sizes 5014 15.6 0.006 0.007 3.353 5.634 mp_alltoall_i44 5014 16.6 3.347 5.628 3.347 5.628 rs_pw_transfer_RS2PW_140 139 11.5 0.277 0.297 2.184 5.359 mp_sum_d 4465 12.1 3.816 4.889 3.816 4.889 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=236.061000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=599.636364, yerr=9.650932 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/09e2302ee41b295db331f66d9974babba7191ee1_performance_tests/16/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 2.183246E+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 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 828.350464E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100942E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834332048 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58363. MP_Allreduce 10978 967. MP_Sync 87 MP_Alltoall 1969 5618844. MP_SendRecv 12032 47072. MP_ISendRecv 12032 47072. MP_Wait 25916 MP_ISend 11748 212467. MP_IRecv 11748 212467. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.023 0.044 226.746 226.747 qs_mol_dyn_low 1 2.0 0.003 0.005 226.333 226.346 qs_forces 11 3.9 0.005 0.006 226.243 226.247 qs_energies 11 4.9 0.003 0.010 219.105 219.143 scf_env_do_scf 11 5.9 0.002 0.012 200.918 200.929 scf_env_do_scf_inner_loop 117 6.6 0.008 0.035 158.548 158.554 velocity_verlet 10 3.0 0.001 0.002 143.047 143.049 qs_scf_new_mos 117 7.6 0.001 0.001 113.254 113.812 qs_scf_loop_do_ot 117 8.6 0.001 0.003 113.254 113.811 dbcsr_multiply_generic 2507 12.6 0.188 0.193 108.231 111.569 ot_scf_mini 117 9.6 0.004 0.006 108.105 108.778 multiply_cannon 2507 13.6 0.478 0.539 81.248 89.899 multiply_cannon_loop 2507 14.6 1.249 1.296 76.826 85.305 ot_mini 117 10.6 0.001 0.001 58.062 58.700 mp_waitall_1 214728 16.6 31.658 52.820 31.658 52.820 init_scf_loop 11 6.9 0.004 0.036 42.226 42.231 multiply_cannon_metrocomm3 30084 15.6 0.091 0.100 19.151 41.163 rebuild_ks_matrix 128 8.3 0.001 0.001 36.401 37.383 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.020 36.401 37.383 prepare_preconditioner 11 7.9 0.000 0.002 37.222 37.322 make_preconditioner 11 8.9 0.001 0.003 37.222 37.320 make_full_inverse_cholesky 11 9.9 0.000 0.000 35.700 36.449 multiply_cannon_multrec 30084 15.6 21.216 26.216 30.762 36.174 qs_ot_get_derivative 117 11.6 0.008 0.018 34.939 35.635 qs_ks_update_qs_env 128 7.6 0.001 0.001 32.836 33.749 qs_ot_get_p 128 10.4 0.001 0.001 30.393 31.111 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.476 24.478 apply_single 128 13.6 0.001 0.001 22.476 24.478 qs_ot_p2m_diag 83 11.4 0.188 0.216 23.598 23.735 ot_diis_step 117 11.6 0.015 0.015 22.761 22.768 cp_fm_cholesky_invert 11 10.9 22.097 22.116 22.097 22.116 cp_dbcsr_syevd 83 12.4 0.010 0.042 22.011 22.014 multiply_cannon_sync_h2d 30084 15.6 18.528 21.355 18.528 21.355 make_m2s 5014 13.6 0.089 0.093 17.186 18.813 make_images 5014 14.6 1.155 1.340 16.980 18.606 cp_fm_diag_elpa 83 13.4 0.001 0.001 18.218 18.246 cp_fm_diag_elpa_base 83 14.4 17.881 18.033 18.187 18.212 sum_up_and_integrate 128 10.3 0.116 0.134 17.103 17.134 qs_rho_update_rho_low 128 7.7 0.001 0.001 16.970 17.077 calculate_rho_elec 128 8.7 0.088 0.106 16.969 17.076 integrate_v_rspace 128 11.3 0.004 0.004 16.986 17.021 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 14.306 14.793 init_scf_run 11 5.9 0.000 0.001 13.096 13.098 scf_env_initial_rho_setup 11 6.9 0.000 0.001 13.096 13.097 make_images_data 5014 15.6 0.063 0.069 10.281 12.446 density_rs2pw 128 9.7 0.006 0.008 10.014 12.437 hybrid_alltoall_any 5200 16.5 0.344 1.516 8.595 11.662 rs_pw_transfer 1046 11.9 0.014 0.017 8.839 11.372 mp_sum_l 7870 13.0 8.031 11.029 8.031 11.029 multiply_cannon_metrocomm4 27577 15.6 0.099 0.113 3.889 11.007 pw_transfer 1547 11.6 0.088 0.118 10.714 10.808 mp_irecv_dv 69486 16.3 3.695 10.625 3.695 10.625 fft_wrap_pw1pw2 1291 12.7 0.011 0.013 10.486 10.585 cp_fm_cholesky_decompose 22 10.9 9.730 9.819 9.730 9.819 wfi_extrapolate 11 7.9 0.001 0.002 9.672 9.672 dbcsr_mm_accdrv_process 62242 16.2 4.553 5.483 9.035 9.614 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.117 8.963 fft3d_ps 1291 14.7 2.761 2.932 8.856 8.914 fft_wrap_pw1pw2_140 523 13.2 0.470 0.534 8.481 8.596 grid_integrate_task_list 128 12.3 7.124 7.664 7.124 7.664 potential_pw2rs 128 12.3 0.016 0.019 7.183 7.229 calculate_dm_sparse 128 9.5 0.001 0.001 6.832 7.011 mp_allgather_i34 2507 14.6 2.991 6.636 2.991 6.636 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.572 6.584 qs_ot_get_orbitals 117 10.6 0.001 0.001 6.410 6.509 mp_alltoall_d11v 2415 14.1 5.302 6.140 5.302 6.140 grid_collocate_task_list 128 9.7 4.704 5.873 4.704 5.873 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 5.528 5.721 mp_sum_d 4465 12.1 3.896 5.538 3.896 5.538 mp_waitany 11748 13.9 3.013 5.511 3.013 5.511 dbcsr_complete_redistribute 395 12.7 0.770 0.848 3.898 4.901 dbcsr_dot_sd 1318 12.0 0.774 0.854 3.368 4.713 rs_pw_transfer_RS2PW_140 139 11.5 0.351 0.378 2.242 4.663 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=226.747000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=788.545455, yerr=4.599677 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/09e2302ee41b295db331f66d9974babba7191ee1_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430457200640 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986252263424 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992001093632 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613077360640 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239162695680 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239162695680 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.232994E+12 0.0% 0.0% 100.0% flops max/rank 2.951022E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806382528 0.0% 0.0% 100.0% number of processed stacks 4017216 0.0% 0.0% 100.0% average stack size 0.0 0.0 1694.3 marketing flops 145.650931E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 937.091072E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1052064 MPI messages size (bytes): total size 2.737208E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.601750E+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 284544 37295751168 131072 < size <= 4194304 665856 1004519030784 4194304 < size <= 16777216 66080 937890532432 16777216 < size 28896 757491302400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 57635. MP_Allreduce 11104 1003. MP_Sync 88 MP_Alltoall 1724 9394234. MP_SendRecv 7998 75008. MP_ISendRecv 7998 75008. MP_Wait 21986 MP_ISend 11836 275177. MP_IRecv 11836 275177. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.028 0.054 212.006 212.010 qs_mol_dyn_low 1 2.0 0.003 0.005 211.434 211.446 qs_forces 11 3.9 0.009 0.045 211.326 211.329 qs_energies 11 4.9 0.004 0.025 203.981 204.036 scf_env_do_scf 11 5.9 0.007 0.052 185.718 185.720 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 142.015 142.017 velocity_verlet 10 3.0 0.003 0.004 136.516 136.519 qs_scf_new_mos 118 7.6 0.001 0.001 96.789 97.167 qs_scf_loop_do_ot 118 8.6 0.001 0.001 96.788 97.165 dbcsr_multiply_generic 2529 12.6 0.181 0.187 90.976 93.913 ot_scf_mini 118 9.6 0.004 0.004 92.210 92.666 multiply_cannon 2529 13.6 0.509 0.530 65.999 73.253 multiply_cannon_loop 2529 14.6 0.863 0.891 61.568 68.467 ot_mini 118 10.6 0.001 0.001 49.311 49.838 mp_waitall_1 172006 16.6 31.169 46.130 31.169 46.130 init_scf_loop 11 6.9 0.001 0.003 43.548 43.551 prepare_preconditioner 11 7.9 0.000 0.001 38.877 38.954 make_preconditioner 11 8.9 0.000 0.002 38.877 38.954 make_full_inverse_cholesky 11 9.9 0.000 0.000 36.394 37.913 rebuild_ks_matrix 129 8.3 0.001 0.001 35.162 36.206 qs_ks_build_kohn_sham_matrix 129 9.3 0.016 0.017 35.162 36.205 multiply_cannon_metrocomm3 20232 15.6 0.059 0.064 17.344 32.755 qs_ks_update_qs_env 129 7.6 0.001 0.001 31.617 32.520 qs_ot_get_derivative 118 11.6 0.002 0.005 28.331 28.802 qs_ot_get_p 129 10.4 0.001 0.001 26.822 27.405 multiply_cannon_multrec 20232 15.6 13.351 16.266 22.048 25.002 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 20.380 21.728 apply_single 129 13.6 0.001 0.001 20.379 21.727 qs_ot_p2m_diag 84 11.4 0.269 0.276 20.958 21.059 ot_diis_step 118 11.6 0.018 0.018 20.725 20.727 cp_dbcsr_syevd 84 12.4 0.006 0.007 19.287 19.292 cp_fm_cholesky_invert 11 10.9 19.202 19.223 19.202 19.223 make_m2s 5058 13.6 0.081 0.085 17.365 18.910 make_images 5058 14.6 1.190 1.273 17.133 18.676 qs_rho_update_rho_low 129 7.7 0.001 0.001 18.214 18.299 calculate_rho_elec 129 8.7 0.133 0.149 18.213 18.299 sum_up_and_integrate 129 10.3 0.133 0.146 17.877 17.904 integrate_v_rspace 129 11.3 0.004 0.005 17.744 17.776 multiply_cannon_sync_h2d 20232 15.6 14.149 15.910 14.149 15.910 cp_fm_diag_elpa 84 13.4 0.001 0.001 15.580 15.591 cp_fm_diag_elpa_base 84 14.4 15.063 15.304 15.574 15.585 make_images_data 5058 15.6 0.060 0.067 10.746 12.850 density_rs2pw 129 9.7 0.006 0.007 10.486 12.668 init_scf_run 11 5.9 0.000 0.001 12.368 12.369 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.368 12.369 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 11.633 12.007 hybrid_alltoall_any 5245 16.5 0.438 1.987 9.064 11.708 pw_transfer 1559 11.6 0.089 0.111 11.212 11.371 rs_pw_transfer 1054 12.0 0.014 0.015 9.250 11.332 fft_wrap_pw1pw2 1301 12.7 0.011 0.012 10.983 11.145 cp_fm_cholesky_decompose 22 10.9 10.402 10.437 10.402 10.437 multiply_cannon_metrocomm4 17703 15.6 0.064 0.077 3.502 9.479 fft3d_ps 1301 14.7 2.691 2.917 9.240 9.355 mp_irecv_dv 50659 16.2 3.380 9.229 3.380 9.229 fft_wrap_pw1pw2_140 527 13.2 0.474 0.524 9.023 9.186 wfi_extrapolate 11 7.9 0.001 0.002 8.840 8.840 dbcsr_mm_accdrv_process 41846 16.2 4.416 5.371 8.173 8.292 mp_sum_l 7936 13.1 5.801 8.075 5.801 8.075 grid_integrate_task_list 129 12.3 7.240 8.048 7.240 8.048 potential_pw2rs 129 12.3 0.021 0.023 7.480 7.550 cp_fm_upper_to_full 106 14.8 5.830 7.505 5.830 7.505 mp_alltoall_d11v 2429 14.1 5.929 7.201 5.929 7.201 dbcsr_complete_redistribute 397 12.7 1.162 1.192 5.265 7.152 mp_allgather_i34 2529 14.6 2.928 6.402 2.928 6.402 calculate_dm_sparse 129 9.5 0.001 0.001 6.224 6.329 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 5.397 6.093 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.043 6.056 grid_collocate_task_list 129 9.7 4.944 5.725 4.944 5.725 copy_fm_to_dbcsr 210 11.7 0.002 0.003 3.819 5.708 mp_waitany 11836 13.9 3.054 5.169 3.054 5.169 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 4.733 5.025 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.895 4.969 mp_alltoall_z22v 1301 16.7 4.548 4.804 4.548 4.804 mp_alltoall_i22 720 14.1 2.499 4.625 2.499 4.625 mp_sum_d 4507 12.2 3.191 4.561 3.191 4.561 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.464 4.331 rs_pw_transfer_RS2PW_140 140 11.5 0.332 0.360 2.147 4.299 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=212.010000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=890.454545, yerr=6.840569 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/09e2302ee41b295db331f66d9974babba7191ee1_performance_tests/18/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 4.353788E+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 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 1.135935E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1143192 MPI messages size (bytes): total size 2.023815E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.770319E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 319024 36042702848 131072 < size <= 4194304 715736 785529176064 4194304 < size <= 16777216 70320 665379241840 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58354. MP_Allreduce 10977 1091. 