=== 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: c60ebb48d63dcff4bc85c6fc137b78e381e1214e ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, Cray-FFTW 3.3.8.10, # COSMA 2.6.6, ELPA 2022.11.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.2.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.2, # 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 (31.05.2023) # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.2 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(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.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/c60ebb48d63dcff4bc85c6fc137b78e381e1214e_performance_tests/01 job id: 47181505 --- 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/c60ebb48d63dcff4bc85c6fc137b78e381e1214e_performance_tests/02 job id: 47181506 --- 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/c60ebb48d63dcff4bc85c6fc137b78e381e1214e_performance_tests/03 job id: 47181507 --- 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/c60ebb48d63dcff4bc85c6fc137b78e381e1214e_performance_tests/04 job id: 47181509 --- 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/c60ebb48d63dcff4bc85c6fc137b78e381e1214e_performance_tests/05 job id: 47181510 --- 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/c60ebb48d63dcff4bc85c6fc137b78e381e1214e_performance_tests/06 job id: 47181511 --- 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/c60ebb48d63dcff4bc85c6fc137b78e381e1214e_performance_tests/07 job id: 47181512 --- 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/c60ebb48d63dcff4bc85c6fc137b78e381e1214e_performance_tests/08 job id: 47181513 --- 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/c60ebb48d63dcff4bc85c6fc137b78e381e1214e_performance_tests/09 job id: 47181514 --- 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/c60ebb48d63dcff4bc85c6fc137b78e381e1214e_performance_tests/10 job id: 47181515 --- 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/c60ebb48d63dcff4bc85c6fc137b78e381e1214e_performance_tests/11 job id: 47181516 --- 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/c60ebb48d63dcff4bc85c6fc137b78e381e1214e_performance_tests/12 job id: 47181517 --- 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/c60ebb48d63dcff4bc85c6fc137b78e381e1214e_performance_tests/13 job id: 47181518 --- 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/c60ebb48d63dcff4bc85c6fc137b78e381e1214e_performance_tests/14 job id: 47181520 --- 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/c60ebb48d63dcff4bc85c6fc137b78e381e1214e_performance_tests/15 job id: 47181521 --- 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/c60ebb48d63dcff4bc85c6fc137b78e381e1214e_performance_tests/16 job id: 47181522 --- 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/c60ebb48d63dcff4bc85c6fc137b78e381e1214e_performance_tests/17 job id: 47181523 --- 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/c60ebb48d63dcff4bc85c6fc137b78e381e1214e_performance_tests/18 job id: 47181524 --- 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/c60ebb48d63dcff4bc85c6fc137b78e381e1214e_performance_tests/19 job id: 47181525 --- 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/c60ebb48d63dcff4bc85c6fc137b78e381e1214e_performance_tests/20 job id: 47181526 --- 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/c60ebb48d63dcff4bc85c6fc137b78e381e1214e_performance_tests/21 job id: 47181527 --- 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/c60ebb48d63dcff4bc85c6fc137b78e381e1214e_performance_tests/22 job id: 47181528 --- 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/c60ebb48d63dcff4bc85c6fc137b78e381e1214e_performance_tests/23 job id: 47181529 --- 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/c60ebb48d63dcff4bc85c6fc137b78e381e1214e_performance_tests/24 job id: 47181530 --- 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/c60ebb48d63dcff4bc85c6fc137b78e381e1214e_performance_tests/25 job id: 47181531 --- 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/c60ebb48d63dcff4bc85c6fc137b78e381e1214e_performance_tests/26 job id: 47181532 --- 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/c60ebb48d63dcff4bc85c6fc137b78e381e1214e_performance_tests/27 job id: 47181533 --- 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/c60ebb48d63dcff4bc85c6fc137b78e381e1214e_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 424 8. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.037 134.258 134.259 farming_run 1 2.0 133.675 133.677 134.229 134.232 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.452863E+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 571 1938539. MP_Sync 25 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 10 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.008 0.025 115.409 115.410 qs_energies 1 2.0 0.000 0.000 115.209 115.213 mp2_main 1 3.0 0.000 0.000 113.123 113.126 mp2_gpw_main 1 4.0 0.019 0.023 112.129 112.132 mp2_ri_gpw_compute_in 1 5.0 0.172 0.173 93.200 93.639 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 54.897 55.336 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.167 41.343 46.595 get_2c_integrals 1 6.0 0.008 0.009 37.203 38.130 integrate_v_rspace 273 8.0 0.435 0.451 24.742 29.729 pw_transfer 6555 10.6 0.373 0.394 27.366 27.857 fft_wrap_pw1pw2 5465 11.4 0.045 0.048 26.002 26.521 grid_integrate_task_list 273 9.0 20.563 26.037 20.563 26.037 fft_wrap_pw1pw2_100 2178 12.4 1.179 1.399 23.521 24.056 compute_2c_integrals 1 7.0 0.002 0.002 19.566 19.567 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.860 19.287 mp2_eri_2c_integrate_gpw 1 9.0 2.379 2.433 18.857 19.284 rpa_ri_compute_en 1 5.0 0.019 0.023 18.819 19.027 cp_fm_cholesky_decompose 12 8.2 17.566 18.503 17.566 18.503 cholesky_decomp 1 7.0 0.000 0.000 16.472 17.399 fft3d_s 5443 13.4 16.144 16.607 16.167 16.628 ao_to_mo_and_store_B_mult_1 272 7.0 10.724 15.302 10.724 15.302 calculate_wavefunction 272 8.0 5.402 5.553 12.514 13.280 rpa_num_int 1 6.0 0.000 0.000 10.596 10.596 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.581 10.594 calc_mat_Q 8 8.0 0.000 0.000 9.451 9.532 contract_S_to_Q 8 9.0 0.000 0.000 8.874 8.956 calc_potential_gpw 544 9.5 0.005 0.006 8.264 8.657 parallel_gemm_fm 14 9.1 0.000 0.000 8.467 8.552 parallel_gemm_fm_cosma 14 10.1 8.467 8.552 8.467 8.552 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 8.205 8.493 potential_pw2rs 545 10.0 0.107 0.109 7.670 8.314 collocate_single_gaussian 272 10.0 0.039 0.041 7.461 7.694 create_integ_mat 1 6.0 0.022 0.028 7.690 7.690 array2fm 1 7.0 0.000 0.000 6.636 7.033 pw_scatter_s 2720 13.7 4.433 4.681 4.433 4.681 pw_gather_s 2722 13.2 3.840 4.232 3.840 4.232 array2fm_buffer_send 1 8.0 2.958 3.124 2.958 3.124 pw_poisson_solve 545 10.5 1.095 1.141 2.167 2.385 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.127663, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2797.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c60ebb48d63dcff4bc85c6fc137b78e381e1214e_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 205321. MP_Allreduce 424 9. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.032 0.040 396.974 396.975 farming_run 1 2.0 396.167 396.175 396.922 396.926 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827128576 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788821 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.220379E+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 2021 21391. MP_Sync 37 MP_Alltoall 77 14602601. 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 0.010 0.029 209.206 209.206 qs_energies 1 2.0 0.001 0.006 208.933 208.973 scf_env_do_scf 1 3.0 0.000 0.000 105.983 105.983 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.008 105.016 rebuild_ks_matrix 4 6.0 0.000 0.000 105.007 105.015 qs_ks_build_kohn_sham_matrix 4 7.0 0.059 0.067 105.007 105.015 hfx_ks_matrix 4 8.0 0.001 0.001 104.604 104.608 integrate_four_center 4 9.0 0.143 0.454 104.603 104.607 mp2_main 1 3.0 0.003 0.022 102.625 102.660 mp2_gpw_main 1 4.0 0.036 0.103 101.661 101.694 integrate_four_center_main 4 10.0 0.123 0.588 96.692 98.396 integrate_four_center_bin 273 11.0 96.568 98.382 96.568 98.382 init_scf_loop 1 4.0 0.000 0.000 91.784 91.784 mp2_ri_gpw_compute_in 1 5.0 0.068 0.102 74.520 75.591 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.028 55.101 mp2_eri_3c_integrate_gpw 91 7.0 0.145 0.162 41.825 46.844 integrate_v_rspace 95 8.0 0.397 0.574 28.217 33.204 pw_transfer 2240 10.6 0.144 0.171 29.957 30.400 ao_to_mo_and_store_B_mult_1 91 7.0 10.521 29.537 10.521 29.537 fft_wrap_pw1pw2 1868 11.4 0.018 0.022 28.966 29.393 mp2_ri_gpw_compute_en 1 5.0 0.074 0.151 26.952 28.712 grid_integrate_task_list 95 9.0 23.487 28.678 23.487 28.678 fft_wrap_pw1pw2_100 730 12.4 1.276 1.463 26.660 27.121 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.842 1.882 25.161 25.173 get_2c_integrals 1 6.0 0.007 0.065 20.378 20.425 compute_2c_integrals 1 7.0 0.003 0.004 19.351 19.357 compute_2c_integrals_loop_lm 1 8.0 0.003 0.016 18.896 19.228 mp2_eri_2c_integrate_gpw 1 9.0 1.743 1.880 18.893 19.227 fft3d_s 1823 13.4 18.456 18.947 18.470 18.960 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.198 14.198 calculate_wavefunction 91 8.0 2.010 2.047 9.738 9.994 mp2_ri_gpw_compute_en_expansio 172 7.0 0.560 0.581 8.852 9.261 potential_pw2rs 186 10.0 0.033 0.035 8.637 9.251 local_gemm 172 8.0 8.293 8.705 8.293 8.705 mp2_ri_gpw_compute_en_comm 22 7.0 0.495 0.513 8.079 8.639 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.255 8.542 calc_potential_gpw 182 9.5 0.002 0.002 7.944 8.231 collocate_single_gaussian 91 10.0 0.017 0.033 7.895 8.096 mp_sendrecv_dm3 2068 8.0 6.120 6.642 6.120 6.642 mp_sync 37 10.5 2.585 6.585 2.585 6.585 mp2_ri_gpw_compute_en_ener 172 7.0 6.342 6.434 6.342 6.434 pw_gather_s 912 13.2 4.897 5.458 4.897 5.458 pw_scatter_s 910 13.7 3.937 4.219 3.937 4.219 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.647239, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1504.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c60ebb48d63dcff4bc85c6fc137b78e381e1214e_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 451.506176E+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 10329 270. MP_Sync 530 MP_Alltoall 2083 588782. MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.045 0.085 53.958 53.968 qs_mol_dyn_low 1 2.0 0.004 0.011 53.272 53.281 qs_forces 11 3.9 0.006 0.027 53.183 53.185 qs_energies 11 4.9 0.001 0.001 51.663 51.677 scf_env_do_scf 11 5.9 0.000 0.001 45.516 45.516 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 43.457 43.457 qs_scf_new_mos 108 7.5 0.000 0.001 33.152 33.439 qs_scf_loop_do_ot 108 8.5 0.000 0.001 33.152 33.439 dbcsr_multiply_generic 2286 12.5 0.095 0.097 32.912 33.324 ot_scf_mini 108 9.5 0.002 0.002 31.453 31.654 multiply_cannon 2286 13.5 0.179 0.186 25.586 27.072 multiply_cannon_loop 2286 14.5 1.458 1.557 24.906 26.431 velocity_verlet 10 3.0 0.001 0.001 25.392 25.392 ot_mini 108 10.5 0.001 0.001 18.903 19.174 qs_ot_get_derivative 108 11.5 0.001 0.001 15.951 16.149 mp_waitall_1 245248 16.5 8.198 14.476 8.198 14.476 multiply_cannon_metrocomm3 54864 15.5 0.067 0.072 5.798 12.541 multiply_cannon_multrec 54864 15.5 4.282 6.618 7.533 11.001 qs_ot_get_p 119 10.4 0.001 0.001 7.995 8.291 rebuild_ks_matrix 119 8.3 0.000 0.000 7.964 8.135 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 7.964 8.134 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.005 7.164 mp_sum_l 7287 12.8 5.363 6.968 5.363 6.968 multiply_cannon_sync_h2d 54864 15.5 5.934 6.816 5.934 6.816 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.646 5.770 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.256 5.704 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.323 5.402 init_scf_run 11 5.9 0.000 0.001 4.795 4.796 scf_env_initial_rho_setup 11 6.9 0.017 0.028 4.795 4.795 sum_up_and_integrate 119 10.3 0.012 0.014 4.566 4.573 integrate_v_rspace 119 11.3 0.002 0.002 4.554 4.562 dbcsr_mm_accdrv_process 76910 16.1 1.118 1.800 3.173 4.540 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.536 4.537 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.337 4.462 calculate_rho_elec 119 8.7 0.012 0.017 4.337 4.461 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.321 4.321 cp_fm_redistribute_end 50 14.0 2.207 4.297 2.212 4.300 cp_fm_diag_elpa_base 50 14.0 2.082 4.189 2.086 4.197 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.889 3.097 apply_single 119 13.6 0.000 0.000 2.889 3.097 calculate_dm_sparse 119 9.5 0.000 0.001 2.953 3.069 multiply_cannon_metrocomm1 54864 15.5 0.053 0.059 1.717 2.974 rs_pw_transfer 974 11.9 0.012 0.013 2.802 2.912 ot_diis_step 108 11.5 0.006 0.006 2.682 2.683 jit_kernel_multiply 13 15.8 1.994 2.680 1.994 2.680 calculate_first_density_matrix 1 7.0 0.000 0.000 2.571 2.592 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.364 2.443 density_rs2pw 119 9.7 0.004 0.005 2.269 2.382 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.299 2.301 acc_transpose_blocks 54864 15.5 0.224 0.246 1.722 2.190 wfi_extrapolate 11 7.9 0.001 0.001 2.132 2.132 grid_integrate_task_list 119 12.3 2.023 2.131 2.023 2.131 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.022 2.074 init_scf_loop 11 6.9 0.000 0.000 2.026 2.027 mp_sum_d 4135 12.0 1.340 1.956 1.340 1.956 potential_pw2rs 119 12.3 0.004 0.004 1.881 1.896 pw_transfer 1439 11.6 0.052 0.059 1.798 1.870 make_m2s 4572 13.5 0.053 0.055 1.742 1.803 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.722 1.796 make_images 4572 14.5 0.133 0.139 1.660 1.720 mp_alltoall_d11v 2130 13.8 1.356 1.546 1.356 1.546 mp_waitany 12084 13.8 1.296 1.436 1.296 1.436 fft3d_ps 1201 14.6 0.376 0.485 1.365 1.434 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.001 1.415 1.432 grid_collocate_task_list 119 9.7 1.356 1.410 1.356 1.410 fft_wrap_pw1pw2_140 487 13.2 0.185 0.201 1.332 1.405 dbcsr_dot_sd 1205 11.9 0.049 0.060 0.717 1.137 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=53.968000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=430.636364, yerr=0.642824 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c60ebb48d63dcff4bc85c6fc137b78e381e1214e_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 487.174144E+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 10306 303. MP_Sync 54 MP_Alltoall 2060 1030466. 