=== 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: a007b1b0cc940e469a6f710d029fae2a5232a07b ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, Cray-FFTW 3.3.8.10, # COSMA 2.6.2, ELPA 2022.05.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.0.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.0, # SIRIUS 7.3.2, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Author: Matthias Krack (26.10.2022) # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2022.05.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.0.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.0 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.3.2 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MPI_VERSION=3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta_prefixed_scalapack.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a007b1b0cc940e469a6f710d029fae2a5232a07b_performance_tests/01 job id: 43161074 --- 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/a007b1b0cc940e469a6f710d029fae2a5232a07b_performance_tests/02 job id: 43161075 --- 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/a007b1b0cc940e469a6f710d029fae2a5232a07b_performance_tests/03 job id: 43161076 --- 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/a007b1b0cc940e469a6f710d029fae2a5232a07b_performance_tests/04 job id: 43161077 --- 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/a007b1b0cc940e469a6f710d029fae2a5232a07b_performance_tests/05 job id: 43161078 --- 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/a007b1b0cc940e469a6f710d029fae2a5232a07b_performance_tests/06 job id: 43161079 --- 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/a007b1b0cc940e469a6f710d029fae2a5232a07b_performance_tests/07 job id: 43161080 --- 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/a007b1b0cc940e469a6f710d029fae2a5232a07b_performance_tests/08 job id: 43161082 --- 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/a007b1b0cc940e469a6f710d029fae2a5232a07b_performance_tests/09 job id: 43161083 --- 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/a007b1b0cc940e469a6f710d029fae2a5232a07b_performance_tests/10 job id: 43161084 --- 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/a007b1b0cc940e469a6f710d029fae2a5232a07b_performance_tests/11 job id: 43161086 --- 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/a007b1b0cc940e469a6f710d029fae2a5232a07b_performance_tests/12 job id: 43161087 --- 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/a007b1b0cc940e469a6f710d029fae2a5232a07b_performance_tests/13 job id: 43161088 --- 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/a007b1b0cc940e469a6f710d029fae2a5232a07b_performance_tests/14 job id: 43161089 --- 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/a007b1b0cc940e469a6f710d029fae2a5232a07b_performance_tests/15 job id: 43161090 --- 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/a007b1b0cc940e469a6f710d029fae2a5232a07b_performance_tests/16 job id: 43161091 --- 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/a007b1b0cc940e469a6f710d029fae2a5232a07b_performance_tests/17 job id: 43161092 --- 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/a007b1b0cc940e469a6f710d029fae2a5232a07b_performance_tests/18 job id: 43161093 --- 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/a007b1b0cc940e469a6f710d029fae2a5232a07b_performance_tests/19 job id: 43161094 --- 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/a007b1b0cc940e469a6f710d029fae2a5232a07b_performance_tests/20 job id: 43161095 --- 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/a007b1b0cc940e469a6f710d029fae2a5232a07b_performance_tests/21 job id: 43161096 --- 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/a007b1b0cc940e469a6f710d029fae2a5232a07b_performance_tests/22 job id: 43161097 --- 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/a007b1b0cc940e469a6f710d029fae2a5232a07b_performance_tests/23 job id: 43161099 --- 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/a007b1b0cc940e469a6f710d029fae2a5232a07b_performance_tests/24 job id: 43161100 --- 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/a007b1b0cc940e469a6f710d029fae2a5232a07b_performance_tests/25 job id: 43161101 --- 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/a007b1b0cc940e469a6f710d029fae2a5232a07b_performance_tests/26 job id: 43161102 --- Point --- name: 510 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/1r/12t) --- Point --- name: 511 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ === END TESTS (description) === ===== PLOTS (description) ===== ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2_mem", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md_mem", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md_mem", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md_mem", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls_mem", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" === END PLOTS (description) === ============ RESULTS ============ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a007b1b0cc940e469a6f710d029fae2a5232a07b_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 344 9. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.023 0.035 134.550 134.551 farming_run 1 2.0 133.995 134.035 134.519 134.526 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.460781E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 228 1113141. MP_Allreduce 485 2282278. MP_Sync 27 MP_Alltoall 38 9316958. MP_SendRecv 30 829726. MP_ISendRecv 135 235435. MP_Wait 281 MP_comm_split 8 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.033 116.184 116.185 qs_energies 1 2.0 0.000 0.000 115.841 115.842 mp2_main 1 3.0 0.000 0.000 113.795 113.797 mp2_gpw_main 1 4.0 0.023 0.033 112.880 112.882 mp2_ri_gpw_compute_in 1 5.0 0.174 0.179 93.791 94.153 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 55.565 55.929 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.167 41.882 47.166 get_2c_integrals 1 6.0 0.000 0.000 37.424 38.057 integrate_v_rspace 273 8.0 0.437 0.450 25.243 30.371 pw_transfer 6555 10.6 0.376 0.398 27.467 27.965 grid_integrate_task_list 273 9.0 21.035 26.665 21.035 26.665 fft_wrap_pw1pw2 5465 11.4 0.044 0.045 26.127 26.639 fft_wrap_pw1pw2_100 2178 12.4 1.204 1.434 23.691 24.204 compute_2c_integrals 1 7.0 0.011 0.012 19.689 19.696 compute_2c_integrals_loop_lm 1 8.0 0.003 0.003 18.935 19.371 mp2_eri_2c_integrate_gpw 1 9.0 2.364 2.422 18.932 19.370 rpa_ri_compute_en 1 5.0 0.000 0.000 18.978 19.233 cp_fm_cholesky_decompose 12 8.2 17.777 18.395 17.777 18.395 cholesky_decomp 1 7.0 0.000 0.000 16.584 17.219 fft3d_s 5443 13.4 16.182 16.644 16.204 16.665 ao_to_mo_and_store_B_mult_1 272 7.0 10.853 15.595 10.853 15.595 calculate_wavefunction 272 8.0 5.397 5.540 12.560 13.147 rpa_num_int 1 6.0 0.000 0.000 10.709 10.718 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.612 10.710 calc_mat_Q 8 8.0 0.000 0.000 9.375 9.557 contract_S_to_Q 8 9.0 0.000 0.000 8.797 8.977 calc_potential_gpw 544 9.5 0.005 0.006 8.280 8.648 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.260 8.567 parallel_gemm_fm 14 9.1 0.000 0.000 8.381 8.551 parallel_gemm_fm_cosma 14 10.1 8.381 8.551 8.381 8.551 potential_pw2rs 545 10.0 0.107 0.109 7.724 8.524 create_integ_mat 1 6.0 0.015 0.028 7.810 7.819 collocate_single_gaussian 272 10.0 0.040 0.043 7.482 7.756 array2fm 1 7.0 0.000 0.000 6.623 7.273 pw_scatter_s 2720 13.7 4.439 4.666 4.439 4.666 pw_gather_s 2722 13.2 3.933 4.255 3.933 4.255 array2fm_buffer_send 1 8.0 2.927 3.227 2.927 3.227 pw_poisson_solve 545 10.5 1.124 1.177 2.169 2.354 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.880905, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2731.