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 0.068 0.131 223.573 223.575 qs_mol_dyn_low 1 2.0 0.003 0.004 223.012 223.025 qs_forces 11 3.9 0.004 0.004 222.864 222.872 qs_energies 11 4.9 0.003 0.007 215.118 215.126 scf_env_do_scf 11 5.9 0.002 0.006 196.612 196.619 velocity_verlet 10 3.0 0.004 0.005 147.859 147.863 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 139.554 139.556 qs_scf_new_mos 117 7.6 0.001 0.001 96.317 96.652 qs_scf_loop_do_ot 117 8.6 0.001 0.001 96.316 96.651 ot_scf_mini 117 9.6 0.003 0.004 91.480 91.780 dbcsr_multiply_generic 2507 12.6 0.187 0.195 87.506 88.832 multiply_cannon 2507 13.6 0.557 0.582 57.734 60.816 init_scf_loop 11 6.9 0.001 0.003 56.918 56.920 multiply_cannon_loop 2507 14.6 1.181 1.209 52.904 56.475 prepare_preconditioner 11 7.9 0.000 0.000 52.354 52.385 make_preconditioner 11 8.9 0.000 0.003 52.354 52.385 make_full_inverse_cholesky 11 9.9 0.000 0.000 45.569 50.778 ot_mini 117 10.6 0.001 0.001 47.751 48.100 mp_waitall_1 147882 16.7 23.084 36.672 23.084 36.672 rebuild_ks_matrix 128 8.3 0.001 0.001 33.957 34.473 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.019 33.956 34.473 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.557 31.001 multiply_cannon_multrec 30084 15.6 13.861 19.008 25.761 30.446 qs_ot_get_p 128 10.4 0.001 0.001 28.979 29.390 qs_ot_get_derivative 117 11.6 0.002 0.002 27.388 27.705 make_m2s 5014 13.6 0.097 0.100 23.600 25.518 make_images 5014 14.6 1.949 2.244 23.291 25.207 qs_ot_p2m_diag 83 11.4 0.344 0.391 23.063 23.148 cp_fm_cholesky_invert 11 10.9 22.703 22.723 22.703 22.723 cp_dbcsr_syevd 83 12.4 0.006 0.007 21.248 21.250 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 19.672 20.565 apply_single 128 13.6 0.001 0.001 19.671 20.564 ot_diis_step 117 11.6 0.018 0.018 20.150 20.151 multiply_cannon_metrocomm3 30084 15.6 0.047 0.049 7.662 19.496 sum_up_and_integrate 128 10.3 0.140 0.153 17.828 17.851 integrate_v_rspace 128 11.3 0.004 0.004 17.687 17.714 cp_fm_upper_to_full 105 14.8 11.911 17.556 11.911 17.556 cp_fm_diag_elpa 83 13.4 0.001 0.001 17.300 17.313 cp_fm_diag_elpa_base 83 14.4 16.224 16.559 17.291 17.303 qs_rho_update_rho_low 128 7.7 0.001 0.001 17.214 17.270 calculate_rho_elec 128 8.7 0.176 0.190 17.214 17.270 make_images_data 5014 15.6 0.063 0.067 13.216 16.319 hybrid_alltoall_any 5200 16.5 0.527 2.212 11.315 13.992 dbcsr_complete_redistribute 395 12.7 1.501 1.633 10.000 13.964 multiply_cannon_sync_h2d 30084 15.6 11.659 12.859 11.659 12.859 init_scf_run 11 5.9 0.000 0.001 12.390 12.392 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.390 12.391 copy_fm_to_dbcsr 209 11.7 0.002 0.002 8.239 12.258 dbcsr_mm_accdrv_process 62264 16.2 7.316 8.349 11.480 11.952 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 11.486 11.709 pw_transfer 1547 11.6 0.089 0.107 11.322 11.401 fft_wrap_pw1pw2 1291 12.7 0.011 0.012 11.092 11.177 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.762 10.779 density_rs2pw 128 9.7 0.006 0.007 9.337 10.707 cp_fm_cholesky_decompose 22 10.9 10.342 10.447 10.342 10.447 mp_alltoall_i22 716 14.1 6.000 9.859 6.000 9.859 fft_wrap_pw1pw2_140 523 13.2 0.480 0.495 9.396 9.505 fft3d_ps 1291 14.7 2.786 2.887 9.321 9.380 rs_pw_transfer 1046 11.9 0.014 0.015 7.848 9.268 wfi_extrapolate 11 7.9 0.002 0.005 8.844 8.844 grid_integrate_task_list 128 12.3 7.456 8.046 7.456 8.046 mp_alltoall_d11v 2415 14.1 6.670 7.965 6.670 7.965 potential_pw2rs 128 12.3 0.023 0.023 7.093 7.133 multiply_cannon_metrocomm4 25070 15.6 0.084 0.097 2.804 7.050 mp_irecv_dv 76098 16.2 2.658 6.782 2.658 6.782 calculate_dm_sparse 128 9.5 0.001 0.001 6.552 6.671 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.165 6.229 grid_collocate_task_list 128 9.7 5.080 5.881 5.080 5.881 mp_sum_l 7870 13.0 4.275 5.705 4.275 5.705 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 5.336 5.415 mp_allgather_i34 2507 14.6 2.922 4.917 2.922 4.917 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.800 4.914 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 4.592 4.773 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 4.577 4.578 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=223.575000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1071.909091, yerr=19.378782 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/09e2302ee41b295db331f66d9974babba7191ee1_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430457200640 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986252263424 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992001093632 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613077360640 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239162695680 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239162695680 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.232994E+12 0.0% 0.0% 100.0% flops max/rank 5.910120E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806382528 0.0% 0.0% 100.0% number of processed stacks 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 1.484501E+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 5.526746E+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 726592214928 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 9010 51. MP_Alltoall 9724 794507. MP_ISend 40420 2100460. MP_IRecv 40420 2099564. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 57630. MP_Allreduce 11104 1171. MP_Sync 88 MP_Alltoall 1724 18848081. 