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 0.021 0.072 40.372 40.373 qs_mol_dyn_low 1 2.0 0.003 0.006 40.097 40.105 qs_forces 11 3.9 0.004 0.012 40.013 40.014 qs_energies 11 4.9 0.002 0.007 38.296 38.300 scf_env_do_scf 11 5.9 0.001 0.001 32.534 32.534 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 29.983 29.983 dbcsr_multiply_generic 2286 12.5 0.102 0.104 21.709 22.081 qs_scf_new_mos 108 7.5 0.001 0.001 20.519 20.768 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.519 20.767 ot_scf_mini 108 9.5 0.003 0.003 19.602 19.779 velocity_verlet 10 3.0 0.001 0.002 18.595 18.597 multiply_cannon 2286 13.5 0.206 0.215 16.531 18.299 multiply_cannon_loop 2286 14.5 0.905 0.971 15.388 16.941 ot_mini 108 10.5 0.001 0.001 11.995 12.241 mp_waitall_1 200699 16.5 5.836 10.954 5.836 10.954 qs_ot_get_derivative 108 11.5 0.001 0.001 9.494 9.672 multiply_cannon_metrocomm3 27432 15.5 0.066 0.068 4.215 9.535 multiply_cannon_multrec 27432 15.5 1.993 4.321 5.938 8.678 rebuild_ks_matrix 119 8.3 0.000 0.000 7.391 7.531 qs_ks_build_kohn_sham_matrix 119 9.3 0.014 0.024 7.391 7.530 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.534 6.661 dbcsr_mm_accdrv_process 47894 16.0 3.018 4.999 3.875 5.730 qs_ot_get_p 119 10.4 0.001 0.001 4.661 4.897 init_scf_run 11 5.9 0.000 0.001 4.503 4.503 scf_env_initial_rho_setup 11 6.9 0.003 0.017 4.503 4.503 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.605 4.454 sum_up_and_integrate 119 10.3 0.025 0.030 4.313 4.318 integrate_v_rspace 119 11.3 0.002 0.003 4.288 4.296 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.070 4.141 apply_single 119 13.6 0.000 0.000 3.070 4.141 mp_sum_l 7287 12.8 2.160 4.041 2.160 4.041 qs_rho_update_rho_low 119 7.7 0.001 0.003 3.949 3.988 calculate_rho_elec 119 8.7 0.021 0.024 3.949 3.985 qs_ot_p2m_diag 50 11.0 0.009 0.013 3.160 3.181 rs_pw_transfer 974 11.9 0.010 0.011 2.768 3.136 calculate_first_density_matrix 1 7.0 0.003 0.026 2.932 2.937 multiply_cannon_sync_h2d 27432 15.5 2.170 2.829 2.170 2.829 make_m2s 4572 13.5 0.052 0.053 2.566 2.799 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.714 2.715 make_images 4572 14.5 0.201 0.238 2.476 2.706 density_rs2pw 119 9.7 0.004 0.004 2.221 2.595 init_scf_loop 11 6.9 0.001 0.003 2.530 2.531 ot_diis_step 108 11.5 0.013 0.018 2.442 2.446 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.361 2.361 qs_ot_get_derivative_diag 49 12.0 0.001 0.002 2.250 2.340 cp_fm_redistribute_end 50 14.0 1.197 2.332 1.201 2.336 cp_fm_diag_elpa_base 50 14.0 1.101 2.237 1.131 2.277 calculate_dm_sparse 119 9.5 0.000 0.001 2.169 2.246 pw_transfer 1439 11.6 0.065 0.070 2.010 2.041 potential_pw2rs 119 12.3 0.006 0.006 2.025 2.035 jit_kernel_multiply 10 16.1 0.803 2.001 0.803 2.001 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.918 1.950 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.944 1.946 grid_integrate_task_list 119 12.3 1.830 1.924 1.830 1.924 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.687 1.731 prepare_preconditioner 11 7.9 0.000 0.001 1.591 1.618 make_preconditioner 11 8.9 0.000 0.002 1.591 1.618 make_images_data 4572 15.5 0.044 0.050 1.166 1.604 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.489 1.548 acc_transpose_blocks 27432 15.5 0.111 0.117 1.238 1.538 fft_wrap_pw1pw2_140 487 13.2 0.201 0.211 1.480 1.512 wfi_extrapolate 11 7.9 0.001 0.003 1.506 1.506 fft3d_ps 1201 14.6 0.524 0.583 1.471 1.496 hybrid_alltoall_any 4725 16.4 0.051 0.111 1.026 1.489 grid_collocate_task_list 119 9.7 1.272 1.408 1.272 1.408 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.358 1.368 mp_alltoall_d11v 2130 13.8 1.243 1.347 1.243 1.347 mp_allgather_i34 2286 14.5 0.581 1.341 0.581 1.341 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.198 1.247 mp_sum_d 4135 12.0 0.658 1.113 0.658 1.113 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 0.984 0.986 mp_waitany 5720 13.7 0.565 0.966 0.565 0.966 rs_pw_transfer_RS2PW_140 130 11.5 0.138 0.146 0.585 0.950 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.916 0.929 parallel_gemm_fm 81 9.0 0.000 0.000 0.903 0.908 parallel_gemm_fm_cosma 81 10.0 0.903 0.908 0.903 0.908 acc_transpose_blocks_kernels 27432 16.5 0.183 0.272 0.687 0.899 make_images_sizes 4572 15.5 0.005 0.005 0.582 0.840 mp_alltoall_i44 4572 16.5 0.578 0.836 0.578 0.836 rs_pw_transfer_PW2RS_50 119 14.3 0.588 0.605 0.783 0.831 mp_alltoall_z22v 1201 16.6 0.731 0.809 0.731 0.809 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=40.373000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=463.818182, yerr=1.465865 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c60ebb48d63dcff4bc85c6fc137b78e381e1214e_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 519.970816E+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 10155 305. MP_Sync 54 MP_Alltoall 1821 2528239. 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 0.023 0.038 33.848 33.849 qs_mol_dyn_low 1 2.0 0.003 0.003 33.027 33.035 qs_forces 11 3.9 0.003 0.009 32.966 32.967 qs_energies 11 4.9 0.001 0.005 31.350 31.354 scf_env_do_scf 11 5.9 0.001 0.001 26.615 26.615 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 23.977 23.977 dbcsr_multiply_generic 2286 12.5 0.097 0.100 16.684 16.844 velocity_verlet 10 3.0 0.001 0.001 15.745 15.747 qs_scf_new_mos 108 7.5 0.001 0.001 15.466 15.479 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.466 15.478 ot_scf_mini 108 9.5 0.002 0.003 14.729 14.738 multiply_cannon 2286 13.5 0.194 0.201 13.134 14.133 multiply_cannon_loop 2286 14.5 0.635 0.664 12.314 13.373 ot_mini 108 10.5 0.001 0.001 9.015 9.031 qs_ot_get_derivative 108 11.5 0.001 0.001 7.448 7.459 multiply_cannon_multrec 18288 15.5 1.937 2.988 6.738 7.071 rebuild_ks_matrix 119 8.3 0.000 0.000 6.597 6.623 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.019 6.596 6.623 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.829 5.854 dbcsr_mm_accdrv_process 38222 16.0 4.133 5.425 4.717 5.483 sum_up_and_integrate 119 10.3 0.030 0.031 4.141 4.146 integrate_v_rspace 119 11.3 0.003 0.004 4.111 4.118 mp_waitall_1 158411 16.6 2.893 4.082 2.893 4.082 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.619 3.633 calculate_rho_elec 119 8.7 0.031 0.031 3.619 3.632 qs_ot_get_p 119 10.4 0.001 0.001 3.484 3.509 init_scf_run 11 5.9 0.000 0.001 3.504 3.504 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.504 3.504 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.745 3.377 rs_pw_transfer 974 11.9 0.009 0.010 2.522 2.779 init_scf_loop 11 6.9 0.001 0.003 2.621 2.621 multiply_cannon_metrocomm3 18288 15.5 0.043 0.045 1.526 2.458 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.083 2.387 apply_single 119 13.6 0.000 0.000 2.082 2.387 qs_ot_p2m_diag 50 11.0 0.012 0.015 2.373 2.381 density_rs2pw 119 9.7 0.004 0.004 2.091 2.349 calculate_first_density_matrix 1 7.0 0.000 0.002 2.241 2.242 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.098 2.100 make_m2s 4572 13.5 0.044 0.046 1.912 2.035 pw_transfer 1439 11.6 0.066 0.069 1.993 2.007 make_images 4572 14.5 0.191 0.204 1.825 1.947 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.900 1.916 grid_integrate_task_list 119 12.3 1.811 1.909 1.811 1.909 potential_pw2rs 119 12.3 0.007 0.008 1.883 1.889 jit_kernel_multiply 10 16.3 0.532 1.860 0.532 1.860 calculate_dm_sparse 119 9.5 0.000 0.000 1.831 1.843 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.818 1.827 cp_fm_diag_elpa_base 50 14.0 1.794 1.809 1.817 1.825 prepare_preconditioner 11 7.9 0.000 0.000 1.810 1.812 make_preconditioner 11 8.9 0.000 0.001 1.810 1.812 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.657 1.745 qs_ot_get_derivative_diag 49 12.0 0.001 0.002 1.685 1.695 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.680 1.682 mp_sum_l 7287 12.8 1.245 1.617 1.245 1.617 multiply_cannon_sync_h2d 18288 15.5 1.359 1.558 1.359 1.558 ot_diis_step 108 11.5 0.011 0.011 1.541 1.542 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.527 1.534 fft_wrap_pw1pw2_140 487 13.2 0.252 0.258 1.497 1.509 fft3d_ps 1201 14.6 0.532 0.552 1.388 1.402 grid_collocate_task_list 119 9.7 1.246 1.388 1.246 1.388 acc_transpose_blocks 18288 15.5 0.077 0.079 1.255 1.280 wfi_extrapolate 11 7.9 0.001 0.001 1.205 1.205 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.188 1.192 make_images_data 4572 15.5 0.044 0.048 0.866 1.014 qs_energies_init_hamiltonians 11 5.9 0.000 0.002 1.000 1.001 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.932 0.956 hybrid_alltoall_any 4725 16.4 0.055 0.113 0.744 0.925 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.841 0.843 mp_waitany 9880 13.7 0.578 0.842 0.578 0.842 mp_alltoall_d11v 2130 13.8 0.707 0.837 0.707 0.837 rs_pw_transfer_RS2PW_140 130 11.5 0.118 0.121 0.568 0.820 acc_transpose_blocks_kernels 18288 16.5 0.212 0.222 0.799 0.813 mp_alltoall_z22v 1201 16.6 0.703 0.784 0.703 0.784 cp_fm_cholesky_invert 11 10.9 0.769 0.772 0.769 0.772 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.658 0.740 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=33.849000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=494.909091, yerr=1.443137 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c60ebb48d63dcff4bc85c6fc137b78e381e1214e_performance_tests/06/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 114.044384E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3805952 0.0% 0.0% 100.0% average stack size 0.0 0.0 38.6 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 553.721856E+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 10154 346. MP_Sync 54 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.025 0.042 36.102 36.104 qs_mol_dyn_low 1 2.0 0.009 0.028 35.743 35.753 qs_forces 11 3.9 0.002 0.003 35.670 35.676 qs_energies 11 4.9 0.001 0.005 33.920 33.931 scf_env_do_scf 11 5.9 0.001 0.003 28.858 28.859 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 25.353 25.354 dbcsr_multiply_generic 2286 12.5 0.100 0.104 18.239 18.361 velocity_verlet 10 3.0 0.001 0.001 18.213 18.215 qs_scf_new_mos 108 7.5 0.001 0.001 16.638 16.690 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.638 16.689 ot_scf_mini 108 9.5 0.002 0.003 15.693 15.740 multiply_cannon 2286 13.5 0.221 0.230 14.490 14.897 multiply_cannon_loop 2286 14.5 0.946 0.981 13.521 13.906 ot_mini 108 10.5 0.001 0.001 9.392 9.446 multiply_cannon_multrec 27432 15.5 2.359 3.039 8.607 8.986 qs_ot_get_derivative 108 11.5 0.001 0.001 7.564 7.614 dbcsr_mm_accdrv_process 47916 15.9 5.440 7.166 6.159 7.410 rebuild_ks_matrix 119 8.3 0.000 0.000 6.719 6.782 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.015 6.718 6.782 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.956 6.013 sum_up_and_integrate 119 10.3 0.035 0.037 3.929 3.938 integrate_v_rspace 119 11.3 0.003 0.003 3.894 3.904 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.707 3.737 calculate_rho_elec 119 8.7 0.040 0.046 3.706 3.736 qs_ot_get_p 119 10.4 0.001 0.001 3.649 3.715 init_scf_run 11 5.9 0.000 0.001 3.664 3.664 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.664 3.664 init_scf_loop 11 6.9 0.001 0.006 3.483 3.484 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.815 3.214 prepare_preconditioner 11 7.9 0.000 0.000 2.626 2.632 make_preconditioner 11 8.9 0.000 0.001 2.626 2.632 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.232 2.558 mp_waitall_1 137007 16.6 1.957 2.495 1.957 2.495 rs_pw_transfer 974 11.9 0.009 0.010 2.235 2.442 make_m2s 4572 13.5 0.054 0.055 2.296 2.374 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.310 2.319 density_rs2pw 119 9.7 0.004 0.004 2.092 2.273 make_images 4572 14.5 0.271 0.332 2.188 2.265 calculate_first_density_matrix 1 7.0 0.000 0.002 2.239 2.241 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.047 2.193 apply_single 119 13.6 0.000 0.000 2.047 2.193 pw_transfer 1439 11.6 0.065 0.070 2.083 2.124 calculate_dm_sparse 119 9.5 0.000 0.000 2.032 2.088 qs_ot_get_derivative_diag 49 12.0 0.001 0.002 2.016 2.041 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.990 2.035 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.979 1.979 grid_integrate_task_list 119 12.3 1.806 1.920 1.806 1.920 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.805 1.807 ot_diis_step 108 11.5 0.012 0.012 1.788 1.788 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.700 1.710 cp_fm_diag_elpa_base 50 14.0 1.663 1.681 1.698 1.708 potential_pw2rs 119 12.3 0.009 0.009 1.674 1.685 fft_wrap_pw1pw2_140 487 13.2 0.288 0.302 1.607 1.655 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.620 1.636 acc_transpose_blocks 27432 15.5 0.112 0.115 1.472 1.497 jit_kernel_multiply 8 16.0 0.656 1.480 0.656 1.480 fft3d_ps 1201 14.6 0.557 0.612 1.429 1.456 wfi_extrapolate 11 7.9 0.001 0.001 1.365 1.365 grid_collocate_task_list 119 9.7 1.249 1.354 1.249 1.354 multiply_cannon_metrocomm3 27432 15.5 0.038 0.040 0.770 1.343 mp_sum_l 7287 12.8 1.030 1.287 1.030 1.287 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.213 1.222 cp_fm_upper_to_full 72 14.2 0.815 1.162 0.815 1.162 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.139 1.155 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.130 1.136 dbcsr_complete_redistribute 329 12.2 0.119 0.145 0.812 1.082 multiply_cannon_sync_h2d 27432 15.5 0.985 1.081 0.985 1.081 make_images_data 4572 15.5 0.044 0.047 0.910 1.014 hybrid_alltoall_any 4725 16.4 0.062 0.151 0.779 0.936 mp_alltoall_d11v 2130 13.8 0.788 0.905 0.788 0.905 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.792 0.873 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.856 0.862 copy_fm_to_dbcsr 176 11.2 0.001 0.002 0.582 0.850 acc_transpose_blocks_kernels 27432 16.5 0.270 0.279 0.826 0.834 cp_fm_cholesky_invert 11 10.9 0.802 0.806 0.802 0.806 mp_alltoall_z22v 1201 16.6 0.755 0.781 0.755 0.781 mp_alltoall_i22 627 13.8 0.440 0.735 0.440 0.735 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=36.104000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=526.272727, yerr=3.017853 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c60ebb48d63dcff4bc85c6fc137b78e381e1214e_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 598.269952E+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 10154 346. MP_Sync 54 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.067 0.100 29.242 29.243 qs_mol_dyn_low 1 2.0 0.003 0.003 28.977 28.984 qs_forces 11 3.9 0.002 0.002 28.915 28.915 qs_energies 11 4.9 0.001 0.001 27.180 27.183 scf_env_do_scf 11 5.9 0.000 0.001 22.133 22.133 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 19.577 19.577 velocity_verlet 10 3.0 0.001 0.001 14.688 14.691 dbcsr_multiply_generic 2286 12.5 0.094 0.099 12.410 12.481 qs_scf_new_mos 108 7.5 0.001 0.001 11.374 11.395 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.374 11.394 ot_scf_mini 108 9.5 0.002 0.002 10.678 10.704 multiply_cannon 2286 13.5 0.230 0.236 9.756 10.283 multiply_cannon_loop 2286 14.5 0.332 0.343 8.851 9.019 multiply_cannon_multrec 9144 15.5 1.713 1.946 5.950 6.183 rebuild_ks_matrix 119 8.3 0.000 0.000 6.102 6.124 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.101 6.124 ot_mini 108 10.5 0.001 0.001 5.946 5.979 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.428 5.448 qs_ot_get_derivative 108 11.5 0.001 0.001 4.611 4.638 dbcsr_mm_accdrv_process 12550 15.8 3.294 3.957 4.138 