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a007b1b0cc940e469a6f710d029fae2a5232a07b_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 205321. MP_Allreduce 344 10. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.030 0.038 400.504 400.505 farming_run 1 2.0 399.684 399.689 400.455 400.457 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827141120 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788822 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.225662E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 703 408373. MP_Allreduce 1821 23730. MP_Sync 38 MP_Alltoall 77 MP_SendRecv 2171 2843495. MP_ISendRecv 1739 144022. MP_Wait 2051 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.041 210.870 210.870 qs_energies 1 2.0 0.000 0.000 210.638 210.649 scf_env_do_scf 1 3.0 0.000 0.000 106.726 106.726 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.741 105.749 rebuild_ks_matrix 4 6.0 0.000 0.000 105.740 105.748 qs_ks_build_kohn_sham_matrix 4 7.0 0.057 0.065 105.740 105.748 hfx_ks_matrix 4 8.0 0.001 0.001 105.341 105.344 integrate_four_center 4 9.0 0.143 0.452 105.340 105.343 mp2_main 1 3.0 0.000 0.000 103.608 103.619 mp2_gpw_main 1 4.0 0.037 0.055 102.455 102.471 integrate_four_center_main 4 10.0 0.090 0.553 96.821 99.540 integrate_four_center_bin 266 11.0 96.730 99.489 96.730 99.489 init_scf_loop 1 4.0 0.000 0.000 92.388 92.388 mp2_ri_gpw_compute_in 1 5.0 0.070 0.093 74.980 76.104 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.557 55.679 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.164 42.318 47.439 integrate_v_rspace 95 8.0 0.398 0.571 28.710 33.647 pw_transfer 2240 10.6 0.146 0.162 29.847 30.428 ao_to_mo_and_store_B_mult_1 91 7.0 10.544 29.473 10.544 29.473 fft_wrap_pw1pw2 1868 11.4 0.018 0.020 28.875 29.384 grid_integrate_task_list 95 9.0 23.989 29.122 23.989 29.122 mp2_ri_gpw_compute_en 1 5.0 0.060 0.075 27.313 29.078 fft_wrap_pw1pw2_100 730 12.4 1.243 1.436 26.598 26.950 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.834 1.897 25.545 25.556 get_2c_integrals 1 6.0 0.009 0.084 20.285 20.358 compute_2c_integrals 1 7.0 0.009 0.021 19.248 19.267 compute_2c_integrals_loop_lm 1 8.0 0.002 0.008 18.741 19.087 mp2_eri_2c_integrate_gpw 1 9.0 1.742 1.869 18.739 19.086 fft3d_s 1823 13.4 18.404 18.705 18.417 18.718 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.335 14.335 calculate_wavefunction 91 8.0 2.016 2.041 9.740 9.983 mp2_ri_gpw_compute_en_expansio 172 7.0 0.557 0.580 8.823 9.276 potential_pw2rs 186 10.0 0.033 0.035 8.620 9.239 mp2_ri_gpw_compute_en_comm 22 7.0 0.484 0.499 8.504 9.123 local_gemm 172 8.0 8.266 8.709 8.266 8.709 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.177 8.510 calc_potential_gpw 182 9.5 0.002 0.002 7.875 8.164 collocate_single_gaussian 91 10.0 0.017 0.030 7.826 8.049 mp_sendrecv_dm3 2068 8.0 6.584 7.195 6.584 7.195 mp2_ri_gpw_compute_en_ener 172 7.0 6.340 6.441 6.340 6.441 mp_sync 38 10.4 3.222 6.255 3.222 6.255 pw_gather_s 912 13.2 4.930 5.350 4.930 5.350 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=102.446494, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1513.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a007b1b0cc940e469a6f710d029fae2a5232a07b_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 452.005888E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62385. MP_Allreduce 10249 271. MP_Sync 580 MP_Alltoall 2083 902740. MP_ISendRecv 45220 5520. MP_Wait 60486 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.036 54.866 54.866 qs_mol_dyn_low 1 2.0 0.003 0.003 54.560 54.568 qs_forces 11 3.9 0.001 0.002 54.494 54.495 qs_energies 11 4.9 0.004 0.021 52.947 52.960 scf_env_do_scf 11 5.9 0.000 0.001 45.753 45.754 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 43.190 43.191 dbcsr_multiply_generic 2286 12.5 0.103 0.112 33.664 34.064 qs_scf_new_mos 108 7.5 0.000 0.001 31.868 32.128 qs_scf_loop_do_ot 108 8.5 0.000 0.001 31.867 32.127 ot_scf_mini 108 9.5 0.002 0.002 30.221 30.399 multiply_cannon 2286 13.5 0.191 0.207 26.336 27.845 multiply_cannon_loop 2286 14.5 1.515 1.587 25.393 26.990 velocity_verlet 10 3.0 0.001 0.001 26.655 26.655 ot_mini 108 10.5 0.001 0.001 19.244 19.487 qs_ot_get_derivative 108 11.5 0.001 0.001 16.236 16.405 mp_waitall_1 267858 16.1 9.214 15.329 9.214 15.329 multiply_cannon_metrocomm3 54864 15.5 0.067 0.073 5.666 12.902 multiply_cannon_multrec 54864 15.5 4.225 6.674 8.000 11.922 rebuild_ks_matrix 119 8.3 0.000 0.000 8.845 8.974 qs_ks_build_kohn_sham_matrix 119 9.3 0.015 0.032 8.845 8.974 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.850 7.973 qs_ot_get_p 119 10.4 0.001 0.001 6.404 6.666 multiply_cannon_sync_h2d 54864 15.5 5.858 6.524 5.858 6.524 mp_sum_l 7207 12.9 4.686 6.217 4.686 6.217 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.514 5.990 init_scf_run 11 5.9 0.000 0.001 5.721 5.722 scf_env_initial_rho_setup 11 6.9 0.003 0.007 5.721 5.722 dbcsr_mm_accdrv_process 76910 16.1 1.186 1.828 3.697 5.485 sum_up_and_integrate 119 10.3 0.012 0.014 5.141 5.150 integrate_v_rspace 119 11.3 0.002 0.002 5.129 5.139 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 4.944 5.041 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.908 5.007 calculate_rho_elec 119 8.7 0.011 0.016 4.908 5.006 rs_pw_transfer 974 11.9 0.011 0.012 3.524 3.642 jit_kernel_multiply 13 15.8 2.450 3.615 2.450 3.615 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.524 3.553 calculate_first_density_matrix 1 7.0 0.001 0.011 3.371 3.380 calculate_dm_sparse 119 9.5 0.000 0.000 3.075 3.204 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.945 3.151 apply_single 119 13.6 0.000 0.000 2.945 3.150 density_rs2pw 119 9.7 0.004 0.004 2.891 3.017 multiply_cannon_metrocomm1 54864 15.5 0.051 0.057 1.870 3.001 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.763 2.765 ot_diis_step 108 11.5 0.006 0.006 2.748 2.749 cp_dbcsr_syevd 50 12.0 0.002 0.003 2.672 2.673 init_scf_loop 11 6.9 0.000 0.000 2.536 2.538 potential_pw2rs 119 12.3 0.004 0.004 2.397 2.427 pw_transfer 1439 11.6 0.051 0.055 2.327 2.419 make_m2s 4572 13.5 0.054 0.056 2.339 2.406 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.323 2.385 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.360 2.361 fft_wrap_pw1pw2 1201 12.6 0.006 0.007 2.252 2.345 cp_fm_redistribute_end 50 14.0 2.135 2.331 2.140 2.333 make_images 4572 14.5 0.133 0.139 2.255 2.321 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.246 2.303 cp_fm_diag_elpa_base 50 14.0 0.190 2.282 0.191 2.291 mp_sum_d 4125 12.0 1.621 2.289 1.621 2.289 wfi_extrapolate 11 7.9 0.001 0.001 2.243 2.243 acc_transpose_blocks 54864 15.5 0.232 0.248 1.789 2.227 grid_integrate_task_list 119 12.3 2.043 2.152 2.043 2.152 fft3d_ps 1201 14.6 0.361 0.466 2.033 2.123 mp_alltoall_d11v 2130 13.8 1.706 1.866 1.706 1.866 fft_wrap_pw1pw2_140 487 13.2 0.077 0.092 1.725 1.816 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.600 1.628 mp_waitany 12084 13.8 1.369 1.545 1.369 1.545 grid_collocate_task_list 119 9.7 1.286 1.356 1.286 1.356 dbcsr_dot_sd 1205 11.9 0.047 0.057 0.892 1.337 make_images_sizes 4572 15.5 0.004 0.004 0.970 1.244 mp_alltoall_i44 4572 16.5 0.966 1.240 0.966 1.240 mp_alltoall_z22v 1201 16.6 1.174 1.236 1.174 1.236 prepare_preconditioner 11 7.9 0.000 0.000 1.193 1.225 make_preconditioner 11 8.9 0.000 0.000 1.193 1.225 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.125 1.163 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=54.866000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=430.909091, yerr=1.504813 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a007b1b0cc940e469a6f710d029fae2a5232a07b_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 488.845312E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2194560 MPI messages size (bytes): total size 310.646604E+09 min size 0.000000E+00 max size 1.145520E+06 average size 141.553031E+03 MPI breakdown and total messages size (bytes): size <= 128 724648 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 281952 4619501568 32768 < size <= 131072 494448 39143342080 131072 < size <= 4194304 440000 264807943488 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62664. MP_Allreduce 10226 305. MP_Sync 104 MP_Alltoall 2060 1028288. MP_ISendRecv 33558 37093. MP_Wait 40318 MP_comm_split 50 MP_ISend 5720 128509. MP_IRecv 5720 128509. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.035 0.058 41.399 41.400 qs_mol_dyn_low 1 2.0 0.003 0.004 41.061 41.069 qs_forces 11 3.9 0.002 0.002 40.998 40.999 qs_energies 11 4.9 0.012 0.027 39.106 39.109 scf_env_do_scf 11 5.9 0.001 0.001 33.609 33.609 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 30.756 30.759 dbcsr_multiply_generic 2286 12.5 0.101 0.105 22.061 22.495 qs_scf_new_mos 108 7.5 0.001 0.001 20.907 21.148 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.906 21.147 ot_scf_mini 108 9.5 0.003 0.003 19.966 20.132 velocity_verlet 10 3.0 0.001 0.001 19.629 19.630 multiply_cannon 2286 13.5 0.210 0.217 16.526 18.331 multiply_cannon_loop 2286 14.5 0.907 0.978 15.304 16.935 ot_mini 108 10.5 0.001 0.001 12.324 12.558 mp_waitall_1 217478 16.2 6.915 11.937 6.915 11.937 qs_ot_get_derivative 108 11.5 0.001 0.001 9.820 9.993 multiply_cannon_metrocomm3 27432 15.5 0.067 0.068 4.268 9.680 multiply_cannon_multrec 27432 15.5 1.959 4.425 5.857 8.724 rebuild_ks_matrix 119 8.3 0.000 0.000 7.763 7.903 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 7.763 7.903 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.775 6.904 dbcsr_mm_accdrv_process 47894 16.0 2.945 4.982 3.829 5.696 qs_ot_get_p 119 10.4 0.001 0.001 4.654 4.877 sum_up_and_integrate 119 10.3 0.024 0.027 4.692 4.722 integrate_v_rspace 119 11.3 0.002 0.003 4.667 4.699 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.773 4.634 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.147 4.180 calculate_rho_elec 119 8.7 0.021 0.024 4.147 4.180 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.073 4.168 apply_single 119 13.6 0.000 0.000 3.072 4.168 mp_sum_l 7207 12.9 2.264 4.138 2.264 4.138 init_scf_run 11 5.9 0.000 0.001 4.095 4.095 scf_env_initial_rho_setup 11 6.9 0.001 0.002 4.095 4.095 rs_pw_transfer 974 11.9 0.009 0.011 3.253 3.412 make_m2s 4572 13.5 0.053 0.054 2.832 3.090 make_images 4572 14.5 0.199 0.235 2.742 2.998 qs_ot_p2m_diag 50 11.0 0.008 0.013 2.970 2.994 init_scf_loop 11 6.9 0.001 0.004 2.823 2.829 multiply_cannon_sync_h2d 27432 15.5 2.149 2.802 2.149 2.802 density_rs2pw 119 9.7 0.004 0.004 2.470 2.584 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.504 2.504 potential_pw2rs 119 12.3 0.006 0.006 2.394 2.445 ot_diis_step 108 11.5 0.011 0.011 2.443 2.444 calculate_first_density_matrix 1 7.0 0.001 0.002 2.407 2.409 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.307 2.396 calculate_dm_sparse 119 9.5 0.000 0.000 2.100 2.176 pw_transfer 1439 11.6 0.063 0.066 2.109 2.168 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.088 2.090 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 2.020 2.080 cp_fm_redistribute_end 50 14.0 1.708 2.050 1.713 2.052 cp_fm_diag_elpa_base 50 14.0 0.328 1.989 0.337 2.020 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.996 1.999 grid_integrate_task_list 119 12.3 1.835 1.945 1.835 1.945 prepare_preconditioner 11 7.9 0.000 0.000 1.815 1.841 make_preconditioner 11 8.9 0.000 0.001 1.815 1.841 fft3d_ps 1201 14.6 0.504 0.555 1.728 1.784 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.733 1.774 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.712 1.765 make_images_data 4572 15.5 0.044 0.052 1.246 1.726 jit_kernel_multiply 9 16.1 0.831 1.726 0.831 1.726 wfi_extrapolate 11 7.9 0.001 0.001 1.626 1.626 hybrid_alltoall_any 4725 16.4 0.051 0.112 1.093 1.622 acc_transpose_blocks 27432 15.5 0.109 0.115 1.228 1.572 mp_alltoall_d11v 2130 13.8 1.395 1.561 1.395 1.561 fft_wrap_pw1pw2_140 487 13.2 0.075 0.082 1.495 1.555 mp_allgather_i34 2286 14.5 0.664 1.436 0.664 1.436 grid_collocate_task_list 119 9.7 1.242 1.319 1.242 1.319 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.249 1.293 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.279 1.286 mp_sum_d 4125 12.0 0.697 1.153 0.697 1.153 make_images_sizes 4572 15.5 0.005 0.005 0.776 1.085 mp_alltoall_i44 4572 16.5 0.771 1.080 0.771 1.080 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.049 1.062 mp_alltoall_z22v 1201 16.6 0.944 1.045 0.944 1.045 qs_energies_init_hamiltonians 11 5.9 0.002 0.006 0.977 0.983 acc_transpose_blocks_kernels 27432 16.5 0.182 0.270 0.691 0.944 rs_pw_transfer_PW2RS_50 119 14.3 0.588 0.610 0.877 0.934 cp_fm_cholesky_invert 11 10.9 0.852 0.858 0.852 0.858 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=41.400000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=465.818182, yerr=1.266217 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a007b1b0cc940e469a6f710d029fae2a5232a07b_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 519.688192E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62660. MP_Allreduce 10225 303. MP_Sync 104 MP_Alltoall 1821 1607811. MP_ISendRecv 22134 57667. MP_Wait 33054 MP_comm_split 50 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.089 0.251 35.810 35.814 qs_mol_dyn_low 1 2.0 0.003 0.005 34.033 34.041 qs_forces 11 3.9 0.002 0.002 33.944 33.974 qs_energies 11 4.9 0.004 0.022 32.340 32.372 scf_env_do_scf 11 5.9 0.001 0.001 26.715 26.715 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 23.919 23.920 dbcsr_multiply_generic 2286 12.5 0.099 0.104 17.095 17.255 velocity_verlet 10 3.0 0.001 0.001 16.134 16.136 qs_scf_new_mos 108 7.5 0.001 0.001 15.533 15.554 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.533 15.554 ot_scf_mini 108 9.5 0.002 0.003 14.776 14.791 multiply_cannon 2286 13.5 0.196 0.208 13.323 14.329 multiply_cannon_loop 2286 14.5 0.643 0.672 12.450 13.532 ot_mini 108 10.5 0.001 0.001 9.215 9.236 qs_ot_get_derivative 108 11.5 0.001 0.001 7.627 7.638 multiply_cannon_multrec 18288 15.5 1.915 2.863 6.794 7.042 rebuild_ks_matrix 119 8.3 0.000 0.000 6.596 6.626 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.595 6.626 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.839 5.868 dbcsr_mm_accdrv_process 38222 16.0 4.095 5.321 4.795 5.667 mp_waitall_1 169478 16.3 3.530 4.798 3.530 4.798 init_scf_run 11 5.9 0.000 0.001 4.367 4.370 scf_env_initial_rho_setup 11 6.9 0.005 0.033 4.367 4.369 sum_up_and_integrate 119 10.3 0.030 0.031 4.179 4.184 integrate_v_rspace 119 11.3 0.002 0.003 4.149 4.157 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.482 3.497 calculate_rho_elec 119 8.7 0.030 0.031 3.482 3.496 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.831 3.471 qs_ot_get_p 119 10.4 0.001 0.001 3.295 3.311 calculate_first_density_matrix 1 7.0 0.001 0.003 3.021 3.029 init_scf_loop 11 6.9 0.001 0.002 2.774 2.775 rs_pw_transfer 974 11.9 0.009 0.010 2.615 2.764 multiply_cannon_metrocomm3 18288 15.5 0.044 0.045 1.553 2.641 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.109 2.434 apply_single 119 13.6 0.000 0.000 2.109 2.434 make_m2s 4572 13.5 0.045 0.046 2.065 2.250 make_images 4572 14.5 