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 0.041 0.061 214.377 214.391 qs_mol_dyn_low 1 2.0 0.003 0.004 213.879 213.902 qs_forces 11 3.9 0.004 0.005 213.381 213.389 qs_energies 11 4.9 0.002 0.002 205.397 205.411 scf_env_do_scf 11 5.9 0.001 0.003 185.996 185.999 scf_env_do_scf_inner_loop 118 6.6 0.003 0.010 140.433 140.436 velocity_verlet 10 3.0 0.003 0.004 140.209 140.216 qs_scf_new_mos 118 7.6 0.001 0.001 95.285 95.464 qs_scf_loop_do_ot 118 8.6 0.001 0.001 95.284 95.464 ot_scf_mini 118 9.6 0.003 0.004 90.287 90.407 dbcsr_multiply_generic 2529 12.6 0.181 0.194 85.328 86.677 multiply_cannon 2529 13.6 0.592 0.632 60.027 65.599 multiply_cannon_loop 2529 14.6 0.448 0.456 53.739 56.943 ot_mini 118 10.6 0.001 0.001 47.517 47.683 init_scf_loop 11 6.9 0.000 0.000 45.321 45.324 mp_waitall_1 126876 16.7 32.945 44.826 32.945 44.826 prepare_preconditioner 11 7.9 0.000 0.000 40.682 40.715 make_preconditioner 11 8.9 0.000 0.000 40.681 40.715 make_full_inverse_cholesky 11 9.9 0.000 0.000 38.120 38.635 rebuild_ks_matrix 129 8.3 0.001 0.001 33.499 33.864 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.018 33.498 33.864 qs_ks_update_qs_env 129 7.6 0.001 0.001 30.510 30.842 qs_ot_get_p 129 10.4 0.001 0.001 28.544 28.790 qs_ot_get_derivative 118 11.6 0.002 0.003 25.598 25.724 cp_fm_cholesky_invert 11 10.9 24.098 24.106 24.098 24.106 multiply_cannon_metrocomm3 10116 15.6 0.023 0.024 13.622 23.982 make_m2s 5058 13.6 0.067 0.075 19.833 23.286 qs_ot_p2m_diag 84 11.4 0.502 0.507 23.070 23.091 make_images 5058 14.6 2.312 2.666 19.527 22.986 ot_diis_step 118 11.6 0.020 0.021 21.729 21.729 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 20.937 21.686 apply_single 129 13.6 0.001 0.001 20.937 21.686 multiply_cannon_multrec 10116 15.6 10.414 14.168 17.713 21.650 cp_dbcsr_syevd 84 12.4 0.005 0.005 21.276 21.282 qs_rho_update_rho_low 129 7.7 0.001 0.001 19.171 19.260 calculate_rho_elec 129 8.7 0.261 0.271 19.170 19.259 sum_up_and_integrate 129 10.3 0.183 0.194 17.363 17.416 cp_fm_diag_elpa 84 13.4 0.000 0.000 17.405 17.414 cp_fm_diag_elpa_base 84 14.4 17.075 17.232 17.394 17.402 integrate_v_rspace 129 11.3 0.004 0.004 17.179 17.242 make_images_data 5058 15.6 0.051 0.060 12.230 16.052 hybrid_alltoall_any 5245 16.5 0.823 3.649 11.583 15.021 pw_transfer 1559 11.6 0.088 0.095 12.530 12.599 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 12.303 12.379 multiply_cannon_sync_h2d 10116 15.6 11.534 12.278 11.534 12.278 init_scf_run 11 5.9 0.000 0.001 12.083 12.083 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.083 12.083 multiply_cannon_metrocomm1 10116 15.6 0.028 0.029 7.172 11.362 density_rs2pw 129 9.7 0.005 0.006 10.081 10.980 mp_allgather_i34 2529 14.6 4.466 10.677 4.466 10.677 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 10.510 10.595 fft_wrap_pw1pw2_140 527 13.2 0.494 0.515 10.507 10.594 fft3d_ps 1301 14.7 2.739 2.822 10.458 10.508 cp_fm_cholesky_decompose 22 10.9 10.408 10.504 10.408 10.504 wfi_extrapolate 11 7.9 0.001 0.001 8.953 8.953 mp_alltoall_d11v 2429 14.1 7.539 8.495 7.539 8.495 grid_integrate_task_list 129 12.3 7.737 8.272 7.737 8.272 rs_pw_transfer 1054 12.0 0.013 0.014 6.638 7.609 dbcsr_mm_accdrv_process 20934 16.1 2.650 3.506 6.941 7.561 dbcsr_complete_redistribute 397 12.7 2.130 2.175 6.646 7.137 calculate_dm_sparse 129 9.5 0.001 0.001 6.708 6.834 grid_collocate_task_list 129 9.7 5.440 6.157 5.440 6.157 potential_pw2rs 129 12.3 0.026 0.027 6.088 6.123 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.981 5.988 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 5.357 5.359 mp_alltoall_z22v 1301 16.7 4.942 5.160 4.942 5.160 multiply_cannon_metrocomm4 7587 15.6 0.024 0.028 1.912 5.119 mp_irecv_dv 29102 15.9 1.877 5.028 1.877 5.028 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.821 4.844 copy_fm_to_dbcsr 210 11.7 0.001 0.002 4.195 4.593 copy_dbcsr_to_fm 187 11.8 0.004 0.004 4.460 4.585 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.425 4.554 mp_sum_l 7936 13.1 3.372 4.542 3.372 4.542 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.327 4.467 make_images_sizes 5058 15.6 0.007 0.008 2.178 4.428 mp_alltoall_i44 5058 16.6 2.171 4.421 2.171 4.421 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=214.391000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1386.545455, yerr=38.798696 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/09e2302ee41b295db331f66d9974babba7191ee1_performance_tests/20/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 11.786061E+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 1980288 0.0% 0.0% 100.0% average stack size 0.0 0.0 3437.1 marketing flops 145.650931E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 3.104616E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101160 MPI messages size (bytes): total size 1.144970E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.318403E+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 45648 35433480192 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592466352 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 58564. MP_Allreduce 11104 1511. MP_Sync 88 MP_Alltoall 1724 36993632. MP_SendRecv 1806 218624. MP_ISendRecv 1806 218624. MP_Wait 9876 MP_ISend 6456 1080169. MP_IRecv 6456 1080169. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.040 0.067 341.356 341.361 qs_mol_dyn_low 1 2.0 0.003 0.004 340.799 340.821 qs_forces 11 3.9 0.004 0.005 340.672 340.679 qs_energies 11 4.9 0.001 0.002 330.539 330.555 scf_env_do_scf 11 5.9 0.020 0.021 305.947 305.960 velocity_verlet 10 3.0 0.002 0.002 244.972 244.980 scf_env_do_scf_inner_loop 118 6.6 0.006 0.024 166.213 166.215 init_scf_loop 11 6.9 0.000 0.000 139.385 139.388 prepare_preconditioner 11 7.9 0.000 0.000 133.534 133.563 make_preconditioner 11 8.9 0.000 0.000 133.534 133.563 make_full_inverse_cholesky 11 9.9 0.000 0.000 108.664 130.265 qs_scf_new_mos 118 7.6 0.001 0.001 108.094 108.217 qs_scf_loop_do_ot 118 8.6 0.001 0.001 108.093 108.216 ot_scf_mini 118 9.6 0.004 0.004 102.917 103.031 dbcsr_multiply_generic 2529 12.6 0.212 0.224 93.042 94.072 cp_fm_upper_to_full 106 14.8 53.567 77.137 53.567 77.137 multiply_cannon 2529 13.6 0.721 0.775 63.423 67.310 multiply_cannon_loop 2529 14.6 0.475 0.480 59.257 61.612 ot_mini 118 10.6 0.001 0.001 51.659 51.883 dbcsr_complete_redistribute 397 12.7 4.005 4.048 31.288 44.390 rebuild_ks_matrix 129 8.3 0.001 0.001 43.352 43.671 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.018 43.351 43.670 cp_fm_cholesky_invert 11 10.9 42.774 42.783 42.774 42.783 copy_fm_to_dbcsr 210 11.7 0.002 0.002 27.371 40.461 qs_ks_update_qs_env 129 7.6 0.001 0.001 39.639 39.939 mp_waitall_1 104580 