4.215 sum_up_and_integrate 119 10.3 0.037 0.041 3.768 3.773 integrate_v_rspace 119 11.3 0.003 0.003 3.731 3.736 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.650 3.657 calculate_rho_elec 119 8.7 0.060 0.061 3.649 3.656 init_scf_run 11 5.9 0.000 0.001 3.621 3.621 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.621 3.621 qs_ot_get_p 119 10.4 0.001 0.001 2.926 2.976 init_scf_loop 11 6.9 0.000 0.000 2.536 2.537 calculate_first_density_matrix 1 7.0 0.000 0.000 2.449 2.450 pw_transfer 1439 11.6 0.066 0.069 2.121 2.132 density_rs2pw 119 9.7 0.004 0.004 1.944 2.084 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.027 2.039 mp_waitall_1 115863 16.7 1.547 2.034 1.547 2.034 make_m2s 4572 13.5 0.034 0.035 1.809 1.977 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.939 1.942 grid_integrate_task_list 119 12.3 1.854 1.926 1.854 1.926 rs_pw_transfer 974 11.9 0.008 0.008 1.792 1.915 make_images 4572 14.5 0.268 0.300 1.718 1.885 jit_kernel_multiply 10 15.6 0.805 1.823 0.805 1.823 prepare_preconditioner 11 7.9 0.000 0.000 1.791 1.796 make_preconditioner 11 8.9 0.000 0.000 1.791 1.796 calculate_dm_sparse 119 9.5 0.000 0.000 1.759 1.778 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.733 1.734 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.680 1.705 fft_wrap_pw1pw2_140 487 13.2 0.364 0.372 1.648 1.661 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.593 1.606 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.520 1.521 potential_pw2rs 119 12.3 0.010 0.011 1.472 1.475 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.453 1.461 cp_fm_diag_elpa_base 50 14.0 1.425 1.443 1.451 1.460 grid_collocate_task_list 119 9.7 1.299 1.413 1.299 1.413 fft3d_ps 1201 14.6 0.561 0.570 1.361 1.371 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.356 1.364 ot_diis_step 108 11.5 0.013 0.013 1.318 1.319 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.214 1.215 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.185 1.215 apply_single 119 13.6 0.000 0.000 1.185 1.214 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.200 1.214 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.156 1.162 wfi_extrapolate 11 7.9 0.001 0.001 1.119 1.119 make_images_data 4572 15.5 0.038 0.042 0.845 1.045 hybrid_alltoall_any 4725 16.4 0.063 0.176 0.796 1.032 cp_fm_cholesky_invert 11 10.9 0.928 0.931 0.928 0.931 acc_transpose_blocks 9144 15.5 0.038 0.039 0.914 0.920 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.867 0.918 mp_alltoall_d11v 2130 13.8 0.781 0.866 0.781 0.866 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.340 0.813 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.798 0.800 multiply_cannon_sync_h2d 9144 15.5 0.711 0.798 0.711 0.798 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.751 0.761 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.666 0.716 mp_alltoall_z22v 1201 16.6 0.674 0.703 0.674 0.703 parallel_gemm_fm 81 9.0 0.000 0.000 0.671 0.672 parallel_gemm_fm_cosma 81 10.0 0.671 0.672 0.671 0.672 acc_transpose_blocks_kernels 9144 16.5 0.117 0.121 0.665 0.668 mp_allgather_i34 2286 14.5 0.235 0.662 0.235 0.662 make_basis_sm 11 9.8 0.003 0.004 0.638 0.639 mp_waitany 5200 13.7 0.467 0.586 0.467 0.586 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=29.243000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=567.818182, yerr=3.809514 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c60ebb48d63dcff4bc85c6fc137b78e381e1214e_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 755.601408E+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 10154 429. MP_Sync 54 MP_Alltoall 1582 7383731. MP_SendRecv 2499 189067. MP_ISendRecv 2499 189067. MP_Wait 6399 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.065 0.096 42.546 42.547 qs_mol_dyn_low 1 2.0 0.003 0.003 42.259 42.267 qs_forces 11 3.9 0.002 0.002 42.197 42.200 qs_energies 11 4.9 0.001 0.001 40.180 40.186 scf_env_do_scf 11 5.9 0.001 0.001 34.235 34.235 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 26.150 26.151 velocity_verlet 10 3.0 0.001 0.001 23.922 23.927 dbcsr_multiply_generic 2286 12.5 0.101 0.104 17.702 17.879 qs_scf_new_mos 108 7.5 0.001 0.001 16.166 16.257 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.165 16.256 ot_scf_mini 108 9.5 0.002 0.002 15.066 15.162 multiply_cannon 2286 13.5 0.305 0.310 13.687 14.684 multiply_cannon_loop 2286 14.5 0.345 0.351 12.392 13.360 ot_mini 108 10.5 0.001 0.001 8.933 9.048 multiply_cannon_multrec 9144 15.5 3.410 4.779 8.574 8.652 init_scf_loop 11 6.9 0.000 0.000 8.057 8.058 rebuild_ks_matrix 119 8.3 0.000 0.000 7.397 7.539 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.396 7.539 prepare_preconditioner 11 7.9 0.000 0.000 7.085 7.099 make_preconditioner 11 8.9 0.000 0.000 7.085 7.099 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.568 6.978 qs_ot_get_derivative 108 11.5 0.001 0.001 6.863 6.961 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.669 6.797 dbcsr_mm_accdrv_process 12550 15.8 4.218 5.629 5.044 6.353 cp_fm_upper_to_full 72 14.2 3.238 4.698 3.238 4.698 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.372 4.379 calculate_rho_elec 119 8.7 0.118 0.121 4.372 4.379 sum_up_and_integrate 119 10.3 0.064 0.066 4.091 4.097 integrate_v_rspace 119 11.3 0.003 0.003 4.027 4.032 init_scf_run 11 5.9 0.000 0.001 3.911 3.911 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.910 3.910 mp_waitall_1 94719 16.7 2.500 3.551 2.500 3.551 qs_ot_get_p 119 10.4 0.001 0.001 3.352 3.488 dbcsr_complete_redistribute 329 12.2 0.287 0.290 2.079 2.951 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.456 2.909 pw_transfer 1439 11.6 0.069 0.069 2.736 2.743 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.638 2.645 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.752 2.600 make_m2s 4572 13.5 0.037 0.037 2.364 2.566 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.260 2.494 apply_single 119 13.6 0.000 0.000 2.259 2.494 make_images 4572 14.5 0.352 0.384 2.242 2.444 mp_alltoall_i22 627 13.8 1.555 2.437 1.555 2.437 calculate_first_density_matrix 1 7.0 0.000 0.000 2.401 2.405 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.512 2.352 density_rs2pw 119 9.7 0.004 0.004 2.263 2.289 multiply_cannon_metrocomm3 9144 15.5 0.019 0.020 1.362 2.276 calculate_dm_sparse 119 9.5 0.000 0.000 2.227 2.245 fft_wrap_pw1pw2_140 487 13.2 0.616 0.627 2.207 2.215 grid_integrate_task_list 119 12.3 2.068 2.091 2.068 2.091 ot_diis_step 108 11.5 0.014 0.015 2.039 2.039 qs_ot_p2m_diag 50 11.0 0.042 0.043 2.031 2.032 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.940 1.941 mp_sum_l 7287 12.8 1.106 1.799 1.106 1.799 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.728 1.777 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.764 1.767 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.755 1.755 fft3d_ps 1201 14.6 0.595 0.606 1.655 1.658 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.614 1.649 rs_pw_transfer 974 11.9 0.009 0.009 1.565 1.590 grid_collocate_task_list 119 9.7 1.514 1.530 1.514 1.530 cp_fm_cholesky_invert 11 10.9 1.524 1.528 1.524 1.528 potential_pw2rs 119 12.3 0.014 0.014 1.453 1.455 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.453 1.453 cp_fm_diag_elpa_base 50 14.0 1.304 1.360 1.451 1.451 hybrid_alltoall_any 4725 16.4 0.088 0.149 1.166 1.448 wfi_extrapolate 11 7.9 0.001 0.001 1.435 1.435 make_images_data 4572 15.5 0.042 0.045 1.134 1.389 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.277 1.286 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.156 1.180 mp_alltoall_d11v 2130 13.8 1.132 1.159 1.132 1.159 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.098 1.108 multiply_cannon_sync_h2d 9144 15.5 1.037 1.041 1.037 1.041 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.988 1.022 jit_kernel_multiply 6 15.6 0.798 1.013 0.798 1.013 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.941 0.954 mp_alltoall_z22v 1201 16.6 0.925 0.946 0.925 0.946 qs_create_task_list 11 7.9 0.000 0.000 0.929 0.943 generate_qs_task_list 11 8.9 0.367 0.385 0.929 0.943 acc_transpose_blocks 9144 15.5 0.039 0.040 0.908 0.913 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=42.547000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=709.545455, yerr=13.956840 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c60ebb48d63dcff4bc85c6fc137b78e381e1214e_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 501.559296E+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 9776 488. MP_Sync 52 MP_Alltoall 1938 2286762. 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 0.025 0.103 82.298 82.299 qs_mol_dyn_low 1 2.0 0.003 0.003 81.924 81.934 qs_forces 11 3.9 0.003 0.009 81.852 81.854 qs_energies 11 4.9 0.001 0.007 78.951 78.962 scf_env_do_scf 11 5.9 0.001 0.007 69.660 69.662 scf_env_do_scf_inner_loop 99 6.5 0.003 0.012 63.943 63.945 dbcsr_multiply_generic 2055 12.4 0.110 0.112 50.198 50.453 qs_scf_new_mos 99 7.5 0.000 0.001 46.320 46.427 qs_scf_loop_do_ot 99 8.5 0.001 0.001 46.320 46.427 ot_scf_mini 99 9.5 0.002 0.003 43.992 44.118 multiply_cannon 2055 13.4 0.177 0.182 42.054 42.900 velocity_verlet 10 3.0 0.001 0.001 42.859 42.860 multiply_cannon_loop 2055 14.4 1.515 1.550 41.063 41.901 ot_mini 99 10.5 0.001 0.001 25.902 26.023 qs_ot_get_derivative 99 11.5 0.001 0.001 19.198 19.328 multiply_cannon_multrec 49320 15.4 12.496 13.268 17.427 18.178 rebuild_ks_matrix 110 8.3 0.000 0.000 14.814 14.963 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.813 14.962 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.967 13.097 mp_waitall_1 220248 16.4 10.443 11.151 10.443 11.151 multiply_cannon_sync_h2d 49320 15.4 10.379 11.027 10.379 11.027 qs_ot_get_p 110 10.4 0.001 0.001 9.847 9.955 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.329 7.862 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.215 7.801 apply_single 110 13.6 0.000 0.001 7.215 7.800 multiply_cannon_metrocomm3 49320 15.4 0.078 0.081 6.199 7.289 sum_up_and_integrate 110 10.3 0.036 0.042 7.229 7.241 integrate_v_rspace 110 11.3 0.003 0.004 7.193 7.213 init_scf_run 11 5.9 0.000 0.001 7.157 7.157 scf_env_initial_rho_setup 11 6.9 0.000 0.001 7.157 7.157 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.805 6.911 calculate_rho_elec 110 8.6 0.021 0.026 6.804 6.910 qs_ot_p2m_diag 48 11.0 0.012 0.018 6.602 6.623 ot_diis_step 99 11.5 0.006 0.006 6.543 6.544 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.797 5.797 init_scf_loop 11 6.9 0.001 0.003 5.684 5.687 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.338 5.389 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.220 5.253 cp_fm_diag_elpa_base 48 14.0 5.206 5.240 5.218 5.251 dbcsr_mm_accdrv_process 87628 16.1 1.927 1.998 4.806 5.064 mp_sum_l 6594 12.7 3.762 4.463 3.762 4.463 rs_pw_transfer 902 11.9 0.011 0.013 3.863 4.414 wfi_extrapolate 11 7.9 0.001 0.001 4.080 4.080 make_m2s 4110 13.4 0.061 0.065 3.918 4.040 density_rs2pw 110 9.6 0.004 0.005 3.533 3.997 make_images 4110 14.4 0.178 0.192 3.822 3.945 calculate_dm_sparse 110 9.5 0.001 0.001 3.770 3.862 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.557 3.562 grid_integrate_task_list 110 12.3 3.249 3.431 3.249 3.431 multiply_cannon_metrocomm1 49320 15.4 0.062 0.064 2.322 3.405 prepare_preconditioner 11 7.9 0.000 0.000 3.325 3.349 make_preconditioner 11 8.9 0.000 0.001 3.325 3.349 pw_transfer 1331 11.6 0.055 0.067 3.251 3.333 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.233 3.300 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.162 3.244 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.125 3.169 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.052 3.094 calculate_first_density_matrix 1 7.0 0.000 0.002 2.971 2.977 fft_wrap_pw1pw2_140 451 13.1 0.454 0.503 2.689 2.777 potential_pw2rs 110 12.3 0.006 0.006 2.731 2.758 mp_alltoall_d11v 2046 13.8 2.076 2.736 2.076 2.736 jit_kernel_multiply 13 15.9 2.602 2.643 2.602 2.643 mp_waitany 14300 13.8 1.856 2.423 1.856 2.423 fft3d_ps 1111 14.6 0.794 0.879 2.356 2.416 grid_collocate_task_list 110 9.6 2.162 2.399 2.162 2.399 acc_transpose_blocks 49320 15.4 0.209 0.218 2.085 2.145 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.979 2.000 mp_sum_d 3889 11.9 1.379 1.981 1.379 1.981 make_images_data 4110 15.4 0.042 0.046 1.786 1.918 cp_fm_cholesky_invert 11 10.9 1.856 1.860 1.856 1.860 hybrid_alltoall_any 4261 16.3 0.082 0.480 1.552 1.815 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.720 1.747 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=82.299000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.636364, yerr=2.804954 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c60ebb48d63dcff4bc85c6fc137b78e381e1214e_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 590.475264E+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 9775 566. MP_Sync 52 MP_Alltoall 1717 2523422. 