0.191 0.203 1.979 2.164 density_rs2pw 119 9.7 0.004 0.004 2.050 2.151 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.086 2.096 prepare_preconditioner 11 7.9 0.000 0.000 1.953 1.956 make_preconditioner 11 8.9 0.000 0.000 1.953 1.956 potential_pw2rs 119 12.3 0.007 0.008 1.944 1.954 jit_kernel_multiply 10 16.2 0.648 1.927 0.648 1.927 grid_integrate_task_list 119 12.3 1.793 1.917 1.793 1.917 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.906 1.908 calculate_dm_sparse 119 9.5 0.000 0.000 1.881 1.894 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.801 1.884 pw_transfer 1439 11.6 0.063 0.066 1.863 1.874 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.796 1.797 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.772 1.785 mp_sum_l 7207 12.9 1.348 1.763 1.348 1.763 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.712 1.718 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.626 1.635 multiply_cannon_sync_h2d 18288 15.5 1.344 1.594 1.344 1.594 ot_diis_step 108 11.5 0.011 0.011 1.550 1.552 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.525 1.526 cp_fm_redistribute_end 50 14.0 1.128 1.501 1.130 1.501 cp_fm_diag_elpa_base 50 14.0 0.355 1.440 0.369 1.477 fft3d_ps 1201 14.6 0.508 0.524 1.460 1.473 fft_wrap_pw1pw2_140 487 13.2 0.086 0.089 1.342 1.357 grid_collocate_task_list 119 9.7 1.204 1.296 1.204 1.296 acc_transpose_blocks 18288 15.5 0.075 0.078 1.260 1.286 wfi_extrapolate 11 7.9 0.001 0.001 1.285 1.285 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.221 1.227 make_images_data 4572 15.5 0.045 0.048 0.920 1.140 hybrid_alltoall_any 4725 16.4 0.055 0.114 0.788 1.019 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.975 1.005 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.956 0.976 mp_alltoall_d11v 2130 13.8 0.805 0.975 0.805 0.975 create_qs_kind_set 1 2.0 0.005 0.017 0.764 0.938 read_qs_kind 2 3.0 0.040 0.122 0.759 0.938 parallel_gemm_fm 81 9.0 0.000 0.000 0.930 0.932 parallel_gemm_fm_cosma 81 10.0 0.930 0.932 0.930 0.932 parser_read_line 2807 4.0 0.001 0.001 0.719 0.930 parser_read_line_low 5 5.0 0.002 0.010 0.719 0.930 broadcast_input_information 5 6.0 0.000 0.000 0.717 0.929 mp_alltoall_z22v 1201 16.6 0.778 0.852 0.778 0.852 cp_fm_cholesky_invert 11 10.9 0.841 0.847 0.841 0.847 make_basis_sm 11 9.8 0.000 0.000 0.840 0.843 mp_bcast_am 5 7.0 0.336 0.842 0.337 0.842 acc_transpose_blocks_kernels 18288 16.5 0.212 0.221 0.815 0.834 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.827 0.830 mp_bcast_i 105 3.0 0.382 0.816 0.382 0.816 dbcsr_complete_redistribute 329 12.2 0.115 0.196 0.655 0.767 mp_waitany 9880 13.7 0.586 0.761 0.586 0.761 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.650 0.733 rs_pw_transfer_RS2PW_140 130 11.5 0.120 0.131 0.570 0.730 qs_init_subsys 1 2.0 0.069 0.127 0.646 0.728 mp_allgather_i34 2286 14.5 0.334 0.718 0.334 0.718 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=35.814000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=495.363636, yerr=1.067940 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a007b1b0cc940e469a6f710d029fae2a5232a07b_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 549.937152E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62659. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 2412273. MP_ISendRecv 16422 74133. MP_Wait 24482 MP_comm_split 50 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.076 0.290 36.353 36.354 qs_mol_dyn_low 1 2.0 0.003 0.004 35.814 35.821 qs_forces 11 3.9 0.002 0.002 35.753 35.754 qs_energies 11 4.9 0.002 0.004 33.990 33.995 scf_env_do_scf 11 5.9 0.001 0.001 28.774 28.775 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 25.133 25.134 dbcsr_multiply_generic 2286 12.5 0.101 0.104 18.573 18.693 velocity_verlet 10 3.0 0.002 0.007 18.350 18.352 qs_scf_new_mos 108 7.5 0.001 0.001 16.699 16.757 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.698 16.757 ot_scf_mini 108 9.5 0.003 0.003 15.739 15.793 multiply_cannon 2286 13.5 0.231 0.274 14.560 15.043 multiply_cannon_loop 2286 14.5 0.945 0.978 13.513 13.934 ot_mini 108 10.5 0.001 0.001 9.586 9.649 multiply_cannon_multrec 27432 15.5 2.335 3.027 8.613 8.958 qs_ot_get_derivative 108 11.5 0.001 0.001 7.761 7.815 dbcsr_mm_accdrv_process 47916 15.9 5.280 7.167 6.182 7.375 rebuild_ks_matrix 119 8.3 0.000 0.000 6.723 6.776 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.723 6.775 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.959 6.005 sum_up_and_integrate 119 10.3 0.035 0.038 3.972 3.978 integrate_v_rspace 119 11.3 0.002 0.003 3.937 3.944 init_scf_run 11 5.9 0.000 0.001 3.820 3.821 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.820 3.821 init_scf_loop 11 6.9 0.001 0.002 3.618 3.619 qs_ot_get_p 119 10.4 0.001 0.001 3.488 3.562 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.404 3.439 calculate_rho_elec 119 8.7 0.040 0.046 3.404 3.438 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.825 3.292 mp_waitall_1 145218 16.4 2.448 3.044 2.448 3.044 prepare_preconditioner 11 7.9 0.000 0.000 2.752 2.760 make_preconditioner 11 8.9 0.000 0.001 2.752 2.760 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.359 2.687 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.135 2.598 apply_single 119 13.6 0.000 0.000 2.135 2.598 make_m2s 4572 13.5 0.054 0.057 2.426 2.547 make_images 4572 14.5 0.271 0.332 2.318 2.436 rs_pw_transfer 974 11.9 0.008 0.009 2.261 2.419 calculate_first_density_matrix 1 7.0 0.001 0.002 2.380 2.382 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.117 2.130 calculate_dm_sparse 119 9.5 0.000 0.000 2.055 2.114 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.005 2.032 density_rs2pw 119 9.7 0.003 0.004 1.877 2.011 mp_sum_l 7207 12.9 1.328 1.930 1.328 1.930 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.920 1.921 grid_integrate_task_list 119 12.3 1.828 1.905 1.828 1.905 pw_transfer 1439 11.6 0.063 0.067 1.835 1.869 jit_kernel_multiply 10 16.0 0.841 1.838 0.841 1.838 ot_diis_step 108 11.5 0.012 0.012 1.784 1.785 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 1.745 1.782 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.773 1.774 potential_pw2rs 119 12.3 0.008 0.009 1.730 1.743 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.726 1.739 acc_transpose_blocks 27432 15.5 0.111 0.113 1.462 1.504 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.491 1.492 cp_fm_redistribute_end 50 14.0 0.979 1.461 0.980 1.462 fft3d_ps 1201 14.6 0.535 0.586 1.427 1.452 cp_fm_diag_elpa_base 50 14.0 0.458 1.398 0.479 1.439 fft_wrap_pw1pw2_140 487 13.2 0.084 0.092 1.393 1.432 wfi_extrapolate 11 7.9 0.001 0.001 1.386 1.386 grid_collocate_task_list 119 9.7 1.221 1.324 1.221 1.324 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.778 1.248 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.219 1.229 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.145 1.164 cp_fm_upper_to_full 72 13.5 0.816 1.129 0.816 1.129 dbcsr_complete_redistribute 329 12.2 0.121 0.145 0.847 1.127 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.126 1.127 make_images_data 4572 15.5 0.044 0.048 0.942 1.085 multiply_cannon_sync_h2d 27432 15.5 0.985 1.044 0.985 1.044 mp_alltoall_d11v 2130 13.8 0.846 1.024 0.846 1.024 hybrid_alltoall_any 4725 16.4 0.062 0.150 0.812 1.007 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.596 0.872 cp_fm_cholesky_invert 11 10.9 0.866 0.870 0.866 0.870 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.792 0.869 acc_transpose_blocks_kernels 27432 16.5 0.269 0.278 0.835 0.868 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.855 0.860 mp_alltoall_z22v 1201 16.6 0.774 0.799 0.774 0.799 mp_alltoall_i22 627 13.8 0.452 0.764 0.452 0.764 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=36.354000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=520.181818, yerr=3.638636 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a007b1b0cc940e469a6f710d029fae2a5232a07b_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 605.437952E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62658. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 3682667. MP_ISendRecv 10710 94533. MP_Wait 16690 MP_comm_split 50 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.042 29.208 29.209 qs_mol_dyn_low 1 2.0 0.003 0.004 29.022 29.030 qs_forces 11 3.9 0.002 0.002 28.959 28.960 qs_energies 11 4.9 0.001 0.002 27.219 27.222 scf_env_do_scf 11 5.9 0.000 0.001 22.306 22.306 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 19.575 19.576 velocity_verlet 10 3.0 0.001 0.001 15.073 15.076 dbcsr_multiply_generic 2286 12.5 0.093 0.096 12.724 12.819 qs_scf_new_mos 108 7.5 0.001 0.001 11.643 11.664 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.642 11.663 ot_scf_mini 108 9.5 0.002 0.002 10.934 10.966 multiply_cannon 2286 13.5 0.230 0.235 9.842 10.396 multiply_cannon_loop 2286 14.5 0.334 0.347 8.844 9.057 ot_mini 108 10.5 0.001 0.001 6.059 6.099 rebuild_ks_matrix 119 8.3 0.000 0.000 6.068 6.091 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.068 6.090 multiply_cannon_multrec 9144 15.5 1.584 1.836 5.799 6.023 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.406 5.426 qs_ot_get_derivative 108 11.5 0.001 0.001 4.712 4.744 dbcsr_mm_accdrv_process 12550 15.8 3.290 3.880 4.113 4.195 sum_up_and_integrate 119 10.3 0.037 0.041 3.755 3.758 integrate_v_rspace 119 11.3 0.002 0.003 3.717 3.721 init_scf_run 11 5.9 0.000 0.001 3.462 3.462 scf_env_initial_rho_setup 11 6.9 0.001 0.002 3.462 3.462 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.403 3.413 calculate_rho_elec 119 8.7 0.059 0.061 3.402 3.413 qs_ot_get_p 119 10.4 0.001 0.001 2.996 3.041 mp_waitall_1 121218 16.5 2.226 2.758 2.226 2.758 init_scf_loop 11 6.9 0.000 0.000 2.708 2.709 calculate_first_density_matrix 1 7.0 0.000 0.000 2.254 2.256 make_m2s 4572 13.5 0.035 0.036 1.996 2.212 make_images 4572 14.5 0.266 0.300 1.906 2.120 rs_pw_transfer 974 11.9 0.008 0.008 1.909 2.003 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.989 1.992 prepare_preconditioner 11 7.9 0.000 0.000 1.929 1.935 make_preconditioner 11 8.9 0.000 0.000 1.929 1.935 grid_integrate_task_list 119 12.3 1.845 1.904 1.845 1.904 density_rs2pw 119 9.7 0.003 0.004 1.805 1.899 pw_transfer 1439 11.6 0.063 0.065 1.870 1.881 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.808 1.839 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.780 1.791 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.780 1.781 calculate_dm_sparse 119 9.5 0.000 0.000 1.736 1.755 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.684 1.685 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.617 1.638 potential_pw2rs 119 12.3 0.010 0.010 1.525 1.529 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.489 1.497 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.479 1.480 fft3d_ps 1201 14.6 0.541 0.552 1.448 1.460 cp_fm_redistribute_end 50 14.0 0.736 1.455 0.737 1.456 jit_kernel_multiply 8 15.8 0.784 1.437 0.784 1.437 cp_fm_diag_elpa_base 50 14.0 0.672 1.378 0.716 1.433 fft_wrap_pw1pw2_140 487 13.2 0.081 0.086 1.404 1.414 grid_collocate_task_list 119 9.7 1.273 1.340 1.273 1.340 ot_diis_step 108 11.5 0.013 0.013 1.331 1.331 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.274 1.289 make_images_data 4572 15.5 0.039 0.042 0.973 1.249 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.193 1.228 apply_single 119 13.6 0.000 0.000 1.193 1.228 hybrid_alltoall_any 4725 16.4 0.062 0.176 0.889 1.220 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.216 1.218 wfi_extrapolate 11 7.9 0.001 0.001 1.161 1.161 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.132 1.138 cp_fm_cholesky_invert 11 10.9 1.005 1.008 1.005 1.008 acc_transpose_blocks 9144 15.5 0.038 0.039 0.929 0.956 mp_alltoall_d11v 2130 13.8 0.890 0.950 0.890 0.950 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.867 0.923 mp_allgather_i34 2286 14.5 0.331 0.831 0.331 0.831 mp_alltoall_z22v 1201 16.6 0.781 0.813 0.781 0.813 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.787 0.790 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.781 0.790 multiply_cannon_sync_h2d 9144 15.5 0.709 0.782 0.709 0.782 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.675 0.722 acc_transpose_blocks_kernels 9144 16.5 0.117 0.120 0.686 0.711 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.375 0.680 dbcsr_complete_redistribute 329 12.2 0.163 0.182 0.613 0.642 mp_sum_l 7207 12.9 0.466 0.636 0.466 0.636 jit_kernel_transpose 5 15.6 0.569 0.595 0.569 0.595 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=29.209000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=570.272727, yerr=7.097899 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a007b1b0cc940e469a6f710d029fae2a5232a07b_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 757.067776E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63729. MP_Allreduce 10074 433. MP_Sync 54 MP_Alltoall 1582 7383731. MP_ISendRecv 4998 189067. MP_Wait 8898 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.038 41.929 41.930 qs_mol_dyn_low 1 2.0 0.003 0.004 41.682 41.691 qs_forces 11 3.9 0.002 0.003 41.619 41.621 qs_energies 11 4.9 0.001 0.001 39.655 39.658 scf_env_do_scf 11 5.9 0.001 0.001 33.764 33.764 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 25.602 25.603 velocity_verlet 10 3.0 0.001 0.001 23.629 23.634 dbcsr_multiply_generic 2286 12.5 0.103 0.104 17.911 18.148 qs_scf_new_mos 108 7.5 0.001 0.001 16.429 16.518 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.428 16.517 ot_scf_mini 108 9.5 0.002 0.002 15.326 15.420 multiply_cannon 2286 13.5 0.304 0.311 13.756 14.833 multiply_cannon_loop 2286 14.5 0.346 0.352 12.428 13.479 ot_mini 108 10.5 0.001 0.001 9.051 9.168 multiply_cannon_multrec 9144 15.5 3.402 4.721 8.629 8.737 init_scf_loop 11 6.9 0.000 0.000 8.133 8.136 prepare_preconditioner 11 7.9 0.000 0.000 7.195 7.210 make_preconditioner 11 8.9 0.000 0.000 7.195 7.210 rebuild_ks_matrix 119 8.3 0.000 0.000 7.006 7.146 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.006 7.145 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.783 7.080 qs_ot_get_derivative 108 11.5 0.001 0.001 6.982 7.077 dbcsr_mm_accdrv_process 12550 15.8 3.910 5.289 5.103 6.471 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.337 6.463 cp_fm_upper_to_full 72 14.2 3.206 4.552 3.206 4.552 mp_waitall_1 97218 16.6 2.807 3.947 2.807 3.947 init_scf_run 11 5.9 0.000 0.001 3.872 3.872 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.872 3.872 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.822 3.842 calculate_rho_elec 119 8.7 0.118 0.121 3.821 3.841 sum_up_and_integrate 119 10.3 0.064 0.065 3.806 3.811 integrate_v_rspace 119 11.3 0.003 0.003 3.741 3.747 qs_ot_get_p 119 10.4 0.001 0.001 3.452 3.580 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.534 2.988 dbcsr_complete_redistribute 329 12.2 0.288 0.292 2.016 2.839 make_m2s 4572 13.5 0.039 0.039 2.482 2.710 make_images 4572 14.5 0.350 0.380 2.360 2.588 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.256 2.503 apply_single 119 13.6 0.000 0.000 2.256 2.502 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.662 2.487 calculate_first_density_matrix 1 7.0 0.000 0.000 2.405 2.409 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 