16.8 34.653 38.958 34.653 38.958 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 24.823 37.792 mp_alltoall_i22 720 14.1 22.479 35.666 22.479 35.666 qs_ot_get_p 129 10.4 0.001 0.001 34.424 34.693 qs_ot_get_derivative 118 11.6 0.024 0.030 29.490 29.627 qs_ot_p2m_diag 84 11.4 0.889 0.894 29.053 29.098 cp_dbcsr_syevd 84 12.4 0.006 0.006 26.778 26.781 multiply_cannon_metrocomm3 10116 15.6 0.023 0.024 22.465 25.155 qs_rho_update_rho_low 129 7.7 0.001 0.001 24.731 24.770 calculate_rho_elec 129 8.7 0.486 0.487 24.730 24.769 make_m2s 5058 13.6 0.078 0.081 23.259 24.322 make_images 5058 14.6 3.779 3.853 22.775 23.840 sum_up_and_integrate 129 10.3 0.324 0.325 23.188 23.263 cp_fm_diag_elpa 84 13.4 0.000 0.000 22.982 22.984 cp_fm_diag_elpa_base 84 14.4 18.312 20.069 22.978 22.979 integrate_v_rspace 129 11.3 0.004 0.004 22.863 22.938 ot_diis_step 118 11.6 0.022 0.023 22.029 22.030 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 20.658 21.409 apply_single 129 13.6 0.001 0.001 20.657 21.409 multiply_cannon_multrec 10116 15.6 10.663 12.425 18.140 18.259 pw_transfer 1559 11.6 0.094 0.094 16.435 16.450 fft_wrap_pw1pw2 1301 12.7 0.011 0.012 16.198 16.214 multiply_cannon_sync_h2d 10116 15.6 15.778 15.799 15.778 15.799 make_images_data 5058 15.6 0.060 0.063 12.695 15.542 hybrid_alltoall_any 5245 16.5 1.312 3.052 12.580 15.288 fft3d_ps 1301 14.7 2.772 2.779 14.215 14.230 init_scf_run 11 5.9 0.000 0.001 14.182 14.183 scf_env_initial_rho_setup 11 6.9 0.000 0.001 14.181 14.182 fft_wrap_pw1pw2_140 527 13.2 0.546 0.549 14.038 14.056 mp_alltoall_d11v 2429 14.1 11.993 12.965 11.993 12.965 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 11.980 12.073 cp_fm_cholesky_decompose 22 10.9 11.832 11.867 11.832 11.867 density_rs2pw 129 9.7 0.005 0.005 11.211 11.352 wfi_extrapolate 11 7.9 0.001 0.001 10.946 10.946 dbcsr_mm_accdrv_process 20934 16.1 3.880 5.655 7.232 9.112 grid_integrate_task_list 129 12.3 8.666 8.809 8.666 8.809 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 8.131 8.135 copy_dbcsr_to_fm 187 11.8 0.004 0.004 7.198 7.422 rs_scatter_matrices 140 9.7 3.577 4.548 7.130 7.357 potential_pw2rs 129 12.3 0.041 0.041 7.240 7.252 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 7.122 7.153 calculate_dm_sparse 129 9.5 0.001 0.001 7.078 7.132 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=341.361000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2757.000000, yerr=172.977665 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/09e2302ee41b295db331f66d9974babba7191ee1_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 1.262117E+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 7111212. 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 0.026 0.039 86.075 86.075 qs_energies 1 2.0 0.000 0.000 85.623 85.635 ls_scf 1 3.0 0.000 0.000 84.576 84.588 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.983 73.166 multiply_cannon 111 7.7 0.018 0.021 56.021 57.163 multiply_cannon_loop 111 8.7 0.209 0.223 52.613 53.939 ls_scf_main 1 4.0 0.000 0.000 53.209 53.209 density_matrix_trs4 2 5.0 0.002 0.003 47.308 47.406 ls_scf_init_scf 1 4.0 0.000 0.000 28.323 28.324 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.186 27.242 mp_waitall_1 11031 10.9 22.861 25.484 22.861 25.484 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.075 25.102 multiply_cannon_multrec 2664 9.7 8.184 8.972 15.493 17.198 multiply_cannon_sync_h2d 2664 9.7 13.676 15.607 13.676 15.607 make_m2s 222 7.7 0.008 0.011 13.171 13.713 make_images 222 8.7 0.099 0.109 13.149 13.692 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 9.902 13.184 make_images_data 222 9.7 0.004 0.005 7.653 8.199 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.375 8.096 dbcsr_mm_accdrv_process 4760 10.4 0.515 0.606 6.928 7.864 hybrid_alltoall_any 227 10.6 0.216 1.842 6.576 7.833 dbcsr_mm_accdrv_process_sort 4760 11.4 6.214 7.101 6.214 7.101 calculate_norms 4752 9.8 5.535 6.245 5.535 6.245 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.074 5.203 mp_sum_l 807 5.4 3.256 4.683 3.256 4.683 make_images_sizes 222 9.7 0.000 0.000 0.820 3.781 mp_alltoall_i44 222 10.7 0.819 3.781 0.819 3.781 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.449 3.647 arnoldi_extremal 4 6.8 0.000 0.000 3.511 3.534 arnoldi_normal_ev 4 7.8 0.001 0.004 3.510 3.534 build_subspace 16 8.4 0.009 0.012 3.400 3.404 multiply_cannon_metrocomm4 2442 9.7 0.011 0.014 2.016 3.358 mp_irecv_dv 6231 10.9 2.000 3.330 2.000 3.330 ls_scf_post 1 4.0 0.000 0.000 3.043 3.056 ls_scf_store_result 1 5.0 0.000 0.000 2.862 2.905 dbcsr_special_finalize 555 9.7 0.005 0.006 2.366 2.775 dbcsr_merge_single_wm 555 10.7 0.460 0.588 2.358 2.766 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.538 2.759 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.634 2.722 make_images_pack 222 9.7 2.186 2.616 2.188 2.618 dbcsr_sort_data 658 11.4 2.155 2.496 2.155 2.496 dbcsr_matrix_vector_mult_local 304 10.0 2.020 2.415 2.022 2.417 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.104 2.105 rebuild_ks_matrix 3 7.3 0.000 0.000 2.094 2.096 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.094 2.096 buffer_matrices_ensure_size 222 8.7 1.750 2.024 1.750 2.024 sum_up_and_integrate 3 9.3 0.010 0.013 1.760 1.761 integrate_v_rspace 3 10.3 0.000 0.000 1.749 1.751 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=86.075000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1142.