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 0.027 0.038 70.500 70.503 qs_mol_dyn_low 1 2.0 0.003 0.003 70.083 70.092 qs_forces 11 3.9 0.003 0.003 70.017 70.018 qs_energies 11 4.9 0.001 0.005 66.645 66.649 scf_env_do_scf 11 5.9 0.002 0.015 57.483 57.487 scf_env_do_scf_inner_loop 99 6.5 0.004 0.014 49.816 49.818 dbcsr_multiply_generic 2055 12.4 0.117 0.122 37.997 38.159 velocity_verlet 10 3.0 0.001 0.002 36.612 36.613 qs_scf_new_mos 99 7.5 0.001 0.001 33.263 33.384 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.263 33.383 multiply_cannon 2055 13.4 0.220 0.245 31.239 32.490 ot_scf_mini 99 9.5 0.004 0.017 31.591 31.705 multiply_cannon_loop 2055 14.4 0.922 0.947 29.944 30.936 ot_mini 99 10.5 0.001 0.001 18.348 18.475 multiply_cannon_multrec 24660 15.4 7.596 9.069 13.965 15.571 rebuild_ks_matrix 110 8.3 0.000 0.000 13.911 14.061 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 13.911 14.061 qs_ot_get_derivative 99 11.5 0.001 0.001 12.505 12.625 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.226 12.364 mp_waitall_1 176588 16.5 7.756 10.504 7.756 10.504 multiply_cannon_sync_h2d 24660 15.4 7.019 8.107 7.019 8.107 multiply_cannon_metrocomm3 24660 15.4 0.067 0.070 5.234 7.783 init_scf_loop 11 6.9 0.000 0.003 7.631 7.632 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.551 7.344 apply_single 110 13.6 0.000 0.001 6.551 7.344 init_scf_run 11 5.9 0.000 0.001 6.781 6.782 scf_env_initial_rho_setup 11 6.9 0.001 0.002 6.781 6.782 sum_up_and_integrate 110 10.3 0.053 0.060 6.724 6.736 qs_ot_get_p 110 10.4 0.001 0.001 6.565 6.702 integrate_v_rspace 110 11.3 0.002 0.003 6.671 6.682 dbcsr_mm_accdrv_process 52282 16.1 4.559 5.588 6.207 6.538 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.331 6.342 calculate_rho_elec 110 8.6 0.040 0.048 6.330 6.342 ot_diis_step 99 11.5 0.010 0.011 5.790 5.790 prepare_preconditioner 11 7.9 0.000 0.000 5.586 5.610 make_preconditioner 11 8.9 0.000 0.001 5.586 5.610 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.734 5.478 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.154 5.302 make_m2s 4110 13.4 0.057 0.060 4.240 4.694 qs_ot_p2m_diag 48 11.0 0.028 0.043 4.634 4.656 make_images 4110 14.4 0.401 0.445 4.127 4.577 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.197 4.197 density_rs2pw 110 9.6 0.004 0.005 3.378 3.922 pw_transfer 1331 11.6 0.067 0.077 3.719 3.864 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.612 3.758 rs_pw_transfer 902 11.9 0.012 0.014 3.043 3.624 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.595 3.608 cp_fm_diag_elpa_base 48 14.0 3.548 3.563 3.593 3.605 wfi_extrapolate 11 7.9 0.001 0.001 3.559 3.559 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.381 3.383 grid_integrate_task_list 110 12.3 3.147 3.355 3.147 3.355 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.215 3.279 fft_wrap_pw1pw2_140 451 13.1 0.519 0.534 3.055 3.200 calculate_first_density_matrix 1 7.0 0.000 0.002 3.122 3.125 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.031 3.086 calculate_dm_sparse 110 9.5 0.001 0.001 3.008 3.034 make_images_data 4110 15.4 0.046 0.050 2.396 2.838 hybrid_alltoall_any 4261 16.3 0.102 0.441 2.083 2.809 fft3d_ps 1111 14.6 1.110 1.332 2.587 2.744 cp_fm_cholesky_invert 11 10.9 2.636 2.643 2.636 2.643 mp_sum_l 6594 12.7 1.907 2.543 1.907 2.543 grid_collocate_task_list 110 9.6 2.100 2.532 2.100 2.532 potential_pw2rs 110 12.3 0.008 0.009 2.479 2.494 mp_alltoall_d11v 2046 13.8 1.816 2.167 1.816 2.167 jit_kernel_multiply 10 16.2 1.288 2.065 1.288 2.065 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.975 2.000 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 1.865 1.866 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.842 1.857 mp_waitany 10164 13.8 1.250 1.775 1.250 1.775 multiply_cannon_metrocomm4 22605 15.4 0.077 0.081 0.792 1.772 mp_allgather_i34 2055 14.4 0.623 1.685 0.623 1.685 acc_transpose_blocks 24660 15.4 0.113 0.116 1.566 1.586 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.570 1.579 mp_irecv_dv 57340 16.2 0.665 1.554 0.665 1.554 rs_pw_transfer_RS2PW_140 121 11.5 0.205 0.215 0.981 1.551 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.370 1.477 mp_sum_d 3889 11.9 0.934 1.422 0.934 1.422 make_basis_sm 11 9.8 0.000 0.001 1.417 1.420 dbcsr_complete_redistribute 325 12.2 0.231 0.296 1.147 1.416 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=70.503000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=555.181818, yerr=6.899515 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c60ebb48d63dcff4bc85c6fc137b78e381e1214e_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 656.642048E+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 9774 562. MP_Sync 52 MP_Alltoall 1496 4511006. MP_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.039 0.091 63.109 63.114 qs_mol_dyn_low 1 2.0 0.003 0.004 62.621 62.630 qs_forces 11 3.9 0.003 0.003 62.487 62.488 qs_energies 11 4.9 0.007 0.050 59.197 59.200 scf_env_do_scf 11 5.9 0.001 0.002 50.501 50.501 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 41.530 41.530 velocity_verlet 10 3.0 0.001 0.001 33.867 33.871 dbcsr_multiply_generic 2055 12.4 0.131 0.137 29.186 29.547 qs_scf_new_mos 99 7.5 0.001 0.001 26.042 26.162 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.041 26.162 ot_scf_mini 99 9.5 0.002 0.003 24.781 24.912 multiply_cannon 2055 13.4 0.212 0.224 22.446 23.788 multiply_cannon_loop 2055 14.4 0.617 0.632 21.209 22.277 ot_mini 99 10.5 0.001 0.001 14.027 14.164 rebuild_ks_matrix 110 8.3 0.000 0.000 12.673 12.853 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.672 12.852 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.137 11.302 mp_waitall_1 139946 16.5 7.701 11.055 7.701 11.055 multiply_cannon_multrec 16440 15.4 3.688 4.336 9.415 9.992 qs_ot_get_derivative 99 11.5 0.001 0.001 9.409 9.543 init_scf_loop 11 6.9 0.001 0.002 8.933 8.934 multiply_cannon_metrocomm3 16440 15.4 0.042 0.045 4.859 8.007 prepare_preconditioner 11 7.9 0.000 0.000 7.123 7.143 make_preconditioner 11 8.9 0.000 0.002 7.122 7.143 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.449 6.825 sum_up_and_integrate 110 10.3 0.060 0.061 6.735 6.749 integrate_v_rspace 110 11.3 0.003 0.003 6.674 6.688 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.215 6.223 calculate_rho_elec 110 8.6 0.059 0.059 6.214 6.223 init_scf_run 11 5.9 0.000 0.001 6.149 6.149 scf_env_initial_rho_setup 11 6.9 0.001 0.002 6.148 6.149 qs_ot_get_p 110 10.4 0.001 0.001 5.848 6.005 dbcsr_mm_accdrv_process 34862 16.1 4.645 5.126 5.584 5.791 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.116 5.592 apply_single 110 13.6 0.000 0.000 5.116 5.592 make_m2s 4110 13.4 0.050 0.051 4.347 4.712 make_images 4110 14.4 0.394 0.510 4.229 4.596 ot_diis_step 99 11.5 0.011 0.011 4.579 4.579 density_rs2pw 110 9.6 0.004 0.005 3.245 4.531 multiply_cannon_sync_h2d 16440 15.4 3.598 4.165 3.598 4.165 rs_pw_transfer 902 11.9 0.010 0.011 2.937 4.158 qs_ot_p2m_diag 48 11.0 0.042 0.044 4.114 4.120 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.200 3.841 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.767 3.768 pw_transfer 1331 11.6 0.066 0.074 3.758 3.765 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.650 3.661 grid_integrate_task_list 110 12.3 3.181 3.405 3.181 3.405 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.197 3.206 cp_fm_diag_elpa_base 48 14.0 3.126 3.156 3.195 3.205 fft_wrap_pw1pw2_140 451 13.1 0.635 0.644 3.124 3.136 wfi_extrapolate 11 7.9 0.001 0.001 3.078 3.078 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.034 3.035 make_images_data 4110 15.4 0.042 0.046 2.549 3.000 calculate_first_density_matrix 1 7.0 0.000 0.002 2.973 2.975 cp_fm_cholesky_invert 11 10.9 2.930 2.938 2.930 2.938 hybrid_alltoall_any 4261 16.3 0.106 0.375 2.247 2.910 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.622 2.691 calculate_dm_sparse 110 9.5 0.001 0.001 2.538 2.570 multiply_cannon_metrocomm4 14385 15.4 0.044 0.048 0.939 2.568 mp_waitany 17072 13.8 1.302 2.563 1.302 2.563 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.452 2.529 grid_collocate_task_list 110 9.6 2.123 2.527 2.123 2.527 mp_sum_l 6594 12.7 1.792 2.498 1.792 2.498 fft3d_ps 1111 14.6 1.090 1.102 2.475 2.485 mp_irecv_dv 48980 15.7 0.869 2.442 0.869 2.442 potential_pw2rs 110 12.3 0.011 0.011 2.425 2.437 mp_alltoall_d11v 2046 13.8 1.883 2.262 1.883 2.262 rs_pw_transfer_RS2PW_140 121 11.5 0.174 0.178 0.984 2.197 dbcsr_complete_redistribute 325 12.2 0.334 0.417 1.571 2.064 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 2.047 2.048 jit_kernel_multiply 10 16.2 0.543 1.911 0.543 1.911 cp_fm_upper_to_full 70 14.2 1.382 1.792 1.382 1.792 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.732 1.749 mp_allgather_i34 2055 14.4 0.560 1.705 0.560 1.705 cp_fm_cholesky_decompose 22 10.9 1.652 1.669 1.652 1.669 copy_fm_to_dbcsr 174 11.2 0.001 0.003 1.068 1.568 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.359 1.501 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.470 1.483 rs_gather_matrices 110 12.3 0.234 0.263 0.985 1.325 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.304 1.317 acc_transpose_blocks 16440 15.4 0.074 0.077 1.289 1.316 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=63.114000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=623.090909, yerr=7.597303 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c60ebb48d63dcff4bc85c6fc137b78e381e1214e_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 729.964544E+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 9774 603. MP_Sync 52 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.032 67.657 67.657 qs_mol_dyn_low 1 2.0 0.003 0.003 67.270 67.279 qs_forces 11 3.9 0.003 0.003 66.780 66.782 qs_energies 11 4.9 0.001 0.001 63.279 63.283 scf_env_do_scf 11 5.9 0.000 0.001 54.637 54.639 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 42.717 42.717 velocity_verlet 10 3.0 0.001 0.001 38.340 38.359 dbcsr_multiply_generic 2055 12.4 0.116 0.121 29.775 30.003 qs_scf_new_mos 99 7.5 0.001 0.001 27.328 27.433 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.328 27.433 ot_scf_mini 99 9.5 0.003 0.003 25.681 25.770 multiply_cannon 2055 13.4 0.244 0.263 22.514 23.789 multiply_cannon_loop 2055 14.4 0.889 0.914 21.088 21.756 ot_mini 99 10.5 0.001 0.001 14.391 14.506 multiply_cannon_multrec 24660 15.4 4.197 6.877 12.648 13.729 rebuild_ks_matrix 110 8.3 0.000 0.000 12.434 12.541 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.434 12.540 init_scf_loop 11 6.9 0.000 0.000 11.875 11.875 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.953 11.048 qs_ot_get_derivative 99 11.5 0.001 0.001 10.199 10.295 prepare_preconditioner 11 7.9 0.000 0.000 10.103 10.118 make_preconditioner 11 8.9 0.000 0.000 10.103 10.118 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.345 9.803 dbcsr_mm_accdrv_process 52304 16.0 6.822 8.644 8.305 9.221 sum_up_and_integrate 110 10.3 0.067 0.070 6.723 6.737 integrate_v_rspace 110 11.3 0.003 0.003 6.655 6.669 mp_waitall_1 121746 16.5 4.477 6.429 4.477 6.429 qs_ot_get_p 110 10.4 0.001 0.001 6.257 6.374 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.304 6.316 calculate_rho_elec 110 8.6 0.078 0.081 6.304 6.316 init_scf_run 11 5.9 0.000 0.001 5.814 5.814 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.813 5.814 make_m2s 4110 13.4 0.059 0.060 5.558 5.813 make_images 4110 14.4 0.576 0.695 5.416 5.666 cp_fm_upper_to_full 70 14.2 3.361 4.861 3.361 4.861 qs_ot_p2m_diag 48 11.0 0.054 0.062 4.389 4.402 ot_diis_step 99 11.5 0.011 0.011 4.151 4.152 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.051 4.143 apply_single 110 13.6 0.000 0.000 4.051 4.142 pw_transfer 1331 11.6 0.066 0.075 3.907 3.938 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.905 3.905 dbcsr_complete_redistribute 325 12.2 0.409 0.454 2.723 3.855 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.801 3.836 density_rs2pw 110 9.6 0.004 0.004 3.174 3.700 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.407 3.459 grid_integrate_task_list 110 12.3 3.263 3.448 3.263 3.448 multiply_cannon_sync_h2d 24660 15.4 3.153 3.296 3.153 3.296 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.288 3.296 cp_fm_diag_elpa_base 48 14.0 3.133 3.193 3.285 3.293 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.159 3.278 multiply_cannon_metrocomm3 24660 15.4 0.036 0.037 1.437 3.276 fft_wrap_pw1pw2_140 451 13.1 0.666 0.687 3.205 3.241 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.116 3.165 rs_pw_transfer 902 11.9 0.010 0.011 2.563 3.130 wfi_extrapolate 11 7.9 0.001 0.001 3.098 3.098 make_images_data 4110 15.4 0.045 0.048 2.801 3.078 hybrid_alltoall_any 4261 16.3 0.120 0.457 2.406 3.007 calculate_dm_sparse 110 9.5 0.001 0.001 2.949 2.981 cp_fm_cholesky_invert 11 10.9 2.865 2.876 2.865 2.876 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.874 2.876 mp_alltoall_i22 605 13.7 1.659 2.851 1.659 2.851 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.749 2.843 calculate_first_density_matrix 1 7.0 0.000 0.000 2.614 2.616 fft3d_ps 1111 14.6 1.088 1.117 2.566 2.580 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.477 2.515 grid_collocate_task_list 110 9.6 2.219 2.478 2.219 2.478 mp_alltoall_d11v 2046 13.8 1.895 2.461 1.895 2.461 potential_pw2rs 110 12.3 0.012 0.013 2.320 2.332 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 2.314 2.315 jit_kernel_multiply 12 15.8 1.148 2.232 1.148 2.232 cp_fm_cholesky_decompose 22 10.9 1.745 1.796 1.745 1.796 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.743 1.772 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.726 1.737 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.600 1.704 multiply_cannon_metrocomm4 20550 15.4 0.059 0.063 0.866 1.655 mp_sum_l 6594 12.7 1.015 1.647 1.015 1.647 mp_allgather_i34 2055 14.4 0.528 1.631 0.528 1.631 mp_waitany 13376 13.8 1.085 1.607 1.085 1.607 acc_transpose_blocks 24660 15.4 0.105 0.107 1.561 1.585 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.549 1.565 mp_irecv_dv 62702 16.1 0.766 1.492 0.766 1.492 rs_gather_matrices 110 12.3 0.267 0.297 0.987 1.470 rs_pw_transfer_RS2PW_140 121 11.5 0.193 0.214 0.848 1.416 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.265 1.373 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=67.657000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=689.272727, yerr=8.236183 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c60ebb48d63dcff4bc85c6fc137b78e381e1214e_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 828.264448E+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 9774 644. MP_Sync 52 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.055 0.075 58.918 58.919 qs_mol_dyn_low 1 2.0 0.003 0.003 58.127 58.542 qs_forces 11 3.9 0.003 0.004 57.971 57.972 qs_energies 11 4.9 0.001 0.001 54.203 54.207 scf_env_do_scf 11 5.9 0.000 0.001 45.071 45.072 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 36.862 36.863 velocity_verlet 10 3.0 0.001 0.001 32.401 32.407 dbcsr_multiply_generic 2055 12.4 0.107 0.111 23.776 23.944 qs_scf_new_mos 99 7.5 0.001 0.001 21.399 21.461 qs_scf_loop_do_ot 99 8.5 0.001 0.001 21.399 21.460 ot_scf_mini 99 9.5 0.002 0.002 20.110 20.149 multiply_cannon 2055 13.4 0.246 0.258 18.006 19.304 multiply_cannon_loop 2055 14.4 0.322 0.334 16.607 17.065 rebuild_ks_matrix 110 8.3 0.000 0.000 12.236 12.271 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 12.236 12.270 ot_mini 99 10.5 0.001 0.001 10.877 10.919 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.856 10.888 multiply_cannon_multrec 8220 15.4 3.219 4.751 7.688 8.804 init_scf_loop 11 6.9 0.000 0.000 8.159 8.161 mp_waitall_1 103326 16.6 6.546 8.103 6.546 8.103 qs_ot_get_derivative 99 11.5 0.001 0.001 6.958 7.000 sum_up_and_integrate 110 10.3 0.080 0.081 6.738 6.750 integrate_v_rspace 110 11.3 0.003 0.003 6.658 6.671 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.482 6.498 calculate_rho_elec 110 8.6 0.114 0.114 6.482 6.498 prepare_preconditioner 11 7.9 0.000 0.000 6.450 6.454 make_preconditioner 11 8.9 0.000 0.000 6.450 6.454 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.022 6.109 init_scf_run 11 5.9 0.000 0.001 5.984 5.985 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.984 5.984 qs_ot_get_p 110 10.4 0.001 0.001 5.216 5.270 dbcsr_mm_accdrv_process 17442 15.9 2.871 3.828 4.338 5.232 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 3.231 4.655 make_m2s 4110 13.4 0.039 0.040 4.335 4.560 make_images 4110 14.4 0.636 0.692 4.202 4.424 pw_transfer 1331 11.6 0.066 0.071 4.127 4.134 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.020 4.030 ot_diis_step 99 11.5 0.012 0.012 3.893 3.893 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.780 3.830 apply_single 110 13.6 0.000 0.000 3.780 3.830 