1.353 2.389 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.408 2.227 calculate_dm_sparse 119 9.5 0.000 0.000 2.189 2.204 mp_alltoall_i22 627 13.8 1.380 2.181 1.380 2.181 pw_transfer 1439 11.6 0.066 0.066 2.152 2.157 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.123 2.125 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.057 2.061 grid_integrate_task_list 119 12.3 2.033 2.052 2.033 2.052 ot_diis_step 108 11.5 0.014 0.014 2.038 2.038 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.944 1.945 density_rs2pw 119 9.7 0.003 0.004 1.887 1.905 jit_kernel_multiply 10 15.6 1.165 1.863 1.165 1.863 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.829 1.829 mp_sum_l 7207 12.9 1.134 1.811 1.134 1.811 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.759 1.805 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.744 1.746 cp_fm_cholesky_invert 11 10.9 1.722 1.725 1.722 1.725 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.678 1.717 fft_wrap_pw1pw2_140 487 13.2 0.087 0.089 1.698 1.704 fft3d_ps 1201 14.6 0.566 0.578 1.691 1.694 hybrid_alltoall_any 4725 16.4 0.087 0.147 1.230 1.533 rs_pw_transfer 974 11.9 0.009 0.009 1.479 1.511 make_images_data 4572 15.5 0.042 0.046 1.215 1.508 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.500 1.500 cp_fm_diag_elpa_base 50 14.0 1.353 1.408 1.497 1.498 grid_collocate_task_list 119 9.7 1.469 1.483 1.469 1.483 wfi_extrapolate 11 7.9 0.001 0.001 1.400 1.400 potential_pw2rs 119 12.3 0.014 0.015 1.298 1.302 mp_alltoall_d11v 2130 13.8 1.234 1.264 1.234 1.264 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.240 1.248 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.193 1.218 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.087 1.106 multiply_cannon_sync_h2d 9144 15.5 1.039 1.043 1.039 1.043 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.982 1.025 mp_alltoall_z22v 1201 16.6 0.989 1.011 0.989 1.011 qs_create_task_list 11 7.9 0.000 0.001 0.941 0.954 generate_qs_task_list 11 8.9 0.372 0.392 0.941 0.953 acc_transpose_blocks 9144 15.5 0.038 0.038 0.900 0.905 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.881 0.893 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=41.930000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=705.454545, yerr=17.499941 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a007b1b0cc940e469a6f710d029fae2a5232a07b_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 502.980608E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65372. MP_Allreduce 9840 486. MP_Sync 100 MP_Alltoall 1938 804508. MP_ISendRecv 41800 9096. MP_Wait 58168 MP_comm_split 48 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.040 0.058 84.575 84.576 qs_mol_dyn_low 1 2.0 0.003 0.005 84.189 84.198 qs_forces 11 3.9 0.002 0.003 84.117 84.118 qs_energies 11 4.9 0.006 0.020 81.128 81.137 scf_env_do_scf 11 5.9 0.001 0.003 72.128 72.130 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 66.209 66.211 dbcsr_multiply_generic 2055 12.4 0.108 0.112 52.670 53.067 qs_scf_new_mos 99 7.5 0.000 0.001 48.788 48.906 qs_scf_loop_do_ot 99 8.5 0.001 0.001 48.788 48.906 ot_scf_mini 99 9.5 0.002 0.003 46.321 46.439 multiply_cannon 2055 13.4 0.179 0.187 42.773 43.744 velocity_verlet 10 3.0 0.001 0.001 42.998 42.999 multiply_cannon_loop 2055 14.4 1.560 1.602 41.722 42.719 ot_mini 99 10.5 0.001 0.001 28.189 28.300 qs_ot_get_derivative 99 11.5 0.001 0.001 21.323 21.441 multiply_cannon_multrec 49320 15.4 12.115 12.833 17.149 18.044 rebuild_ks_matrix 110 8.3 0.000 0.001 14.820 14.996 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.015 14.820 14.995 mp_waitall_1 241148 16.1 12.272 13.534 12.272 13.534 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.939 13.106 multiply_cannon_sync_h2d 49320 15.4 9.941 10.519 9.941 10.519 qs_ot_get_p 110 10.4 0.001 0.002 9.582 9.784 multiply_cannon_metrocomm3 49320 15.4 0.079 0.084 7.044 8.615 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.318 7.871 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.187 7.718 apply_single 110 13.6 0.000 0.000 7.187 7.718 sum_up_and_integrate 110 10.3 0.036 0.042 7.310 7.344 integrate_v_rspace 110 11.3 0.002 0.003 7.274 7.317 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.823 6.883 init_scf_run 11 5.9 0.000 0.001 6.864 6.865 scf_env_initial_rho_setup 11 6.9 0.001 0.002 6.864 6.864 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.499 6.638 calculate_rho_elec 110 8.6 0.020 0.025 6.499 6.638 ot_diis_step 99 11.5 0.006 0.006 6.567 6.567 mp_sum_l 6514 12.8 5.348 6.268 5.348 6.268 qs_ot_p2m_diag 48 11.0 0.012 0.018 6.187 6.216 init_scf_loop 11 6.9 0.001 0.005 5.885 5.887 dbcsr_mm_accdrv_process 87628 16.1 2.068 2.172 4.913 5.241 cp_dbcsr_syevd 48 12.0 0.003 0.005 5.188 5.189 cp_fm_diag_elpa 48 13.0 0.000 0.001 4.626 4.629 cp_fm_redistribute_end 48 14.0 4.019 4.592 4.023 4.595 cp_fm_diag_elpa_base 48 14.0 0.564 4.517 0.568 4.543 rs_pw_transfer 902 11.9 0.011 0.013 4.048 4.237 make_m2s 4110 13.4 0.061 0.065 4.119 4.224 make_images 4110 14.4 0.177 0.190 4.022 4.131 wfi_extrapolate 11 7.9 0.001 0.001 4.114 4.114 calculate_dm_sparse 110 9.5 0.000 0.001 3.835 3.939 multiply_cannon_metrocomm1 49320 15.4 0.057 0.061 2.791 3.908 density_rs2pw 110 9.6 0.004 0.006 3.531 3.707 prepare_preconditioner 11 7.9 0.000 0.000 3.675 3.692 make_preconditioner 11 8.9 0.000 0.002 3.675 3.692 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.596 3.600 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.450 3.495 grid_integrate_task_list 110 12.3 3.233 3.401 3.233 3.401 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.292 3.350 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.249 3.316 pw_transfer 1331 11.6 0.054 0.065 3.156 3.246 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 3.069 3.160 potential_pw2rs 110 12.3 0.006 0.007 2.854 2.904 mp_alltoall_d11v 2046 13.8 2.299 2.741 2.299 2.741 fft3d_ps 1111 14.6 0.753 0.843 2.607 2.688 calculate_first_density_matrix 1 7.0 0.001 0.003 2.647 2.655 fft_wrap_pw1pw2_140 451 13.1 0.168 0.188 2.537 2.629 jit_kernel_multiply 13 15.9 2.564 2.602 2.564 2.602 grid_collocate_task_list 110 9.6 2.090 2.247 2.090 2.247 acc_transpose_blocks 49320 15.4 0.221 0.234 2.171 2.243 mp_sum_d 3879 11.9 1.544 2.174 1.544 2.174 mp_waitany 14300 13.8 1.855 2.049 1.855 2.049 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.023 2.040 cp_fm_cholesky_invert 11 10.9 1.985 1.990 1.985 1.990 make_images_data 4110 15.4 0.042 0.046 1.827 1.950 hybrid_alltoall_any 4261 16.3 0.082 0.480 1.598 1.857 mp_alltoall_z22v 1111 16.6 1.531 1.711 1.531 1.711 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=84.576000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=477.090909, yerr=2.644189 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a007b1b0cc940e469a6f710d029fae2a5232a07b_performance_tests/10/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 390.715586E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 5019072 0.0% 0.0% 100.0% average stack size 0.0 0.0 196.1 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 588.795904E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1972800 MPI messages size (bytes): total size 1.077520E+12 min size 0.000000E+00 max size 4.537280E+06 average size 546.188250E+03 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 222984 1826684928 8192 < size <= 32768 520356 13399818240 32768 < size <= 131072 372336 35386294272 131072 < size <= 4194304 787758 788321309808 4194304 < size <= 16777216 54450 238588003280 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65587. MP_Allreduce 