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/09e2302ee41b295db331f66d9974babba7191ee1_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 2.105119E+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 0.027 0.041 92.560 92.561 qs_energies 1 2.0 0.000 0.000 92.067 92.078 ls_scf 1 3.0 0.000 0.000 90.753 90.763 dbcsr_multiply_generic 111 6.7 0.015 0.016 76.163 76.573 multiply_cannon 111 7.7 0.031 0.041 53.602 57.909 ls_scf_main 1 4.0 0.000 0.000 55.969 55.974 multiply_cannon_loop 111 8.7 0.116 0.123 50.301 53.727 density_matrix_trs4 2 5.0 0.002 0.003 50.182 50.379 ls_scf_init_scf 1 4.0 0.000 0.000 31.265 31.266 mp_waitall_1 9105 10.9 21.785 30.587 21.785 30.587 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.963 30.103 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.529 27.544 multiply_cannon_multrec 1332 9.7 13.109 16.868 22.322 27.218 multiply_cannon_metrocomm3 1332 9.7 0.006 0.007 12.018 20.973 make_m2s 222 7.7 0.006 0.008 15.569 16.283 make_images 222 8.7 1.575 1.952 15.539 16.254 dbcsr_mm_accdrv_process 4041 10.4 0.271 0.456 8.812 10.393 make_images_data 222 9.7 0.004 0.004 9.017 10.075 dbcsr_mm_accdrv_process_sort 4041 11.4 8.404 9.938 8.404 9.938 hybrid_alltoall_any 227 10.6 0.521 2.469 8.428 9.483 mp_sum_l 807 5.4 5.685 9.073 5.685 9.073 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.235 7.843 mp_irecv_dv 3311 11.0 3.216 7.784 3.216 7.784 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.432 7.272 calculate_norms 2376 9.8 5.987 6.662 5.987 6.662 multiply_cannon_sync_h2d 1332 9.7 4.781 5.956 4.781 5.956 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.060 5.262 arnoldi_extremal 4 6.8 0.000 0.000 4.997 5.015 arnoldi_normal_ev 4 7.8 0.001 0.004 4.997 5.015 build_subspace 16 8.4 0.014 0.020 4.741 4.746 ls_scf_post 1 4.0 0.000 0.000 3.519 3.530 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.284 3.507 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 1.365 3.393 ls_scf_store_result 1 5.0 0.000 0.000 3.263 3.328 dbcsr_matrix_vector_mult_local 304 10.0 2.733 3.213 2.735 3.215 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.651 2.746 mp_allgather_i34 111 8.7 1.006 2.481 1.006 2.481 make_images_pack 222 9.7 2.026 2.424 2.028 2.426 dbcsr_sort_data 436 11.2 1.786 1.995 1.786 1.995 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.950 1.952 rebuild_ks_matrix 3 7.3 0.000 0.000 1.937 1.939 qs_ks_build_kohn_sham_matrix 3 8.3 0.003 0.020 1.937 1.939 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=92.561000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1705.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/09e2302ee41b295db331f66d9974babba7191ee1_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 2.709348E+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 0.035 0.063 96.665 96.666 qs_energies 1 2.0 0.000 0.000 95.712 95.715 ls_scf 1 3.0 0.000 0.001 94.154 94.157 dbcsr_multiply_generic 111 6.7 0.016 0.018 77.771 78.067 ls_scf_main 1 4.0 0.000 0.000 59.224 59.230 multiply_cannon 111 7.7 0.045 0.101 53.452 57.616 multiply_cannon_loop 111 8.7 0.099 0.106 49.851 53.678 density_matrix_trs4 2 5.0 0.002 0.003 52.613 52.768 mp_waitall_1 7281 11.0 24.768 34.512 24.768 34.512 ls_scf_init_scf 1 4.0 0.000 0.001 31.403 31.405 ls_scf_init_matrix_S 1 5.0 0.000 0.000 30.119 30.220 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 27.712 27.724 multiply_cannon_multrec 888 9.7 12.578 15.447 21.105 24.575 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.514 23.237 make_m2s 222 7.7 0.006 0.007 17.551 18.807 make_images 222 8.7 1.966 2.287 17.513 18.768 make_images_data 222 9.7 0.003 0.004 10.037 11.089 hybrid_alltoall_any 227 10.6 0.620 2.859 9.529 11.026 dbcsr_mm_accdrv_process 3754 10.4 0.253 0.418 8.045 9.259 mp_sum_l 807 5.4 5.242 8.933 5.242 8.933 dbcsr_mm_accdrv_process_sort 3754 11.4 7.674 8.841 7.674 8.841 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.091 7.254 multiply_cannon_sync_h2d 888 9.7 6.015 7.192 6.015 7.192 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.438 6.801 mp_irecv_dv 2335 11.1 2.423 6.758 2.423 6.758 multiply_cannon_metrocomm1 888 9.7 0.002 0.003 3.830 6.536 arnoldi_extremal 4 6.8 0.000 0.000 5.508 5.523 arnoldi_normal_ev 4 7.8 0.001 0.006 5.508 5.523 build_subspace 16 8.4 0.014 0.020 5.208 5.214 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.948 5.163 calculate_norms 1584 9.8 4.298 4.676 4.298 4.676 dbcsr_matrix_vector_mult 304 9.0 0.010 0.020 3.703 4.036 mp_allgather_i34 111 8.7 1.426 4.001 1.426 4.001 dbcsr_matrix_vector_mult_local 304 10.0 3.002 3.582 3.004 3.584 ls_scf_post 1 4.0 0.000 0.000 3.526 3.529 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.322 3.416 ls_scf_store_result 1 5.0 0.000 0.000 3.256 3.352 make_images_sizes 222 9.7 0.000 0.000 1.272 2.616 mp_alltoall_i44 222 10.7 1.272 2.616 1.272 2.616 make_images_pack 222 9.7 1.813 2.136 1.815 2.139 dbcsr_sort_data 325 11.1 1.824 2.094 1.824 2.094 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.976 1.978 dbcsr_data_release 9322 10.9 1.326 1.967 1.326 1.967 rebuild_ks_matrix 3 7.3 0.000 0.000 1.957 1.959 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.004 1.957 1.959 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=96.666000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2212.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/09e2302ee41b295db331f66d9974babba7191ee1_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 3.328655E+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 0.050 0.103 98.623 98.623 qs_energies 1 2.0 0.000 0.000 97.873 97.878 ls_scf 1 3.0 0.001 0.012 96.214 96.220 dbcsr_multiply_generic 111 6.7 0.017 0.017 78.637 78.854 ls_scf_main 1 4.0 0.001 0.010 59.444 59.446 multiply_cannon 111 7.7 0.072 0.175 51.189 56.577 density_matrix_trs4 2 5.0 0.003 0.016 53.170 53.264 multiply_cannon_loop 111 8.7 0.115 0.125 46.093 49.342 ls_scf_init_scf 1 4.0 0.000 0.002 33.505 33.506 ls_scf_init_matrix_S 1 5.0 0.000 0.002 31.902 31.971 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.439 29.450 mp_waitall_1 6369 11.0 22.342 28.890 22.342 28.890 multiply_cannon_multrec 1332 9.7 14.266 17.785 22.114 24.968 make_m2s 222 7.7 0.007 0.008 21.329 22.784 make_images 222 8.7 3.141 3.595 21.279 22.735 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 8.673 17.475 make_images_data 222 9.7 0.004 0.004 11.876 13.494 hybrid_alltoall_any 227 10.6 0.800 3.798 11.094 12.799 dbcsr_mm_accdrv_process 