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.772 3.777 grid_integrate_task_list 110 12.3 3.374 3.564 3.374 3.564 density_rs2pw 110 9.6 0.004 0.004 3.159 3.530 fft_wrap_pw1pw2_140 451 13.1 0.833 0.844 3.415 3.439 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.437 3.437 cp_fm_cholesky_invert 11 10.9 3.230 3.236 3.230 3.236 multiply_cannon_sync_h2d 8220 15.4 2.893 3.011 2.893 3.011 calculate_first_density_matrix 1 7.0 0.000 0.000 3.008 3.009 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.869 2.876 cp_fm_diag_elpa_base 48 14.0 2.811 2.837 2.867 2.874 wfi_extrapolate 11 7.9 0.001 0.001 2.861 2.861 make_images_data 4110 15.4 0.037 0.043 2.468 2.855 hybrid_alltoall_any 4261 16.3 0.199 0.858 2.385 2.815 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.789 2.791 rs_pw_transfer 902 11.9 0.010 0.010 2.330 2.729 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.682 2.682 calculate_dm_sparse 110 9.5 0.001 0.001 2.573 2.613 fft3d_ps 1111 14.6 1.140 1.180 2.565 2.587 grid_collocate_task_list 110 9.6 2.315 2.574 2.315 2.574 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.424 2.439 potential_pw2rs 110 12.3 0.015 0.015 2.237 2.242 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.976 2.003 mp_alltoall_d11v 2046 13.8 1.746 1.997 1.746 1.997 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.955 1.985 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.768 1.980 cp_fm_cholesky_decompose 22 10.9 1.747 1.761 1.747 1.761 jit_kernel_multiply 9 15.7 1.153 1.737 1.153 1.737 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.713 1.718 mp_allgather_i34 2055 14.4 0.582 1.715 0.582 1.715 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.521 1.644 dbcsr_complete_redistribute 325 12.2 0.553 0.572 1.522 1.620 mp_waitany 9240 13.8 1.127 1.551 1.127 1.551 multiply_cannon_metrocomm1 8220 15.4 0.021 0.022 0.944 1.526 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.493 1.502 qs_create_task_list 11 7.9 0.000 0.001 1.223 1.328 generate_qs_task_list 11 8.9 0.375 0.443 1.223 1.327 multiply_cannon_metrocomm4 6165 15.4 0.017 0.019 0.482 1.249 mp_alltoall_z22v 1111 16.6 1.193 1.223 1.193 1.223 rs_pw_transfer_RS2PW_140 121 11.5 0.162 0.168 0.825 1.221 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.183 1.215 mp_irecv_dv 24056 15.7 0.459 1.211 0.459 1.211 make_basis_sm 11 9.8 0.001 0.001 1.194 1.196 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=58.919000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=780.545455, yerr=11.105802 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c60ebb48d63dcff4bc85c6fc137b78e381e1214e_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 1.375908E+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 9752 812. MP_Sync 52 MP_Alltoall 1474 16505187. MP_SendRecv 2310 360267. MP_ISendRecv 2310 360267. MP_Wait 5214 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.040 0.056 87.402 87.402 qs_mol_dyn_low 1 2.0 0.004 0.017 87.017 87.025 qs_forces 11 3.9 0.007 0.010 86.939 86.942 qs_energies 11 4.9 0.001 0.001 82.707 82.711 scf_env_do_scf 11 5.9 0.001 0.001 72.445 72.445 velocity_verlet 10 3.0 0.001 0.001 55.906 55.913 scf_env_do_scf_inner_loop 99 6.5 0.003 0.006 43.925 43.926 init_scf_loop 11 6.9 0.000 0.000 28.447 28.449 dbcsr_multiply_generic 2055 12.4 0.121 0.126 28.234 28.302 prepare_preconditioner 11 7.9 0.000 0.000 26.384 26.390 make_preconditioner 11 8.9 0.000 0.000 26.384 26.390 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.520 25.848 qs_scf_new_mos 99 7.5 0.001 0.001 25.643 25.679 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.643 25.679 ot_scf_mini 99 9.5 0.002 0.002 23.899 23.920 multiply_cannon 2055 13.4 0.349 0.373 21.347 22.177 multiply_cannon_loop 2055 14.4 0.341 0.344 19.538 19.870 cp_fm_upper_to_full 70 14.2 12.714 18.290 12.714 18.290 rebuild_ks_matrix 110 8.3 0.000 0.001 14.134 14.162 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 14.134 14.161 ot_mini 99 10.5 0.001 0.001 13.277 13.299 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.773 12.798 dbcsr_complete_redistribute 325 12.2 1.017 1.060 7.446 10.710 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.415 9.680 multiply_cannon_multrec 8220 15.4 4.363 4.565 9.545 9.650 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.848 9.076 qs_ot_get_derivative 99 11.5 0.001 0.001 8.815 8.837 mp_alltoall_i22 605 13.7 5.457 8.731 5.457 8.731 mp_waitall_1 84994 16.7 7.294 8.077 7.294 8.077 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.800 7.838 calculate_rho_elec 110 8.6 0.225 0.225 7.800 7.837 sum_up_and_integrate 110 10.3 0.150 0.152 7.340 7.355 integrate_v_rspace 110 11.3 0.004 0.004 7.190 7.205 init_scf_run 11 5.9 0.000 0.001 5.900 5.901 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.900 5.900 make_m2s 4110 13.4 0.042 0.043 5.228 5.720 make_images 4110 14.4 0.879 0.928 5.038 5.530 qs_ot_get_p 110 10.4 0.001 0.001 5.487 5.515 dbcsr_mm_accdrv_process 11614 15.7 3.187 3.322 5.039 5.273 cp_fm_cholesky_invert 11 10.9 5.253 5.257 5.253 5.257 pw_transfer 1331 11.6 0.075 0.076 5.094 5.099 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.594 5.060 apply_single 110 13.6 0.000 0.000 4.593 5.059 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 4.977 4.982 multiply_cannon_metrocomm3 8220 15.4 0.018 0.019 4.544 4.868 ot_diis_step 99 11.5 0.015 0.016 4.446 4.447 fft_wrap_pw1pw2_140 451 13.1 1.335 1.341 4.347 4.356 qs_ot_p2m_diag 48 11.0 0.150 0.155 3.985 3.992 multiply_cannon_sync_h2d 8220 15.4 3.944 3.949 3.944 3.949 density_rs2pw 110 9.6 0.004 0.004 3.785 3.821 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 3.756 3.758 grid_integrate_task_list 110 12.3 3.675 3.739 3.675 3.739 hybrid_alltoall_any 4261 16.3 0.256 0.550 2.876 3.589 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.547 3.547 make_images_data 4110 15.4 0.041 0.043 2.836 3.461 wfi_extrapolate 11 7.9 0.001 0.001 3.321 3.321 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.798 3.250 calculate_dm_sparse 110 9.5 0.001 0.001 3.174 3.192 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.957 2.958 cp_fm_diag_elpa_base 48 14.0 2.417 2.620 2.954 2.954 fft3d_ps 1111 14.6 1.301 1.308 2.911 2.919 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.911 2.912 grid_collocate_task_list 110 9.6 2.670 2.677 2.670 2.677 calculate_first_density_matrix 1 7.0 0.000 0.000 2.453 2.454 potential_pw2rs 110 12.3 0.021 0.021 2.433 2.436 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.418 2.427 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.347 2.360 rs_pw_transfer 902 11.9 0.010 0.011 2.337 2.355 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.269 2.327 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.104 2.205 mp_alltoall_d11v 2046 13.8 1.990 2.077 1.990 2.077 cp_fm_cholesky_decompose 22 10.9 2.005 2.030 2.005 2.030 qs_create_task_list 11 7.9 0.028 0.029 1.914 1.964 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.960 1.963 generate_qs_task_list 11 8.9 0.730 0.785 1.886 1.936 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.789 1.827 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=87.402000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1228.727273, yerr=56.970632 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c60ebb48d63dcff4bc85c6fc137b78e381e1214e_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420241154048 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528896499712 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514757E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755939872 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 630.206464E+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 3175955098256 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 11059 797. MP_Sync 87 MP_Alltoall 2226 1956353. 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 0.015 0.031 216.025 216.026 qs_mol_dyn_low 1 2.0 0.003 0.003 215.525 215.539 qs_forces 11 3.9 0.005 0.005 215.096 215.097 qs_energies 11 4.9 0.001 0.001 209.426 209.451 scf_env_do_scf 11 5.9 0.001 0.001 185.392 185.396 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 164.782 164.784 dbcsr_multiply_generic 2507 12.6 0.182 0.187 132.036 132.726 velocity_verlet 10 3.0 0.001 0.002 124.781 124.786 qs_scf_new_mos 117 7.6 0.001 0.001 124.306 124.588 qs_scf_loop_do_ot 117 8.6 0.001 0.001 124.306 124.587 ot_scf_mini 117 9.6 0.003 0.003 117.653 117.935 multiply_cannon 2507 13.6 0.238 0.248 108.603 110.721 multiply_cannon_loop 2507 14.6 2.080 2.124 106.386 108.570 ot_mini 117 10.6 0.001 0.001 65.591 65.928 multiply_cannon_multrec 60168 15.6 33.450 35.101 41.678 43.656 qs_ot_get_derivative 117 11.6 0.001 0.001 40.784 41.071 rebuild_ks_matrix 128 8.3 0.001 0.001 33.972 34.431 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 33.971 34.431 mp_waitall_1 267128 16.5 28.370 31.410 28.370 31.410 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.463 30.919 qs_ot_get_p 128 10.4 0.001 0.002 29.994 30.299 multiply_cannon_sync_h2d 60168 15.6 27.584 29.791 27.584 29.791 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.432 25.214 apply_single 128 13.6 0.001 0.001 24.431 25.214 ot_diis_step 117 11.6 0.008 0.008 24.598 24.599 qs_ot_p2m_diag 83 11.4 0.079 0.091 23.411 23.475 cp_dbcsr_syevd 83 12.4 0.005 0.005 20.832 20.833 init_scf_loop 11 6.9 0.000 0.001 20.535 20.536 init_scf_run 11 5.9 0.000 0.001 19.890 19.891 scf_env_initial_rho_setup 11 6.9 0.001 0.001 19.890 19.891 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.427 19.623 multiply_cannon_metrocomm3 60168 15.6 0.111 0.116 15.838 18.464 cp_fm_diag_elpa 83 13.4 0.000 0.000 17.844 17.876 cp_fm_diag_elpa_base 83 14.4 17.772 17.807 17.840 17.872 prepare_preconditioner 11 7.9 0.000 0.000 15.900 15.947 make_preconditioner 11 8.9 0.000 0.000 15.900 15.947 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.149 15.323 sum_up_and_integrate 128 10.3 0.088 0.105 14.385 14.407 integrate_v_rspace 128 11.3 0.004 0.004 14.297 14.322 make_m2s 5014 13.6 0.105 0.113 13.859 14.211 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.012 14.123 calculate_rho_elec 128 8.7 0.046 0.065 14.011 14.122 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 14.073 14.085 make_images 5014 14.6 0.399 0.417 13.675 14.041 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 13.194 13.305 density_rs2pw 128 9.7 0.006 0.008 7.270 10.544 calculate_first_density_matrix 1 7.0 0.000 0.002 10.459 10.465 acc_transpose_blocks 60168 15.6 0.296 0.306 10.318 10.403 mp_sum_l 7950 12.9 8.304 9.749 8.304 9.749 wfi_extrapolate 11 7.9 0.001 0.001 9.247 9.248 rs_pw_transfer 1046 11.9 0.016 0.018 5.951 9.234 cp_fm_cholesky_invert 11 10.9 9.071 9.079 9.071 9.079 acc_transpose_blocks_kernels 60168 16.6 0.689 0.716 8.726 8.759 calculate_dm_sparse 128 9.5 0.001 0.001 8.472 8.563 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.050 8.228 dbcsr_mm_accdrv_process 124484 16.2 3.161 3.249 7.786 8.212 pw_transfer 1547 11.6 0.074 0.088 7.908 8.089 jit_kernel_transpose 5 15.6 8.036 8.058 8.036 8.058 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.940 8.019 multiply_cannon_metrocomm1 60168 15.6 0.089 0.092 5.886 7.987 fft_wrap_pw1pw2 1291 12.7 0.011 0.012 7.706 7.889 make_images_data 5014 15.6 0.065 0.070 6.828 7.702 grid_integrate_task_list 128 12.3 7.025 7.607 7.025 7.607 hybrid_alltoall_any 5200 16.5 0.291 2.268 5.960 7.353 fft_wrap_pw1pw2_140 523 13.2 1.272 1.319 6.700 6.895 mp_waitany 16020 13.9 2.743 5.999 2.743 5.999 grid_collocate_task_list 128 9.7 4.677 5.990 4.677 5.990 fft3d_ps 1291 14.7 2.145 2.795 5.474 5.806 mp_alltoall_d11v 2415 14.1 4.298 5.493 4.298 5.493 rs_pw_transfer_RS2PW_140 139 11.5 0.277 0.292 2.159 5.449 potential_pw2rs 128 12.3 0.009 0.011 4.727 4.752 cp_fm_cholesky_decompose 22 10.9 4.657 4.671 4.657 4.671 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=216.026000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=597.636364, yerr=6.314093 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c60ebb48d63dcff4bc85c6fc137b78e381e1214e_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 2.199914E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 6022464 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.2 marketing flops 145.647559E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 834.252800E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2425920 MPI messages size (bytes): total size 4.132350E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703416E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 71436 2336489472 32768 < size <= 131072 728832 55956209664 131072 < size <= 4194304 1386864 1409906900992 4194304 < size <= 16777216 155760 1473826772352 16777216 < size 68112 1190343475200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4024 57909. MP_Allreduce 11138 958. MP_Sync 87 MP_Alltoall 1983 6393876. MP_SendRecv 12126 47072. MP_ISendRecv 12126 47072. MP_Wait 26114 MP_ISend 11836 212447. MP_IRecv 11836 212447. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.034 193.565 193.565 qs_mol_dyn_low 1 2.0 0.003 0.004 193.201 193.216 qs_forces 11 3.9 0.004 0.004 193.118 193.119 qs_energies 11 4.9 0.001 0.001 186.241 186.253 scf_env_do_scf 11 5.9 0.001 0.001 169.834 169.844 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 136.277 136.279 velocity_verlet 10 3.0 0.001 0.001 122.392 122.394 dbcsr_multiply_generic 2527 12.6 0.194 0.201 98.379 99.669 qs_scf_new_mos 118 7.6 0.001 0.001 95.500 95.913 qs_scf_loop_do_ot 118 8.6 0.001 0.001 95.499 95.912 ot_scf_mini 118 9.6 0.004 0.004 90.637 91.099 multiply_cannon 2527 13.6 0.481 0.534 77.853 82.573 multiply_cannon_loop 2527 14.6 1.253 1.293 74.542 77.300 ot_mini 118 10.6 0.001 0.001 50.537 50.970 mp_waitall_1 216438 16.6 25.466 39.892 25.466 39.892 multiply_cannon_multrec 30324 15.6 21.761 25.467 31.446 35.408 rebuild_ks_matrix 129 8.3 0.001 0.001 33.754 34.270 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.019 33.753 34.269 init_scf_loop 11 6.9 0.000 0.000 33.465 33.466 qs_ks_update_qs_env 129 7.6 0.001 0.001 30.342 30.811 multiply_cannon_metrocomm3 30324 15.6 0.090 0.096 16.390 29.259 prepare_preconditioner 11 7.9 0.000 0.000 29.061 29.138 make_preconditioner 11 8.9 0.000 0.000 29.061 29.138 qs_ot_get_derivative 118 11.6 0.001 0.002 28.214 28.660 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.670 28.219 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 22.399 23.583 apply_single 129 13.6 0.001 0.001 22.398 23.583 qs_ot_get_p 129 10.4 0.001 0.001 21.709 22.292 ot_diis_step 118 11.6 0.015 0.015 22.152 22.154 multiply_cannon_sync_h2d 30324 15.6 19.045 20.968 19.045 20.968 cp_fm_cholesky_invert 11 10.9 16.920 16.933 16.920 16.933 qs_ot_p2m_diag 83 11.4 0.187 0.215 16.819 16.860 make_m2s 5054 13.6 0.090 0.096 14.595 16.106 make_images 5054 14.6 1.171 1.370 14.377 15.884 cp_dbcsr_syevd 83 12.4 0.006 0.006 15.648 15.649 sum_up_and_integrate 129 10.3 0.117 0.135 15.006 15.036 integrate_v_rspace 129 11.3 0.003 0.004 14.889 14.924 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.445 14.481 calculate_rho_elec 129 8.7 0.088 0.105 14.444 14.481 cp_fm_diag_elpa 83 13.4 0.000 0.000 12.475 12.503 cp_fm_diag_elpa_base 83 14.4 12.211 12.316 12.469 12.493 init_scf_run 11 5.9 0.000 0.001 11.568 11.570 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.568 11.569 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 11.072 11.386 multiply_cannon_metrocomm4 27797 15.6 0.103 0.117 3.878 10.956 make_images_data 5054 15.6 0.063 0.072 8.915 10.849 mp_irecv_dv 70031 16.3 3.676 10.562 3.676 10.562 hybrid_alltoall_any 5240 16.5 0.345 1.502 7.676 10.086 density_rs2pw 129 9.7 0.006 0.007 7.634 10.056 dbcsr_mm_accdrv_process 62734 16.2 4.451 5.638 9.142 9.768 pw_transfer 1559 11.6 0.087 0.101 9.019 9.077 fft_wrap_pw1pw2 1301 12.7 0.011 0.011 8.792 8.854 wfi_extrapolate 11 7.9 0.001 0.001 8.477 8.478 rs_pw_transfer 1054 12.0 0.014 0.016 5.918 8.348 fft_wrap_pw1pw2_140 527 13.2 1.341 1.371 7.725 7.801 grid_integrate_task_list 129 12.3 7.229 7.677 7.229 7.677 cp_fm_cholesky_decompose 22 10.9 7.165 7.247 7.165 7.247 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 6.361 7.166 calculate_dm_sparse 129 9.5 0.001 0.001 6.523 6.648 mp_sum_l 8010 12.9 4.128 6.225 4.128 6.225 fft3d_ps 1301 14.7 2.834 3.015 6.085 6.131 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.106 6.116 grid_collocate_task_list 129 9.7 4.827 6.006 4.827 6.006 qs_ot_get_orbitals 118 10.6 0.001 0.001 5.447 5.508 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.244 5.403 potential_pw2rs 129 12.3 0.016 0.019 5.156 5.180 mp_allgather_i34 2527 14.6 1.869 5.162 1.869 5.162 mp_waitany 11836 13.9 2.550 5.065 2.550 5.065 mp_alltoall_d11v 2423 14.1 4.356 4.930 4.356 4.930 rs_pw_transfer_RS2PW_140 140 11.5 0.351 0.372 2.123 4.546 mp_sum_d 4499 12.1 2.642 4.165 2.642 4.165 dbcsr_complete_redistribute 395 12.7 0.774 0.858 3.233 4.070 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=193.565000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=795.636364, yerr=1.149919 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c60ebb48d63dcff4bc85c6fc137b78e381e1214e_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 961.777664E+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 11184 996. 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 0.090 0.112 180.044 180.045 qs_mol_dyn_low 1 2.0 0.003 0.003 179.410 179.425 qs_forces 11 3.9 0.004 0.004 179.309 179.318 qs_energies 11 4.9 0.001 0.001 172.610 172.621 scf_env_do_scf 11 5.9 0.001 0.001 156.869 156.870 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 121.331 121.332 velocity_verlet 10 3.0 0.001 0.001 115.739 115.741 dbcsr_multiply_generic 2529 12.6 0.182 0.185 82.659 83.643 qs_scf_new_mos 118 7.6 0.001 0.001 82.228 82.576 qs_scf_loop_do_ot 118 8.6 0.001 0.001 82.227 82.575 ot_scf_mini 118 9.6 0.003 0.004 78.027 78.388 multiply_cannon 2529 13.6 0.503 0.523 62.320 66.942 multiply_cannon_loop 2529 14.6 0.868 0.894 59.394 61.959 ot_mini 118 10.6 0.001 0.001 43.099 43.482 init_scf_loop 11 6.9 0.000 0.000 35.439 35.441 mp_waitall_1 172006 16.6 25.932 35.431 25.932 35.431 rebuild_ks_matrix 129 8.3 0.001 0.001 31.328 31.855 qs_ks_build_kohn_sham_matrix 129 9.3 0.016 0.018 31.327 31.854 prepare_preconditioner 11 7.9 0.000 0.000 31.375 31.422 make_preconditioner 11 8.9 0.000 0.000 31.375 31.422 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.990 30.402 qs_ks_update_qs_env 129 7.6 0.001 0.001 28.213 28.695 multiply_cannon_multrec 20232 15.6 13.443 16.700 22.351 25.496 multiply_cannon_metrocomm3 20232 15.6 0.058 0.062 15.728 25.083 qs_ot_get_derivative 118 11.6 0.002 0.002 22.993 23.370 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 20.222 21.203 apply_single 129 13.6 0.001 0.001 20.222 21.203 qs_ot_get_p 129 10.4 0.001 0.001 20.146 20.628 ot_diis_step 118 11.6 0.018 0.020 20.004 20.005 make_m2s 5058 13.6 0.080 0.086 15.274 16.037 multiply_cannon_sync_h2d 20232 15.6 14.298 16.029 14.298 16.029 qs_ot_p2m_diag 84 11.4 0.268 0.275 15.853 15.867 make_images 5058 14.6 1.198 1.293 15.036 15.794 sum_up_and_integrate 129 10.3 0.132 0.146 14.955 14.985 integrate_v_rspace 129 11.3 0.004 0.004 14.822 14.849 cp_dbcsr_syevd 84 12.4 0.006 0.006 14.799 14.800 cp_fm_cholesky_invert 11 10.9 14.790 14.799 14.790 14.799 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.726 14.755 calculate_rho_elec 129 8.7 0.131 0.146 14.725 14.755 cp_fm_diag_elpa 84 13.4 0.000 0.000 11.542 11.560 cp_fm_diag_elpa_base 84 14.4 11.119 11.283 11.539 11.556 make_images_data 5058 15.6 0.058 0.066 9.568 10.803 init_scf_run 11 5.9 0.000 0.001 10.703 10.704 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.703 10.704 hybrid_alltoall_any 5245 16.5 0.438 2.026 8.203 9.898 multiply_cannon_metrocomm4 17703 15.6 0.062 0.073 3.543 9.519 density_rs2pw 129 9.7 0.006 0.007 7.445 9.506 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.111 9.375 mp_irecv_dv 50659 16.2 3.420 9.265 3.420 9.265 pw_transfer 1559 11.6 0.087 0.104 9.123 9.219 fft_wrap_pw1pw2 1301 12.7 0.010 0.012 8.896 8.998 dbcsr_mm_accdrv_process 41846 16.2 4.570 5.449 8.364 8.498 grid_integrate_task_list 129 12.3 7.342 7.975 7.342 7.975 fft_wrap_pw1pw2_140 527 13.2 1.419 1.443 7.836 7.948 wfi_extrapolate 11 7.9 0.001 0.001 7.534 7.534 rs_pw_transfer 1054 12.0 0.014 0.015 5.439 7.516 cp_fm_cholesky_decompose 22 10.9 7.470 7.486 7.470 7.486 cp_fm_upper_to_full 106 14.8 5.691 7.242 5.691 7.242 dbcsr_complete_redistribute 397 12.7 1.173 1.204 4.657 6.401 fft3d_ps 1301 14.7 2.759 2.981 6.003 6.074 grid_collocate_task_list 129 9.7 5.016 5.953 5.016 5.953 calculate_dm_sparse 129 9.5 0.001 0.001 5.802 5.882 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.625 5.633 mp_alltoall_d11v 2429 14.1 4.480 5.347 4.480 5.347 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.548 5.270 copy_fm_to_dbcsr 210 11.7 0.002 0.002 3.456 5.201 potential_pw2rs 129 12.3 0.020 0.023 4.919 4.934 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.706 4.843 mp_sum_l 8016 12.9 3.230 4.703 3.230 4.703 mp_waitany 11836 13.9 2.413 4.623 2.413 4.623 mp_allgather_i34 2529 14.6 1.434 4.594 1.434 4.594 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.094 4.126 transfer_fm_to_dbcsr 11 9.9 0.020 0.027 2.366 4.098 rs_pw_transfer_RS2PW_140 140 11.5 0.333 0.350 1.960 4.028 mp_alltoall_i22 720 14.1 1.964 3.826 1.964 3.826 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 3.765 3.767 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.618 3.664 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=180.045000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=902.272727, yerr=15.112252 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c60ebb48d63dcff4bc85c6fc137b78e381e1214e_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420241154048 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528896499712 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514757E+12 0.0% 0.0% 100.0% flops max/rank 4.353788E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755939872 0.0% 0.0% 100.0% number of processed stacks 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 1.150570E+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 665379345520 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58354. MP_Allreduce 11057 1083. 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 0.014 0.033 189.906 189.906 qs_mol_dyn_low 1 2.0 0.003 0.003 189.412 189.425 qs_forces 11 3.9 0.004 0.004 189.299 189.306 qs_energies 11 4.9 0.001 0.001 182.027 182.036 scf_env_do_scf 11 5.9 0.001 0.001 165.203 165.211 velocity_verlet 10 3.0 0.001 0.001 126.024 126.026 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 118.355 118.356 qs_scf_new_mos 117 7.6 0.001 0.001 80.539 80.852 qs_scf_loop_do_ot 117 8.6 0.001 0.001 80.539 80.851 dbcsr_multiply_generic 2507 12.6 0.188 0.194 79.577 80.380 ot_scf_mini 117 9.6 0.003 0.003 76.040 76.314 multiply_cannon 2507 13.6 0.557 0.592 54.834 58.080 multiply_cannon_loop 2507 14.6 1.186 1.217 51.085 52.858 init_scf_loop 11 6.9 0.000 0.000 46.726 46.727 ot_mini 117 10.6 0.001 0.001 42.375 42.672 prepare_preconditioner 11 7.9 0.000 0.000 42.613 42.646 make_preconditioner 11 8.9 0.000 0.000 42.613 42.646 make_full_inverse_cholesky 11 9.9 0.012 0.020 36.235 41.158 multiply_cannon_multrec 30084 15.6 13.991 18.730 25.975 30.680 rebuild_ks_matrix 128 8.3 0.001 0.001 30.113 30.380 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.019 30.113 30.379 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.085 27.328 mp_waitall_1 147882 16.7 17.554 27.255 17.554 27.255 qs_ot_get_derivative 117 11.6 0.002 0.002 22.671 22.952 make_m2s 5014 13.6 0.095 0.099 20.364 21.450 make_images 5014 14.6 1.945 2.259 20.054 21.139 qs_ot_get_p 128 10.4 0.001 0.001 19.858 20.184 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 19.175 19.783 apply_single 128 13.6 0.001 0.001 19.175 19.782 ot_diis_step 117 11.6 0.017 0.018 19.576 19.578 cp_fm_upper_to_full 105 14.8 11.465 16.772 11.465 16.772 cp_fm_cholesky_invert 11 10.9 16.420 16.428 16.420 16.428 qs_ot_p2m_diag 83 11.4 0.342 0.389 15.753 15.806 multiply_cannon_metrocomm3 30084 15.6 0.047 0.050 6.576 15.139 sum_up_and_integrate 128 10.3 0.140 0.151 14.815 14.842 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.756 14.789 calculate_rho_elec 128 8.7 0.173 0.189 14.755 14.788 integrate_v_rspace 128 11.3 0.004 0.004 14.675 14.707 cp_dbcsr_syevd 83 12.4 0.005 0.006 14.449 14.451 multiply_cannon_sync_h2d 30084 15.6 11.712 12.996 11.712 12.996 make_images_data 5014 15.6 0.060 0.065 11.031 12.889 dbcsr_complete_redistribute 395 12.7 1.505 1.613 8.994 12.801 dbcsr_mm_accdrv_process 62264 16.2 7.418 8.431 11.557 12.120 hybrid_alltoall_any 5200 16.5 0.516 2.072 9.828 11.455 cp_fm_diag_elpa 83 13.4 0.000 0.000 11.315 11.325 copy_fm_to_dbcsr 209 11.7 0.001 0.002 7.610 11.323 cp_fm_diag_elpa_base 83 14.4 10.344 10.664 11.309 11.320 init_scf_run 11 5.9 0.000 0.001 10.940 10.942 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.940 10.942 transfer_fm_to_dbcsr 11 9.9 0.001 0.003 6.356 9.964 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.397 9.604 pw_transfer 1547 11.6 0.086 0.103 9.217 9.305 mp_alltoall_i22 716 14.1 5.506 9.213 5.506 9.213 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.990 9.084 density_rs2pw 128 9.7 0.006 0.007 7.087 8.563 fft_wrap_pw1pw2_140 523 13.2 1.558 1.583 7.975 8.074 grid_integrate_task_list 128 12.3 7.458 7.958 7.458 7.958 wfi_extrapolate 11 7.9 0.001 0.001 7.785 7.785 cp_fm_cholesky_decompose 22 10.9 7.614 7.718 7.614 7.718 multiply_cannon_metrocomm4 25070 15.6 0.079 0.091 2.830 7.482 mp_irecv_dv 76098 16.2 2.682 7.215 2.682 7.215 calculate_dm_sparse 128 9.5 0.001 0.001 6.231 6.312 rs_pw_transfer 1046 11.9 0.013 0.015 4.761 6.299 fft3d_ps 1291 14.7 2.831 2.926 5.920 5.982 mp_alltoall_d11v 2415 14.1 5.077 5.920 5.077 5.920 grid_collocate_task_list 128 9.7 5.143 5.889 5.143 5.889 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.465 5.514 potential_pw2rs 128 12.3 0.023 0.023 4.588 4.613 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.386 4.456 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 4.426 4.427 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.339 4.414 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.201 4.256 mp_allgather_i34 2507 14.6 1.845 3.985 1.845 3.985 mp_waitany 14952 13.9 2.262 3.810 2.262 3.810 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=189.906000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1083.000000, yerr=18.394663 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c60ebb48d63dcff4bc85c6fc137b78e381e1214e_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963542011904 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019182452736 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019182452736 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796564E+12 0.0% 0.0% 100.0% flops max/rank 5.820057E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499488 0.0% 0.0% 100.0% number of processed stacks 1944496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3448.5 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.474499E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 238560 MPI messages size (bytes): total size 1.321104E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.537828E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 112800 59139686400 4194304 < size <= 16777216 104112 545846722560 16777216 < size 20064 716108490000 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8852 52. MP_Alltoall 9584 804353. MP_ISend 39716 2104723. MP_IRecv 39716 2103824. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3995 58308. MP_Allreduce 11065 1167. MP_Sync 86 MP_Alltoall 1700 18828162. MP_SendRecv 3810 122880. MP_ISendRecv 3810 122880. MP_Wait 16000 MP_ISend 10600 423612. MP_IRecv 10600 423612. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.033 175.130 175.131 qs_mol_dyn_low 1 2.0 0.003 0.003 174.735 174.748 qs_forces 11 3.9 0.004 0.005 174.592 174.596 qs_energies 11 4.9 0.001 0.001 166.999 167.011 scf_env_do_scf 11 5.9 0.001 0.001 149.301 149.305 velocity_verlet 10 3.0 0.001 0.001 114.157 114.163 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 112.940 112.941 qs_scf_new_mos 116 7.6 0.001 0.001 74.342 74.434 qs_scf_loop_do_ot 116 8.6 0.001 0.001 74.342 74.433 dbcsr_multiply_generic 2485 12.5 0.201 0.212 73.803 74.208 ot_scf_mini 116 9.6 0.003 0.004 69.928 70.014 multiply_cannon 2485 13.5 0.582 0.619 53.996 58.811 multiply_cannon_loop 2485 14.5 0.442 0.454 49.347 50.105 ot_mini 116 10.6 0.001 0.001 39.286 39.352 init_scf_loop 11 6.9 0.000 0.000 36.214 36.215 mp_waitall_1 124680 16.7 25.794 32.809 25.794 32.809 prepare_preconditioner 11 7.9 0.000 0.000 32.278 32.314 make_preconditioner 11 8.9 0.000 0.000 32.278 32.314 make_full_inverse_cholesky 11 9.9 0.013 0.026 30.213 30.482 rebuild_ks_matrix 127 8.3 0.001 0.001 29.911 29.989 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.018 29.911 29.988 qs_ks_update_qs_env 127 7.6 0.001 0.001 27.111 27.177 multiply_cannon_multrec 9940 15.5 10.299 14.588 17.622 20.721 ot_diis_step 116 11.6 0.020 0.022 19.651 19.651 qs_ot_get_derivative 116 11.6 0.002 0.002 19.565 19.638 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.302 19.521 apply_single 127 13.6 0.001 0.001 19.302 19.521 multiply_cannon_metrocomm3 9940 15.5 0.023 0.024 12.098 19.085 cp_fm_cholesky_invert 11 10.9 18.568 18.574 18.568 18.574 make_m2s 4970 13.5 0.066 0.070 16.212 18.329 make_images 4970 14.5 2.263 2.595 15.904 18.022 qs_ot_get_p 127 10.4 0.001 0.001 17.832 17.922 qs_rho_update_rho_low 127 7.7 0.001 0.001 15.429 15.467 calculate_rho_elec 127 8.7 0.254 0.265 15.428 15.467 sum_up_and_integrate 127 10.3 0.180 0.191 15.089 15.137 integrate_v_rspace 127 11.3 0.004 0.004 14.908 14.966 qs_ot_p2m_diag 82 11.4 0.489 0.495 14.141 14.161 cp_dbcsr_syevd 82 12.4 0.005 0.005 13.011 13.013 make_images_data 4970 15.5 0.049 0.059 9.889 12.288 multiply_cannon_sync_h2d 9940 15.5 11.531 12.175 11.531 12.175 hybrid_alltoall_any 5155 16.4 0.813 3.640 9.693 12.136 init_scf_run 11 5.9 0.000 0.001 10.899 10.899 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.898 10.899 cp_fm_diag_elpa 82 13.4 0.000 0.000 9.945 9.957 cp_fm_diag_elpa_base 82 14.4 9.697 9.781 9.940 9.953 pw_transfer 1535 11.6 0.086 0.095 9.608 9.631 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 9.385 9.414 fft_wrap_pw1pw2_140 519 13.2 1.897 1.931 8.264 8.302 cp_fm_cholesky_decompose 22 10.9 8.113 8.217 8.113 8.217 grid_integrate_task_list 127 12.3 7.675 8.164 7.675 8.164 density_rs2pw 127 9.7 0.006 0.006 7.003 7.907 qs_ot_get_derivative_diag 76 12.4 0.002 0.003 7.819 7.887 dbcsr_mm_accdrv_process 20590 16.1 2.575 3.372 6.953 7.616 multiply_cannon_metrocomm1 9940 15.5 0.028 0.029 4.494 7.611 mp_allgather_i34 2485 14.5 2.830 7.611 2.830 7.611 wfi_extrapolate 11 7.9 0.001 0.001 7.608 7.609 calculate_dm_sparse 127 9.5 0.001 0.001 6.120 6.197 mp_alltoall_d11v 2401 14.1 5.142 6.182 5.142 6.182 grid_collocate_task_list 127 9.7 5.422 6.091 5.422 6.091 fft3d_ps 1281 14.7 2.741 2.812 5.893 5.929 dbcsr_complete_redistribute 393 12.7 2.164 2.224 5.251 5.648 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.396 5.434 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.279 5.279 rs_pw_transfer 1038 11.9 0.012 0.013 4.254 5.162 potential_pw2rs 127 12.3 0.025 0.026 4.527 4.541 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.231 4.250 multiply_cannon_metrocomm4 7455 15.5 0.023 0.026 1.838 4.011 mp_irecv_dv 28618 15.9 1.803 3.936 1.803 3.936 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.565 3.842 copy_fm_to_dbcsr 208 11.6 0.001 0.002 3.405 3.739 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.699 3.711 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.667 3.707 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.568 3.637 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=175.131000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1385.272727, yerr=37.182818 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c60ebb48d63dcff4bc85c6fc137b78e381e1214e_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430450896896 0.0% 0.0% 100.0% flops 32 x 32 x 32 1984274890752 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986253074432 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992012447744 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753953742848 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 11613050818560 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239188242432 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239188242432 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911113097216 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.254474E+12 0.0% 0.0% 100.0% flops max/rank 11.788742E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806708064 0.0% 0.0% 100.0% number of processed stacks 1983968 0.0% 0.0% 100.0% average stack size 0.0 0.0 3430.9 marketing flops 145.672406E+12 ------------------------------------------------------------------------------- # multiplications 2539 max memory usage/rank 3.156677E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101560 MPI messages size (bytes): total size 1.145305E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.277126E+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 46048 35769024512 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592270320 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4073 58147. MP_Allreduce 11259 1498. MP_Sync 88 MP_Alltoall 1724 36993351. 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 0.026 0.048 301.673 301.674 qs_mol_dyn_low 1 2.0 0.003 0.004 301.118 301.130 qs_forces 11 3.9 0.005 0.006 300.946 300.954 qs_energies 11 4.9 0.001 0.001 291.521 291.533 scf_env_do_scf 11 5.9 0.001 0.001 269.117 269.131 velocity_verlet 10 3.0 0.001 0.001 217.450 217.467 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 143.218 143.221 init_scf_loop 11 6.9 0.000 0.000 125.646 125.651 prepare_preconditioner 11 7.9 0.000 0.000 120.531 120.559 make_preconditioner 11 8.9 0.000 0.000 120.531 120.559 make_full_inverse_cholesky 11 9.9 0.035 0.039 96.115 117.695 qs_scf_new_mos 118 7.6 0.001 0.001 92.051 92.166 qs_scf_loop_do_ot 118 8.6 0.001 0.001 92.050 92.165 ot_scf_mini 118 9.6 0.004 0.004 87.181 87.232 dbcsr_multiply_generic 2539 12.6 0.215 0.228 84.463 85.071 cp_fm_upper_to_full 106 14.8 53.542 77.045 53.542 77.045 multiply_cannon 2539 13.6 0.719 0.758 59.622 60.029 multiply_cannon_loop 2539 14.6 0.478 0.487 55.881 57.247 ot_mini 118 10.6 0.001 0.001 45.113 45.171 dbcsr_complete_redistribute 397 12.7 4.000 4.038 30.103 43.175 copy_fm_to_dbcsr 210 11.7 0.001 0.002 26.733 39.759 rebuild_ks_matrix 129 8.3 0.001 0.001 38.624 38.658 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.017 38.624 38.657 transfer_fm_to_dbcsr 11 9.9 0.030 0.033 24.370 37.128 qs_ks_update_qs_env 129 7.6 0.001 0.001 35.603 35.634 mp_alltoall_i22 720 14.1 22.181 35.327 22.181 35.327 cp_fm_cholesky_invert 11 10.9 33.524 33.530 33.524 33.530 mp_waitall_1 104980 16.8 28.960 33.041 28.960 33.041 qs_ot_get_p 129 10.4 0.001 0.001 26.331 26.382 qs_ot_get_derivative 118 11.6 0.002 0.002 24.545 24.599 qs_ot_p2m_diag 84 11.4 0.889 0.895 22.238 22.272 make_m2s 5078 13.6 0.075 0.080 20.116 21.121 qs_rho_update_rho_low 129 7.7 0.001 0.001 20.723 20.743 calculate_rho_elec 129 8.7 0.484 0.486 20.722 20.742 make_images 5078 14.6 3.804 4.019 19.632 20.641 multiply_cannon_metrocomm3 10156 15.6 0.023 0.025 19.459 20.609 ot_diis_step 118 11.6 0.022 0.023 20.538 20.539 cp_dbcsr_syevd 84 12.4 0.006 0.006 20.468 20.469 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.742 19.956 apply_single 129 13.6 0.001 0.001 19.742 19.955 sum_up_and_integrate 129 10.3 0.323 0.324 19.741 19.825 integrate_v_rspace 129 11.3 0.004 0.004 19.418 19.502 multiply_cannon_multrec 10156 15.6 10.587 12.389 18.064 18.160 cp_fm_diag_elpa 84 13.4 0.000 0.000 17.213 17.214 cp_fm_diag_elpa_base 84 14.4 12.679 14.351 17.209 17.209 multiply_cannon_sync_h2d 10156 15.6 15.771 15.787 15.771 15.787 make_images_data 5078 15.6 0.058 0.063 10.743 12.587 hybrid_alltoall_any 5265 16.5 1.309 3.058 10.800 12.544 init_scf_run 11 5.9 0.000 0.001 12.346 12.348 scf_env_initial_rho_setup 11 6.9 0.001 0.002 12.346 12.347 pw_transfer 1559 11.6 0.094 0.094 11.552 11.559 fft_wrap_pw1pw2 1301 12.7 0.011 0.012 11.315 11.321 fft_wrap_pw1pw2_140 527 13.2 3.098 3.135 10.057 10.067 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.628 9.676 mp_alltoall_d11v 2429 14.1 8.427 9.356 8.427 9.356 wfi_extrapolate 11 7.9 0.001 0.001 9.232 9.232 dbcsr_mm_accdrv_process 20974 16.1 3.890 5.807 7.234 9.116 cp_fm_cholesky_decompose 22 10.9 8.954 8.995 8.954 8.995 grid_integrate_task_list 129 12.3 8.652 8.825 8.652 8.825 density_rs2pw 129 9.7 0.005 0.006 8.078 8.264 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 7.990 7.992 calculate_dm_sparse 129 9.5 0.001 0.001 6.722 6.810 grid_collocate_task_list 129 9.7 6.493 6.562 6.493 6.562 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.346 6.484 fft3d_ps 1301 14.7 2.864 2.877 6.348 6.385 rs_scatter_matrices 140 9.7 3.696 4.708 6.119 6.349 copy_dbcsr_to_fm 187 11.8 0.004 0.004 6.029 6.125 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=301.674000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2880.454545, yerr=157.345228 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c60ebb48d63dcff4bc85c6fc137b78e381e1214e_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 1.259594E+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 3139 6114. MP_Sync 4 MP_Alltoall 54 MP_SendRecv 285 19200. MP_ISendRecv 285 19200. MP_Wait 1017 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.039 0.052 85.415 85.416 qs_energies 1 2.0 0.000 0.000 84.705 84.714 ls_scf 1 3.0 0.000 0.000 83.787 83.796 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.664 72.847 multiply_cannon 111 7.7 0.017 0.020 55.906 57.120 multiply_cannon_loop 111 8.7 0.207 0.224 52.488 53.896 ls_scf_main 1 4.0 0.000 0.000 52.301 52.312 density_matrix_trs4 2 5.0 0.002 0.003 46.734 46.795 ls_scf_init_scf 1 4.0 0.000 0.000 28.397 28.399 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.257 27.309 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.097 25.111 mp_waitall_1 11031 10.9 22.475 25.013 22.475 25.013 multiply_cannon_multrec 2664 9.7 8.179 8.929 15.605 17.357 multiply_cannon_sync_h2d 2664 9.7 13.569 15.417 13.569 15.417 make_m2s 222 7.7 0.008 0.010 13.130 13.545 make_images 222 8.7 0.099 0.109 13.108 13.522 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 9.641 12.405 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.472 9.026 make_images_data 222 9.7 0.004 0.005 7.696 8.189 dbcsr_mm_accdrv_process 4760 10.4 0.513 0.633 7.042 8.022 hybrid_alltoall_any 227 10.6 0.215 1.845 6.659 7.605 dbcsr_mm_accdrv_process_sort 4760 11.4 6.327 7.235 6.327 7.235 calculate_norms 4752 9.8 5.546 6.216 5.546 6.216 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.033 5.206 mp_sum_l 887 5.1 3.090 4.408 3.090 4.408 make_images_sizes 222 9.7 0.000 0.000 0.673 3.973 mp_alltoall_i44 222 10.7 0.673 3.973 0.673 3.973 multiply_cannon_metrocomm4 2442 9.7 0.012 0.014 2.048 3.521 mp_irecv_dv 6231 10.9 2.031 3.507 2.031 3.507 arnoldi_extremal 4 6.8 0.000 0.000 3.267 3.291 arnoldi_normal_ev 4 7.8 0.001 0.003 3.267 3.290 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.310 3.232 build_subspace 16 8.4 0.009 0.012 3.174 3.176 ls_scf_post 1 4.0 0.000 0.000 3.089 3.098 ls_scf_store_result 1 5.0 0.000 0.000 2.892 2.926 dbcsr_special_finalize 555 9.7 0.005 0.006 2.411 2.799 dbcsr_merge_single_wm 555 10.7 0.469 0.622 2.403 2.790 make_images_pack 222 9.7 2.203 2.628 2.205 2.629 dbcsr_matrix_vector_mult 304 9.0 0.006 0.014 2.321 2.582 dbcsr_sort_data 658 11.4 2.194 2.536 2.194 2.536 dbcsr_matrix_vector_mult_local 304 10.0 2.069 2.480 2.071 2.482 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.353 2.426 buffer_matrices_ensure_size 222 8.7 1.788 2.138 1.788 2.138 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.786 1.787 rebuild_ks_matrix 3 7.3 0.000 0.000 1.776 1.777 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.776 1.777 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.416000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1140.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c60ebb48d63dcff4bc85c6fc137b78e381e1214e_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 2.104812E+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 3138 10075. MP_Sync 4 MP_Alltoall 47 15335933. MP_SendRecv 141 57600. MP_ISendRecv 141 57600. MP_Wait 687 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.041 0.098 97.968 97.970 qs_energies 1 2.0 0.000 0.000 97.282 97.286 ls_scf 1 3.0 0.000 0.000 95.962 95.965 dbcsr_multiply_generic 111 6.7 0.015 0.017 81.801 82.101 multiply_cannon 111 7.7 0.029 0.042 59.320 63.439 multiply_cannon_loop 111 8.7 0.115 0.121 56.030 59.261 ls_scf_main 1 4.0 0.000 0.000 55.631 55.639 density_matrix_trs4 2 5.0 0.002 0.003 49.859 50.050 ls_scf_init_scf 1 4.0 0.000 0.000 36.814 36.814 ls_scf_init_matrix_S 1 5.0 0.000 0.000 35.596 35.684 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 33.141 33.159 mp_waitall_1 9105 10.9 22.417 31.649 22.417 31.649 multiply_cannon_multrec 1332 9.7 13.100 16.808 22.412 27.249 multiply_cannon_metrocomm3 1332 9.7 0.006 0.007 12.368 21.579 make_m2s 222 7.7 0.006 0.007 15.775 16.508 make_images 222 8.7 1.582 1.972 15.745 16.477 dbcsr_mm_accdrv_process 4041 10.4 0.281 0.438 8.908 10.447 make_images_data 222 9.7 0.004 0.004 9.118 10.161 dbcsr_mm_accdrv_process_sort 4041 11.4 8.503 10.025 8.503 10.025 hybrid_alltoall_any 227 10.6 0.522 2.497 8.490 9.607 mp_sum_l 887 5.1 5.409 8.844 5.409 8.844 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 1.563 8.053 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.202 7.716 mp_irecv_dv 3311 11.0 3.183 7.660 3.183 7.660 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.185 7.147 calculate_norms 2376 9.8 5.976 6.666 5.976 6.666 compute_matrix_preconditioner 1 6.0 0.000 0.001 6.580 6.584 multiply_cannon_sync_h2d 1332 9.7 4.839 6.131 4.839 6.131 acc_transpose_blocks 1332 9.7 0.007 0.007 5.302 6.042 acc_transpose_blocks_kernels 1332 10.7 0.018 0.019 5.254 5.993 jit_kernel_transpose 1 13.0 5.236 5.975 5.236 5.975 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.095 5.350 arnoldi_extremal 4 6.8 0.000 0.000 4.736 4.764 arnoldi_normal_ev 4 7.8 0.001 0.005 4.736 4.764 build_subspace 16 8.4 0.014 0.021 4.475 4.477 ls_scf_post 1 4.0 0.000 0.000 3.518 3.521 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.176 3.410 ls_scf_store_result 1 5.0 0.000 0.000 3.237 3.341 dbcsr_matrix_vector_mult_local 304 10.0 2.747 3.236 2.749 3.238 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.650 2.749 mp_allgather_i34 111 8.7 0.985 2.504 0.985 2.504 make_images_pack 222 9.7 2.054 2.425 2.056 2.427 dbcsr_sort_data 436 11.2 1.841 2.066 1.841 2.066 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=97.970000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1707.