9839 562. MP_Sync 100 MP_Alltoall 1717 3328237. MP_ISendRecv 20680 26400. MP_Wait 32692 MP_comm_split 48 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.020 0.045 73.129 73.130 qs_mol_dyn_low 1 2.0 0.003 0.004 72.821 72.831 qs_forces 11 3.9 0.002 0.002 72.664 72.665 qs_energies 11 4.9 0.037 0.052 69.303 69.309 scf_env_do_scf 11 5.9 0.001 0.001 59.751 59.754 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 51.147 51.148 dbcsr_multiply_generic 2055 12.4 0.139 0.178 38.865 39.126 velocity_verlet 10 3.0 0.001 0.001 38.232 38.235 qs_scf_new_mos 99 7.5 0.001 0.001 34.499 34.651 qs_scf_loop_do_ot 99 8.5 0.001 0.002 34.499 34.650 multiply_cannon 2055 13.4 0.221 0.244 31.395 32.940 ot_scf_mini 99 9.5 0.003 0.004 32.792 32.921 multiply_cannon_loop 2055 14.4 0.929 0.953 29.930 31.028 ot_mini 99 10.5 0.001 0.001 19.135 19.286 multiply_cannon_multrec 24660 15.4 7.517 8.939 13.753 15.138 rebuild_ks_matrix 110 8.3 0.000 0.001 14.062 14.167 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.015 14.062 14.166 qs_ot_get_derivative 99 11.5 0.001 0.002 13.227 13.361 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.421 12.509 mp_waitall_1 186928 16.3 8.745 11.919 8.745 11.919 multiply_cannon_metrocomm3 24660 15.4 0.068 0.070 5.477 8.615 init_scf_loop 11 6.9 0.000 0.002 8.561 8.562 multiply_cannon_sync_h2d 24660 15.4 6.965 7.848 6.965 7.848 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.595 7.304 apply_single 110 13.6 0.000 0.001 6.595 7.304 qs_ot_get_p 110 10.4 0.001 0.001 6.884 7.056 sum_up_and_integrate 110 10.3 0.053 0.059 6.900 6.912 integrate_v_rspace 110 11.3 0.002 0.002 6.847 6.862 init_scf_run 11 5.9 0.000 0.001 6.758 6.758 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.757 6.758 prepare_preconditioner 11 7.9 0.000 0.000 6.471 6.487 make_preconditioner 11 8.9 0.000 0.001 6.471 6.487 dbcsr_mm_accdrv_process 52282 16.1 4.566 5.399 6.079 6.381 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.254 6.264 calculate_rho_elec 110 8.6 0.039 0.047 6.254 6.264 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.009 6.171 ot_diis_step 99 11.5 0.010 0.010 5.848 5.848 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.922 5.632 make_m2s 4110 13.4 0.057 0.060 4.711 5.249 make_images 4110 14.4 0.398 0.441 4.600 5.134 qs_ot_p2m_diag 48 11.0 0.028 0.044 4.773 4.798 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.297 4.298 pw_transfer 1331 11.6 0.066 0.071 3.716 3.883 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.609 3.778 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.775 3.777 density_rs2pw 110 9.6 0.004 0.004 3.484 3.662 wfi_extrapolate 11 7.9 0.001 0.001 3.659 3.659 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.653 3.657 cp_fm_redistribute_end 48 14.0 2.713 3.612 2.717 3.613 cp_fm_diag_elpa_base 48 14.0 0.858 3.479 0.891 3.566 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.487 3.561 rs_pw_transfer 902 11.9 0.012 0.014 3.347 3.556 grid_integrate_task_list 110 12.3 3.148 3.329 3.148 3.329 cp_fm_cholesky_invert 11 10.9 3.137 3.145 3.137 3.145 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.077 3.126 fft3d_ps 1111 14.6 1.074 1.271 2.970 3.122 make_images_data 4110 15.4 0.046 0.050 2.546 3.075 calculate_dm_sparse 110 9.5 0.001 0.001 3.016 3.063 fft_wrap_pw1pw2_140 451 13.1 0.197 0.215 2.862 3.024 calculate_first_density_matrix 1 7.0 0.000 0.001 2.999 3.002 mp_sum_l 6514 12.8 2.131 2.845 2.131 2.845 hybrid_alltoall_any 4261 16.3 0.102 0.444 2.242 2.840 potential_pw2rs 110 12.3 0.008 0.008 2.719 2.745 mp_alltoall_d11v 2046 13.8 2.065 2.480 2.065 2.480 grid_collocate_task_list 110 9.6 2.121 2.294 2.121 2.294 qs_energies_init_hamiltonians 11 5.9 0.038 0.104 2.045 2.104 qs_ot_get_orbitals 99 10.5 0.001 0.001 2.080 2.099 jit_kernel_multiply 10 16.3 1.161 1.985 1.161 1.985 dbcsr_complete_redistribute 325 12.2 0.417 0.569 1.634 1.906 mp_allgather_i34 2055 14.4 0.795 1.884 0.795 1.884 multiply_cannon_metrocomm4 22605 15.4 0.073 0.078 0.802 1.877 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.808 1.824 mp_alltoall_z22v 1111 16.6 1.602 1.698 1.602 1.698 mp_irecv_dv 57340 16.2 0.679 1.666 0.679 1.666 cp_fm_cholesky_decompose 22 10.9 1.641 1.650 1.641 1.650 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.552 1.565 acc_transpose_blocks 24660 15.4 0.110 0.114 1.537 1.563 mp_sum_d 3879 11.9 1.119 1.512 1.119 1.512 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.369 1.474 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=73.130000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=555.636364, yerr=7.713892 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a007b1b0cc940e469a6f710d029fae2a5232a07b_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 656.609280E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65578. MP_Allreduce 9838 559. MP_Sync 100 MP_Alltoall 1496 4511006. MP_ISendRecv 13640 27424. MP_Wait 32318 MP_comm_split 48 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.038 61.974 61.976 qs_mol_dyn_low 1 2.0 0.003 0.005 61.623 61.632 qs_forces 11 3.9 0.004 0.021 61.509 61.518 qs_energies 11 4.9 0.005 0.022 58.134 58.151 scf_env_do_scf 11 5.9 0.003 0.017 49.720 49.720 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 40.531 40.539 velocity_verlet 10 3.0 0.001 0.002 33.432 33.435 dbcsr_multiply_generic 2055 12.4 0.110 0.113 29.123 29.448 qs_scf_new_mos 99 7.5 0.001 0.001 25.721 25.830 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.720 25.830 ot_scf_mini 99 9.5 0.003 0.004 24.453 24.577 multiply_cannon 2055 13.4 0.210 0.218 22.399 23.549 multiply_cannon_loop 2055 14.4 0.618 0.633 21.140 22.441 ot_mini 99 10.5 0.002 0.010 14.161 14.288 rebuild_ks_matrix 110 8.3 0.000 0.000 12.577 12.730 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.020 12.576 12.729 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.008 11.143 multiply_cannon_multrec 16440 15.4 3.800 4.928 9.604 10.765 mp_waitall_1 146766 16.3 7.739 10.689 7.739 10.689 qs_ot_get_derivative 99 11.5 0.009 0.137 9.596 9.721 init_scf_loop 11 6.9 0.000 0.002 9.146 9.152 multiply_cannon_metrocomm3 16440 15.4 0.043 0.044 4.570 7.412 prepare_preconditioner 11 7.9 0.000 0.000 7.246 7.263 make_preconditioner 11 8.9 0.000 0.001 7.246 7.263 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.532 6.888 sum_up_and_integrate 110 10.3 0.061 0.062 6.674 6.706 integrate_v_rspace 110 11.3 0.002 0.003 6.613 6.646 qs_rho_update_rho_low 110 7.6 0.001 0.002 6.017 6.050 calculate_rho_elec 110 8.6 0.058 0.059 6.016 6.049 init_scf_run 11 5.9 0.000 0.001 5.934 5.934 scf_env_initial_rho_setup 11 6.9 0.021 0.146 5.933 5.934 dbcsr_mm_accdrv_process 34862 16.1 4.647 5.280 5.660 5.855 apply_preconditioner_dbcsr 110 12.6 0.002 0.021 5.168 5.690 apply_single 110 13.6 0.000 0.000 5.166 5.690 qs_ot_get_p 110 10.4 0.001 0.001 5.440 5.585 make_m2s 4110 13.4 0.051 0.053 4.356 4.757 make_images 4110 14.4 0.391 0.509 4.240 4.641 ot_diis_step 99 11.5 0.011 0.013 4.526 4.527 multiply_cannon_sync_h2d 16440 15.4 3.656 4.314 3.656 4.314 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.265 4.027 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.693 3.699 pw_transfer 1331 11.6 0.064 0.071 3.420 3.448 density_rs2pw 110 9.6 0.004 0.004 3.076 3.398 grid_integrate_task_list 110 12.3 3.192 3.388 3.192 3.388 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.314 3.345 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.338 3.339 rs_pw_transfer 902 11.9 0.010 0.012 2.996 3.320 wfi_extrapolate 11 7.9 0.003 0.019 3.085 3.086 make_images_data 4110 15.4 0.043 0.047 2.533 3.039 cp_fm_cholesky_invert 11 10.9 2.999 