3641 10.4 0.229 0.413 7.486 8.983 dbcsr_mm_accdrv_process_sort 3641 11.4 7.111 8.574 7.111 8.574 mp_sum_l 807 5.4 4.571 8.270 4.571 8.270 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.553 6.528 multiply_cannon_sync_h2d 1332 9.7 5.463 6.062 5.463 6.062 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.099 6.007 arnoldi_extremal 4 6.8 0.000 0.000 5.938 5.951 arnoldi_normal_ev 4 7.8 0.001 0.004 5.938 5.951 mp_irecv_dv 3229 10.9 2.079 5.935 2.079 5.935 build_subspace 16 8.4 0.014 0.020 5.601 5.609 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.623 5.353 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.546 4.737 mp_allgather_i34 111 8.7 2.197 4.637 2.197 4.637 calculate_norms 2376 9.8 4.181 4.507 4.181 4.507 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 4.005 4.252 dbcsr_matrix_vector_mult_local 304 10.0 3.171 3.670 3.173 3.672 dbcsr_sort_data 658 11.4 3.021 3.432 3.021 3.432 ls_scf_post 1 4.0 0.000 0.000 3.264 3.268 dbcsr_special_finalize 555 9.7 0.006 0.007 2.764 3.216 dbcsr_merge_single_wm 555 10.7 0.532 0.663 2.755 3.207 ls_scf_dm_to_ks 2 5.0 0.001 0.013 3.144 3.190 ls_scf_store_result 1 5.0 0.000 0.000 2.960 3.019 dbcsr_data_release 10477 10.7 1.583 2.409 1.583 2.409 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.301 2.303 rebuild_ks_matrix 3 7.3 0.000 0.000 2.277 2.279 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.277 2.279 make_images_sizes 222 9.7 0.000 0.000 0.849 1.999 mp_alltoall_i44 222 10.7 0.849 1.999 0.849 1.999 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=98.623000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2712.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/09e2302ee41b295db331f66d9974babba7191ee1_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 4.632318E+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 0.043 0.074 102.436 102.437 qs_energies 1 2.0 0.000 0.000 101.591 101.594 ls_scf 1 3.0 0.000 0.000 99.498 99.507 dbcsr_multiply_generic 111 6.7 0.017 0.018 79.029 79.259 ls_scf_main 1 4.0 0.000 0.000 63.809 63.810 multiply_cannon 111 7.7 0.073 0.170 55.650 61.073 density_matrix_trs4 2 5.0 0.002 0.003 55.982 56.059 multiply_cannon_loop 111 8.7 0.070 0.083 50.785 52.335 mp_waitall_1 5436 11.0 26.783 33.085 26.783 33.085 ls_scf_init_scf 1 4.0 0.000 0.000 32.068 32.074 ls_scf_init_matrix_S 1 5.0 0.000 0.000 30.687 30.721 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.403 28.411 multiply_cannon_multrec 444 9.7 13.972 16.279 20.998 23.441 make_m2s 222 7.7 0.004 0.005 18.293 21.212 make_images 222 8.7 3.725 4.430 18.231 21.151 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.756 16.032 make_images_data 222 9.7 0.003 0.004 10.204 13.273 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 5.619 13.053 hybrid_alltoall_any 227 10.6 0.793 3.785 10.063 12.885 multiply_cannon_sync_h2d 444 9.7 6.547 8.129 6.547 8.129 dbcsr_mm_accdrv_process 3003 10.4 0.166 0.352 6.733 7.856 dbcsr_mm_accdrv_process_sort 3003 11.4 6.414 7.514 6.414 7.514 mp_allgather_i34 111 8.7 3.077 7.389 3.077 7.389 arnoldi_extremal 4 6.8 0.000 0.000 6.749 6.757 arnoldi_normal_ev 4 7.8 0.002 0.005 6.749 6.757 build_subspace 16 8.4 0.014 0.019 6.343 6.352 mp_sum_l 807 5.4 3.162 5.681 3.162 5.681 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.989 5.193 dbcsr_matrix_vector_mult 304 9.0 0.011 0.020 4.581 4.733 dbcsr_matrix_vector_mult_local 304 10.0 3.669 4.155 3.671 4.157 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.116 4.148 make_images_sizes 222 9.7 0.000 0.000 1.446 4.000 mp_alltoall_i44 222 10.7 1.445 4.000 1.445 4.000 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.823 3.972 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.662 3.932 mp_irecv_dv 1241 11.2 1.643 3.906 1.643 3.906 calculate_norms 792 9.8 3.541 3.651 3.541 3.651 ls_scf_post 1 4.0 0.000 0.000 3.620 3.624 ls_scf_store_result 1 5.0 0.000 0.000 3.380 3.450 dbcsr_finalize 304 7.8 0.062 0.078 2.189 2.256 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.201 2.202 rebuild_ks_matrix 3 7.3 0.000 0.000 2.169 2.170 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.169 2.170 dbcsr_merge_all 275 8.9 0.472 0.522 2.039 2.085 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.077 2.077 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=102.437000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3634.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/09e2302ee41b295db331f66d9974babba7191ee1_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 8.715665E+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 0.109 0.143 113.152 113.153 qs_energies 1 2.0 0.000 0.000 111.646 111.659 ls_scf 1 3.0 0.000 0.000 108.682 108.695 dbcsr_multiply_generic 111 6.7 0.023 0.027 81.759 81.885 ls_scf_main 1 4.0 0.000 0.000 66.920 66.920 density_matrix_trs4 2 5.0 0.002 0.003 57.587 57.659 multiply_cannon 111 7.7 0.132 0.229 53.832 55.549 multiply_cannon_loop 111 8.7 0.068 0.070 50.388 51.449 ls_scf_init_scf 1 4.0 0.000 0.000 38.026 38.027 ls_scf_init_matrix_S 1 5.0 0.000 0.000 36.290 36.305 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 33.457 33.472 mp_waitall_1 4527 11.1 23.503 27.137 23.503 27.137 make_m2s 222 7.7 0.005 0.005 24.124 25.062 make_images 222 8.7 4.594 4.960 24.019 24.955 multiply_cannon_multrec 444 9.7 17.919 18.591 22.539 23.168 hybrid_alltoall_any 227 10.6 1.662 3.651 13.006 15.926 make_images_data 222 9.7 0.003 0.003 13.319 15.717 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 11.588 15.156 multiply_cannon_sync_h2d 444 9.7 8.847 8.893 8.847 8.893 arnoldi_extremal 4 6.8 0.000 0.000 7.551 7.563 arnoldi_normal_ev 4 7.8 0.005 0.013 7.551 7.562 build_subspace 16 8.4 0.026 0.036 6.992 7.005 dbcsr_matrix_vector_mult 304 9.0 0.017 0.032 5.581 5.746 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.267 5.362 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.050 5.318 dbcsr_matrix_vector_mult_local 304 10.0 