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c60ebb48d63dcff4bc85c6fc137b78e381e1214e_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 2.773520E+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 3138 10896. MP_Sync 4 MP_Alltoall 47 23526250. MP_SendRecv 93 57600. MP_ISendRecv 93 57600. MP_Wait 639 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.045 0.091 95.358 95.360 qs_energies 1 2.0 0.000 0.000 94.653 94.667 ls_scf 1 3.0 0.007 0.032 92.902 92.906 dbcsr_multiply_generic 111 6.7 0.016 0.016 77.550 77.783 ls_scf_main 1 4.0 0.001 0.015 58.294 58.297 multiply_cannon 111 7.7 0.039 0.113 53.304 57.042 multiply_cannon_loop 111 8.7 0.099 0.104 49.752 54.146 density_matrix_trs4 2 5.0 0.002 0.005 52.289 52.435 mp_waitall_1 7281 11.0 24.279 34.284 24.279 34.284 ls_scf_init_scf 1 4.0 0.000 0.002 31.053 31.056 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.837 29.891 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 27.397 27.409 multiply_cannon_multrec 888 9.7 12.586 15.328 21.212 24.578 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.342 23.558 make_m2s 222 7.7 0.006 0.007 17.227 18.443 make_images 222 8.7 1.980 2.300 17.188 18.404 hybrid_alltoall_any 227 10.6 0.623 2.880 9.483 10.939 make_images_data 222 9.7 0.003 0.004 9.775 10.825 mp_sum_l 887 5.1 5.477 9.605 5.477 9.605 dbcsr_mm_accdrv_process 3754 10.4 0.245 0.410 8.152 9.376 dbcsr_mm_accdrv_process_sort 3754 11.4 7.783 8.966 7.783 8.966 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.250 7.592 multiply_cannon_sync_h2d 888 9.7 6.025 7.452 6.025 7.452 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.442 7.000 mp_irecv_dv 2335 11.1 2.428 6.963 2.428 6.963 multiply_cannon_metrocomm1 888 9.7 0.002 0.003 3.787 6.503 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.930 5.155 arnoldi_extremal 4 6.8 0.000 0.000 5.051 5.065 arnoldi_normal_ev 4 7.8 0.001 0.004 5.051 5.065 calculate_norms 1584 9.8 4.313 4.768 4.313 4.768 build_subspace 16 8.4 0.014 0.020 4.751 4.758 mp_allgather_i34 111 8.7 1.396 3.874 1.396 3.874 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.415 3.743 dbcsr_matrix_vector_mult_local 304 10.0 3.003 3.577 3.006 3.579 ls_scf_post 1 4.0 0.000 0.002 3.548 3.558 ls_scf_store_result 1 5.0 0.000 0.000 3.290 3.382 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.896 3.018 make_images_sizes 222 9.7 0.000 0.000 1.143 2.228 mp_alltoall_i44 222 10.7 1.142 2.228 1.142 2.228 dbcsr_sort_data 325 11.1 1.848 2.151 1.848 2.151 make_images_pack 222 9.7 1.816 2.082 1.819 2.085 dbcsr_data_release 9322 10.9 1.318 1.959 1.318 1.959 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.926 1.929 rebuild_ks_matrix 3 7.3 0.000 0.000 1.908 1.911 qs_ks_build_kohn_sham_matrix 3 8.3 0.005 0.020 1.908 1.911 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=95.360000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2192.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c60ebb48d63dcff4bc85c6fc137b78e381e1214e_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 3.355517E+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 3138 13030. MP_Sync 4 MP_Alltoall 47 30278988. MP_SendRecv 69 86400. MP_ISendRecv 69 86400. MP_Wait 531 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.040 0.054 97.791 97.792 qs_energies 1 2.0 0.000 0.000 97.128 97.133 ls_scf 1 3.0 0.000 0.001 95.219 95.225 dbcsr_multiply_generic 111 6.7 0.017 0.017 78.855 79.141 ls_scf_main 1 4.0 0.000 0.003 59.187 59.188 multiply_cannon 111 7.7 0.048 0.158 51.787 56.994 density_matrix_trs4 2 5.0 0.002 0.003 53.012 53.115 multiply_cannon_loop 111 8.7 0.114 0.127 46.725 49.955 ls_scf_init_scf 1 4.0 0.000 0.002 32.785 32.786 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.544 31.624 mp_waitall_1 6369 11.0 22.776 29.131 22.776 29.131 matrix_sqrt_Newton_Schulz 2 6.5 0.002 0.005 28.971 28.997 multiply_cannon_multrec 1332 9.7 14.164 17.803 22.141 24.728 make_m2s 222 7.7 0.007 0.008 21.154 22.604 make_images 222 8.7 3.138 3.584 21.104 22.555 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.015 18.088 make_images_data 222 9.7 0.004 0.004 11.758 13.325 hybrid_alltoall_any 227 10.6 0.794 3.771 11.095 12.694 dbcsr_mm_accdrv_process 3641 10.4 0.214 0.419 7.606 9.150 dbcsr_mm_accdrv_process_sort 3641 11.4 7.223 8.732 7.223 8.732 mp_sum_l 887 5.1 4.356 8.204 4.356 8.204 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.423 6.630 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.831 6.272 multiply_cannon_sync_h2d 1332 9.7 5.499 6.086 5.499 6.086 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.071 6.034 mp_irecv_dv 3229 10.9 2.048 5.951 2.048 5.951 arnoldi_extremal 4 6.8 0.000 0.000 5.224 5.247 arnoldi_normal_ev 4 7.8 0.002 0.009 5.224 5.246 build_subspace 16 8.4 0.014 0.021 4.850 4.859 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.573 4.732 calculate_norms 2376 9.8 4.202 4.524 4.202 4.524 mp_allgather_i34 111 8.7 2.190 4.459 2.190 4.459 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.568 3.847 dbcsr_matrix_vector_mult_local 304 10.0 3.162 3.651 3.164 3.653 dbcsr_sort_data 658 11.4 3.052 3.359 3.052 3.359 dbcsr_special_finalize 555 9.7 0.006 0.007 2.816 3.258 ls_scf_post 1 4.0 0.000 0.001 3.247 3.251 dbcsr_merge_single_wm 555 10.7 0.536 0.664 2.807 3.249 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.066 3.120 ls_scf_store_result 1 5.0 0.000 0.000 2.995 3.048 dbcsr_data_release 10477 10.7 1.574 2.363 1.574 2.363 dbcsr_finalize 304 7.8 0.050 0.061 1.790 1.990 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.792000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2744.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c60ebb48d63dcff4bc85c6fc137b78e381e1214e_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 4.667380E+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 3129 15263. MP_Sync 4 MP_Alltoall 47 46208988. MP_SendRecv 45 115200. MP_ISendRecv 45 115200. MP_Wait 528 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.077 0.080 100.328 100.328 qs_energies 1 2.0 0.000 0.000 99.416 99.421 ls_scf 1 3.0 0.000 0.000 97.448 97.458 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.669 78.928 ls_scf_main 1 4.0 0.000 0.000 62.769 62.770 multiply_cannon 111 7.7 0.090 0.146 55.650 61.208 density_matrix_trs4 2 5.0 0.002 0.003 55.515 55.603 multiply_cannon_loop 111 8.7 0.069 0.076 51.234 53.025 mp_waitall_1 5436 11.0 27.170 33.069 27.170 33.069 ls_scf_init_scf 1 4.0 0.000 0.000 31.063 31.068 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.769 29.820 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.594 27.603 multiply_cannon_multrec 444 9.7 13.931 16.550 21.059 22.892 make_m2s 222 7.7 0.004 0.005 18.079 20.484 make_images 222 8.7 3.724 4.423 18.017 20.425 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.657 16.737 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.104 14.871 make_images_data 222 9.7 0.003 0.004 10.207 12.671 hybrid_alltoall_any 227 10.6 0.788 3.760 9.921 12.487 dbcsr_mm_accdrv_process 3003 10.4 0.188 0.346 6.833 7.983 dbcsr_mm_accdrv_process_sort 3003 11.4 6.517 7.636 6.517 7.636 multiply_cannon_sync_h2d 444 9.7 6.518 7.567 6.518 7.567 mp_allgather_i34 111 8.7 2.632 7.020 2.632 7.020 arnoldi_extremal 4 6.8 0.000 0.000 5.937 5.946 arnoldi_normal_ev 4 7.8 0.002 0.005 5.937 5.946 build_subspace 16 8.4 0.015 0.020 5.537 5.546 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.597 4.705 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.707 4.612 mp_sum_l 887 5.1 2.983 4.600 2.983 4.600 mp_irecv_dv 1241 11.2 1.687 4.583 1.687 4.583 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.186 4.327 dbcsr_matrix_vector_mult_local 304 10.0 3.635 4.098 3.637 4.100 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.636 3.744 make_images_sizes 222 9.7 0.000 0.000 1.215 3.740 mp_alltoall_i44 222 10.7 1.214 3.739 1.214 3.739 calculate_norms 792 9.8 3.555 3.705 3.555 3.705 ls_scf_post 1 4.0 0.000 0.000 3.616 3.620 ls_scf_store_result 1 5.0 0.000 0.000 3.375 3.441 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.075 3.425 dbcsr_finalize 304 7.8 0.062 0.078 2.204 2.294 dbcsr_merge_all 275 8.9 0.475 0.536 2.050 2.131 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.095 2.097 rebuild_ks_matrix 3 7.3 0.000 0.000 2.062 2.065 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.062 2.065 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=100.328000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3732.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c60ebb48d63dcff4bc85c6fc137b78e381e1214e_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 8.779280E+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 3123 21388. MP_Sync 4 MP_Alltoall 47 88727262. MP_SendRecv 42 732600. MP_ISendRecv 42 732600. MP_Wait 267 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.081 0.099 115.620 115.620 qs_energies 1 2.0 0.000 0.000 113.997 114.015 ls_scf 1 3.0 0.000 0.000 111.067 111.085 dbcsr_multiply_generic 111 6.7 0.023 0.027 84.137 84.242 ls_scf_main 1 4.0 0.000 0.000 66.494 66.496 multiply_cannon 111 7.7 0.155 0.215 56.315 58.080 density_matrix_trs4 2 5.0 0.002 0.003 57.146 57.210 multiply_cannon_loop 111 8.7 0.067 0.070 52.964 53.875 ls_scf_init_scf 1 4.0 0.000 0.000 40.865 40.865 ls_scf_init_matrix_S 1 5.0 0.000 0.000 39.215 39.223 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 36.292 36.305 mp_waitall_1 4527 11.1 23.831 29.322 23.831 29.322 make_m2s 222 7.7 0.005 0.005 24.195 25.069 make_images 222 8.7 4.585 4.967 24.089 24.963 multiply_cannon_multrec 444 9.7 17.747 18.457 22.405 22.938 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 11.831 17.903 hybrid_alltoall_any 227 10.6 1.661 3.638 13.133 16.055 make_images_data 222 9.7 0.003 0.003 13.387 15.757 multiply_cannon_sync_h2d 444 9.7 8.842 8.892 8.842 8.892 arnoldi_extremal 4 6.8 0.000 0.000 7.534 7.550 arnoldi_normal_ev 4 7.8 0.003 0.009 7.534 7.550 build_subspace 16 8.4 0.026 0.036 6.977 6.987 compute_matrix_preconditioner 1 6.0 0.002 0.002 6.830 6.841 acc_transpose_blocks 444 9.7 0.002 0.002 5.437 6.192 acc_transpose_blocks_kernels 444 10.7 0.005 0.005 5.402 6.156 jit_kernel_transpose 1 13.0 5.397 6.151 5.397 6.151 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.442 5.520 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.120 5.372 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.251 5.344 dbcsr_matrix_vector_mult_local 304 10.0 4.904 5.183 4.907 5.186 dbcsr_mm_accdrv_process 1814 10.4 0.184 0.318 4.488 4.620 dbcsr_mm_accdrv_process_sort 1814 11.4 4.190 4.326 4.190 4.326 ls_scf_post 1 4.0 0.000 0.000 3.708 3.725 make_images_sizes 222 9.7 0.000 0.000 1.450 3.619 mp_alltoall_i44 222 10.7 1.450 3.619 1.450 3.619 ls_scf_store_result 1 5.0 0.000 0.000 3.442 3.449 calculate_norms 792 9.8 3.245 3.287 3.245 3.287 dbcsr_finalize 304 7.8 0.082 0.089 3.088 3.166 mp_allgather_i34 111 8.7 0.995 3.125 0.995 3.125 dbcsr_merge_all 275 8.9 0.890 0.918 2.874 2.946 dbcsr_data_release 12724 10.6 2.323 2.905 2.323 2.905 qs_energies_init_hamiltonians 1 3.0 0.001 0.002 2.899 2.899 dbcsr_complete_redistribute 5 7.6 1.439 1.484 2.793 2.886 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.560 2.562 matrix_ls_to_qs 2 6.0 0.000 0.000 2.441 2.554 dbcsr_sort_data 325 11.1 2.442 2.501 2.442 2.501 rebuild_ks_matrix 3 7.3 0.000 0.000 2.494 2.496 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.494 2.496 dbcsr_new_transposed 4 7.5 0.285 0.334 2.420 2.435 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=115.620000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6814.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c60ebb48d63dcff4bc85c6fc137b78e381e1214e_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 589.905920E+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 13232 2081. MP_Sync 1064 MP_Alltoall 2588 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 0.042 0.168 287.550 287.552 qs_mol_dyn_low 1 2.0 0.005 0.061 286.403 286.423 qs_forces 5 3.8 0.005 0.044 286.239 286.240 qs_energies 5 4.8 0.002 0.058 282.802 282.833 scf_env_do_scf 5 5.8 0.001 0.008 262.578 262.580 scf_env_do_scf_inner_loop 105 6.6 0.002 0.006 228.808 228.809 qs_scf_new_mos 105 7.6 0.000 0.001 179.552 179.739 qs_scf_loop_do_ot 105 8.6 0.001 0.001 179.552 179.738 ot_scf_mini 105 9.6 0.003 0.004 169.592 169.814 dbcsr_multiply_generic 1445 12.2 0.128 0.136 140.999 141.707 multiply_cannon 1445 13.2 0.274 0.287 119.507 122.889 multiply_cannon_loop 1445 14.2 2.393 2.523 116.990 119.098 velocity_verlet 4 3.0 0.003 0.029 118.823 118.825 qs_ot_get_p 112 10.4 0.001 0.002 76.597 76.902 qs_ot_p2m_diag 40 11.0 0.020 0.030 64.187 64.285 ot_mini 105 10.6 0.001 0.002 62.500 62.641 cp_dbcsr_syevd 40 12.0 0.002 0.003 60.750 60.753 cp_fm_syevd 40 13.0 0.000 0.003 54.577 54.729 mp_waitall_1 488190 16.1 38.591 50.871 38.591 50.871 cp_fm_redistribute_end 40 14.0 24.565 49.013 24.573 49.018 cp_fm_syevd_base 40 14.0 24.425 48.873 24.425 48.873 multiply_cannon_multrec 69360 15.2 31.321 36.609 39.006 44.751 qs_ot_get_derivative 55 11.6 0.001 0.006 40.610 40.761 multiply_cannon_metrocomm3 69360 15.2 0.192 0.202 26.464 39.440 multiply_cannon_sync_h2d 69360 15.2 30.034 34.186 30.034 34.186 init_scf_loop 7 6.6 0.000 0.005 33.712 33.714 rebuild_ks_matrix 110 8.4 0.000 0.000 32.915 33.105 qs_ks_build_kohn_sham_matrix 110 9.4 0.012 0.022 32.915 33.105 qs_ks_update_qs_env 112 7.6 0.001 0.001 30.245 30.422 prepare_preconditioner 7 7.6 0.000 0.000 28.044 28.077 make_preconditioner 7 8.6 0.000 0.003 28.044 28.077 qs_rho_update_rho_low 110 7.6 0.001 0.001 23.174 23.416 calculate_rho_elec 110 8.6 0.030 0.033 23.173 23.416 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 23.116 23.398 apply_single 62 13.6 0.000 0.000 23.116 23.398 ot_new_cg_direction 55 11.6 0.001 0.004 21.119 21.120 rs_pw_transfer 690 11.5 0.010 0.012 18.719 20.027 make_full_inverse_cholesky 7 9.6 0.000 0.000 19.683 19.744 density_rs2pw 110 9.6 0.006 0.007 16.769 18.094 init_scf_run 5 5.8 0.000 0.001 17.213 17.214 scf_env_initial_rho_setup 5 6.8 0.001 0.003 17.213 17.214 qs_ot_get_orbitals 105 10.6 0.001 0.001 15.188 15.445 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 14.629 14.739 mp_sum_l 4764 12.2 12.933 13.853 12.933 13.853 pw_transfer 1645 12.4 0.085 0.113 13.574 13.843 fft_wrap_pw1pw2 1425 13.5 0.014 0.017 13.430 13.703 cp_fm_cholesky_invert 7 10.6 13.565 13.579 13.565 13.579 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.002 12.240 12.258 acc_transpose_blocks 69360 15.2 0.365 0.383 11.101 11.852 calculate_first_density_matrix 1 7.0 0.000 0.007 11.749 11.762 calculate_dm_sparse 110 9.5 0.000 0.001 11.530 11.728 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.001 11.645 11.714 sum_up_and_integrate 60 10.3 0.028 0.031 11.635 11.654 integrate_v_rspace 60 11.3 0.001 0.002 11.607 11.626 fft_wrap_pw1pw2_240 915 15.0 1.211 1.310 11.193 11.402 qs_vxc_create 110 10.4 0.002 0.005 11.126 11.181 check_diag 80 13.5 8.586 8.862 10.909 11.060 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 10.652 10.743 multiply_cannon_metrocomm1 69360 15.2 0.091 0.097 5.119 10.473 acc_transpose_blocks_kernels 69360 16.2 0.857 0.899 9.154 9.795 fft3d_pb 915 16.0 2.397 2.698 9.229 9.571 jit_kernel_transpose 5 15.0 8.297 8.916 8.297 8.916 make_m2s 2890 13.2 0.079 0.086 8.153 8.740 make_images 2890 14.2 0.240 0.260 8.044 8.630 dbcsr_mm_accdrv_process 154766 15.8 4.087 4.280 7.557 8.348 make_full_single_inverse 7 9.6 0.001 0.006 7.980 8.013 xc_rho_set_and_dset_create 110 12.4 0.076 0.098 7.609 7.865 xc_vxc_pw_create 60 11.3 0.039 0.049 7.534 7.589 potential_pw2rs 60 12.3 0.003 0.003 7.492 7.535 mp_alltoall_z22v 2340 17.7 6.355 6.840 6.355 6.840 xc_pw_derive 510 13.4 0.006 0.007 6.769 6.839 rs_pw_transfer_RS2PW_30 110 11.6 1.594 1.677 6.535 6.760 mp_sendrecv_dv 168740 12.6 6.218 6.385 6.218 6.385 mp_waitany 7680 13.5 4.637 5.940 4.637 5.940 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=287.552000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=559.600000, yerr=3.322650 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: c60ebb48d63dcff4bc85c6fc137b78e381e1214e Summary: empty Status: OK