5.006 5.307 5.009 5.309 dbcsr_mm_accdrv_process 1814 10.4 0.272 0.351 4.445 4.563 dbcsr_mm_accdrv_process_sort 1814 11.4 4.124 4.256 4.124 4.256 compute_matrix_preconditioner 1 6.0 0.002 0.002 3.951 3.959 ls_scf_post 1 4.0 0.000 0.000 3.736 3.749 make_images_sizes 222 9.7 0.000 0.000 1.466 3.707 mp_alltoall_i44 222 10.7 1.466 3.706 1.466 3.706 ls_scf_store_result 1 5.0 0.000 0.000 3.433 3.443 mp_allgather_i34 111 8.7 1.089 3.406 1.089 3.406 acc_transpose_blocks 444 9.7 0.002 0.002 2.962 3.356 acc_transpose_blocks_kernels 444 10.7 0.005 0.005 2.927 3.320 jit_kernel_transpose 1 13.0 2.922 3.316 2.922 3.316 calculate_norms 792 9.8 3.240 3.283 3.240 3.283 dbcsr_finalize 304 7.8 0.082 0.090 3.107 3.195 dbcsr_merge_all 275 8.9 0.892 0.919 2.872 2.950 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.933 2.933 dbcsr_complete_redistribute 5 7.6 1.432 1.472 2.765 2.880 dbcsr_data_release 12724 10.6 2.365 2.871 2.365 2.871 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.624 2.626 mp_sum_l 807 5.4 1.464 2.606 1.464 2.606 rebuild_ks_matrix 3 7.3 0.000 0.000 2.557 2.559 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.557 2.559 matrix_ls_to_qs 2 6.0 0.000 0.000 2.410 2.533 dbcsr_sort_data 325 11.1 2.444 2.505 2.444 2.505 dbcsr_new_transposed 4 7.5 0.245 0.253 2.287 2.300 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=113.153000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6802.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/09e2302ee41b295db331f66d9974babba7191ee1_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 590.594048E+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 1124357819. 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 0.111 317.179 317.184 qs_mol_dyn_low 1 2.0 0.004 0.045 315.354 315.705 qs_forces 5 3.8 0.005 0.007 315.227 315.232 qs_energies 5 4.8 0.005 0.092 311.627 311.682 scf_env_do_scf 5 5.8 0.001 0.008 294.602 294.605 scf_env_do_scf_inner_loop 105 6.6 0.006 0.044 253.388 253.395 qs_scf_new_mos 105 7.6 0.000 0.001 197.564 197.919 qs_scf_loop_do_ot 105 8.6 0.001 0.001 197.563 197.919 ot_scf_mini 105 9.6 0.003 0.004 186.565 186.786 dbcsr_multiply_generic 1445 12.2 0.121 0.128 145.566 147.208 velocity_verlet 4 3.0 0.004 0.013 128.711 128.712 multiply_cannon 1445 13.2 0.279 0.292 120.956 125.789 multiply_cannon_loop 1445 14.2 2.370 2.502 117.775 122.240 qs_ot_get_p 112 10.4 0.001 0.001 87.435 87.996 qs_ot_p2m_diag 40 11.0 0.020 0.030 73.587 73.689 cp_dbcsr_syevd 40 12.0 0.002 0.002 69.927 69.931 ot_mini 105 10.6 0.001 0.001 67.328 67.616 cp_fm_syevd 40 13.0 0.000 0.000 62.820 63.015 mp_waitall_1 488190 16.1 46.819 59.936 46.819 59.936 cp_fm_redistribute_end 40 14.0 28.426 56.708 28.436 56.714 cp_fm_syevd_base 40 14.0 28.259 56.547 28.259 56.547 multiply_cannon_metrocomm3 69360 15.2 0.187 0.197 31.190 46.579 qs_ot_get_derivative 55 11.6 0.002 0.010 44.940 45.195 multiply_cannon_multrec 69360 15.2 30.502 35.904 38.301 43.375 init_scf_loop 7 6.6 0.001 0.022 41.160 41.163 rebuild_ks_matrix 110 8.4 0.000 0.000 36.564 37.081 qs_ks_build_kohn_sham_matrix 110 9.4 0.011 0.014 36.564 37.081 prepare_preconditioner 7 7.6 0.000 0.001 34.783 34.834 make_preconditioner 7 8.6 0.000 0.006 34.783 34.834 qs_ks_update_qs_env 112 7.6 0.001 0.001 33.786 34.286 multiply_cannon_sync_h2d 69360 15.2 28.887 33.680 28.887 33.680 qs_rho_update_rho_low 110 7.6 0.001 0.001 26.797 27.351 calculate_rho_elec 110 8.6 0.029 0.032 26.797 27.350 make_full_inverse_cholesky 7 9.6 0.000 0.000 25.354 25.417 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 23.489 23.784 apply_single 62 13.6 0.000 0.000 23.489 23.784 rs_pw_transfer 690 11.5 0.011 0.013 22.171 23.772 ot_new_cg_direction 55 11.6 0.001 0.001 21.651 21.657 density_rs2pw 110 9.6 0.005 0.006 19.884 21.471 multiply_cannon_metrocomm1 69360 15.2 0.088 0.094 7.085 17.617 cp_fm_cholesky_invert 7 10.6 17.491 17.503 17.491 17.503 qs_ot_get_orbitals 105 10.6 0.001 0.001 16.290 16.645 pw_transfer 1645 12.4 0.080 0.098 16.074 16.555 fft_wrap_pw1pw2 1425 13.5 0.012 0.014 15.936 16.418 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 16.022 16.188 mp_sum_l 4684 12.4 14.247 15.863 14.247 15.863 init_scf_run 5 5.8 0.000 0.004 13.827 13.828 scf_env_initial_rho_setup 5 6.8 0.000 0.001 13.827 13.828 fft_wrap_pw1pw2_240 915 15.0 0.421 0.465 13.403 13.807 sum_up_and_integrate 60 10.3 0.028 0.031 13.290 13.303 integrate_v_rspace 60 11.3 0.002 0.002 13.262 13.275 calculate_dm_sparse 110 9.5 0.000 0.001 12.629 12.941 qs_ot_get_derivative_diag 18 12.0 0.000 0.000 12.631 12.757 fft3d_pb 915 16.0 2.301 2.673 12.235 12.687 qs_vxc_create 110 10.4 0.002 0.002 12.633 12.680 check_diag 80 13.5 8.587 8.889 12.114 12.316 make_m2s 2890 13.2 0.075 0.081 9.935 10.590 make_images 2890 14.2 0.246 0.266 9.834 10.489 make_full_single_inverse 7 9.6 0.001 0.002 9.072 9.124 potential_pw2rs 60 12.3 0.003 0.003 8.818 8.886 xc_vxc_pw_create 60 11.3 0.042 0.067 8.726 8.774 xc_rho_set_and_dset_create 110 12.4 0.085 0.106 8.418 8.711 rs_pw_transfer_RS2PW_30 110 11.6 1.559 1.634 8.255 8.539 mp_sendrecv_dv 168740 12.6 8.334 8.480 8.334 8.480 dbcsr_mm_accdrv_process 154766 15.8 4.026 4.222 7.676 8.480 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.002 8.250 8.265 mp_alltoall_z22v 2340 17.7 7.404 8.193 7.404 8.193 xc_pw_derive 510 13.4 0.005 0.006 8.034 8.119 calculate_first_density_matrix 1 7.0 0.002 0.005 7.758 7.771 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 7.564 7.639 acc_transpose_blocks 69360 15.2 0.357 0.376 7.078 7.578 cp_fm_cholesky_decompose 14 10.2 7.360 7.368 7.360 7.368 mp_waitany 7680 13.5 4.960 6.579 4.960 6.579 mp_alltoall_d11v 1300 13.8 5.486 6.440 5.486 6.440 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=317.184000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=560.200000, yerr=3.429286 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 09e2302ee41b295db331f66d9974babba7191ee1 Summary: empty Status: OK