3.006 2.999 3.006 hybrid_alltoall_any 4261 16.3 0.105 0.374 2.230 2.850 fft_wrap_pw1pw2_140 451 13.1 0.209 0.211 2.813 2.845 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.810 2.812 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.774 2.778 cp_fm_redistribute_end 48 14.0 1.722 2.746 1.724 2.747 cp_fm_diag_elpa_base 48 14.0 0.959 2.610 1.016 2.714 fft3d_ps 1111 14.6 1.064 1.077 2.630 2.667 calculate_dm_sparse 110 9.5 0.001 0.001 2.589 2.622 potential_pw2rs 110 12.3 0.010 0.010 2.478 2.510 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.442 2.507 multiply_cannon_metrocomm4 14385 15.4 0.043 0.047 0.895 2.502 mp_sum_l 6514 12.8 1.761 2.492 1.761 2.492 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.432 2.483 calculate_first_density_matrix 1 7.0 0.001 0.003 2.435 2.439 grid_collocate_task_list 110 9.6 2.220 2.430 2.220 2.430 mp_alltoall_d11v 2046 13.8 1.997 2.423 1.997 2.423 mp_irecv_dv 48980 15.7 0.825 2.376 0.825 2.376 dbcsr_complete_redistribute 325 12.2 0.320 0.345 1.560 2.015 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.995 1.998 cp_fm_upper_to_full 70 13.6 1.393 1.837 1.393 1.837 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.768 1.786 cp_fm_cholesky_decompose 22 10.9 1.733 1.748 1.733 1.748 mp_allgather_i34 2055 14.4 0.589 1.685 0.589 1.685 jit_kernel_multiply 8 16.7 0.626 1.662 0.626 1.662 mp_waitany 17072 13.8 1.318 1.586 1.318 1.586 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.509 1.523 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.050 1.506 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.371 1.474 mp_alltoall_z22v 1111 16.6 1.353 1.400 1.353 1.400 rs_pw_transfer_RS2PW_140 121 11.5 0.175 0.178 0.996 1.307 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.293 1.302 rs_gather_matrices 110 12.3 0.137 0.149 0.890 1.295 acc_transpose_blocks 16440 15.4 0.072 0.075 1.237 1.274 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=61.976000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=620.818182, yerr=8.736737 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a007b1b0cc940e469a6f710d029fae2a5232a07b_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 731.529216E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65576. MP_Allreduce 9838 600. MP_Sync 100 MP_Alltoall 1496 5863162. MP_ISendRecv 10120 43184. MP_Wait 25102 MP_comm_split 48 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.022 0.044 66.731 66.732 qs_mol_dyn_low 1 2.0 0.003 0.004 66.387 66.396 qs_forces 11 3.9 0.002 0.002 66.314 66.315 qs_energies 11 4.9 0.009 0.065 62.919 62.924 scf_env_do_scf 11 5.9 0.001 0.001 54.445 54.456 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 42.033 42.034 velocity_verlet 10 3.0 0.001 0.001 37.267 37.269 dbcsr_multiply_generic 2055 12.4 0.121 0.124 30.116 30.349 qs_scf_new_mos 99 7.5 0.001 0.001 27.397 27.517 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.396 27.517 ot_scf_mini 99 9.5 0.003 0.004 25.756 25.858 multiply_cannon 2055 13.4 0.243 0.260 22.881 24.072 multiply_cannon_loop 2055 14.4 0.893 0.912 21.495 22.096 ot_mini 99 10.5 0.001 0.001 14.792 14.914 multiply_cannon_multrec 24660 15.4 4.276 6.996 12.903 14.263 init_scf_loop 11 6.9 0.005 0.040 12.368 12.377 rebuild_ks_matrix 110 8.3 0.000 0.000 12.236 12.351 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.029 12.235 12.350 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.867 10.969 qs_ot_get_derivative 99 11.5 0.001 0.001 10.604 10.712 prepare_preconditioner 11 7.9 0.000 0.000 10.316 10.332 make_preconditioner 11 8.9 0.007 0.029 10.316 10.331 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.520 9.985 dbcsr_mm_accdrv_process 52304 16.0 6.870 8.225 8.480 9.386 mp_waitall_1 126806 16.4 4.876 7.061 4.876 7.061 sum_up_and_integrate 110 10.3 0.068 0.071 6.466 6.482 integrate_v_rspace 110 11.3 0.002 0.003 6.399 6.414 qs_ot_get_p 110 10.4 0.001 0.001 5.979 6.107 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.961 5.970 calculate_rho_elec 110 8.6 0.077 0.081 5.960 5.970 make_m2s 4110 13.4 0.060 0.062 5.549 5.781 init_scf_run 11 5.9 0.000 0.001 5.714 5.714 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.713 5.714 make_images 4110 14.4 0.572 0.693 5.408 5.636 cp_fm_upper_to_full 70 13.8 3.263 4.644 3.263 4.644 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.057 4.205 apply_single 110 13.6 0.000 0.000 4.056 4.205 ot_diis_step 99 11.5 0.011 0.011 4.150 4.150 qs_ot_p2m_diag 48 11.0 0.055 0.064 4.069 4.084 dbcsr_complete_redistribute 325 12.2 0.425 0.478 2.759 3.881 multiply_cannon_metrocomm3 24660 15.4 0.035 0.036 1.571 3.638 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.574 3.575 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.471 3.543 grid_integrate_task_list 110 12.3 3.271 3.416 3.271 3.416 pw_transfer 1331 11.6 0.064 0.072 3.385 3.412 multiply_cannon_sync_h2d 24660 15.4 3.161 3.374 3.161 3.374 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.279 3.310 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.189 3.303 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.139 3.200 density_rs2pw 110 9.6 0.004 0.004 3.031 3.171 make_images_data 4110 15.4 0.046 0.050 2.818 3.078 hybrid_alltoall_any 4261 16.3 0.119 0.457 2.456 3.050 wfi_extrapolate 11 7.9 0.001 0.001 3.031 3.031 cp_fm_cholesky_invert 11 10.9 3.004 3.011 3.004 3.011 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.976 2.979 calculate_dm_sparse 110 9.5 0.001 0.001 2.933 2.960 cp_fm_redistribute_end 48 14.0 1.478 2.940 1.480 2.941 cp_fm_diag_elpa_base 48 14.0 1.377 2.807 1.456 2.915 mp_alltoall_i22 605 13.7 1.717 2.911 1.717 2.911 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.781 2.879 rs_pw_transfer 902 11.9 0.010 0.011 2.680 2.856 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.820 2.822 fft_wrap_pw1pw2_140 451 13.1 0.201 0.212 2.758 2.790 fft3d_ps 1111 14.6 1.064 1.098 2.583 2.601 grid_collocate_task_list 110 9.6 2.257 2.494 2.257 2.494 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.418 2.454 calculate_first_density_matrix 1 7.0 0.000 0.002 2.404 2.407 jit_kernel_multiply 12 15.7 1.281 2.291 1.281 2.291 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.247 2.249 mp_alltoall_d11v 2046 13.8 2.014 2.229 2.014 2.229 potential_pw2rs 110 12.3 0.012 0.013 2.208 2.224 cp_fm_cholesky_decompose 22 10.9 1.721 1.774 1.721 1.774 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.724 1.754 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.597 1.696 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.627 1.642 mp_sum_l 6514 12.8 0.999 1.624 0.999 1.624 mp_allgather_i34 2055 14.4 0.493 1.591 0.493 1.591 acc_transpose_blocks 24660 15.4 0.103 0.105 1.523 1.556 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.532 1.541 multiply_cannon_metrocomm4 20550 15.4 0.056 0.059 0.854 1.511 mp_waitany 13376 13.8 1.236 1.476 1.236 1.476 mp_irecv_dv 62702 16.1 0.756 1.436 0.756 1.436 mp_alltoall_z22v 1111 16.6 1.302 1.361 1.302 1.361 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=66.732000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=693.545455, yerr=9.228056 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a007b1b0cc940e469a6f710d029fae2a5232a07b_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 832.192512E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65574. MP_Allreduce 9838 640. MP_Sync 100 MP_Alltoall 1496 8504061. MP_ISendRecv 6600 54848. MP_Wait 17226 MP_comm_split 48 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.034 0.094 57.515 57.515 qs_mol_dyn_low 1 2.0 0.013 0.175 56.662 56.672 qs_forces 11 3.9 0.053 0.104 56.554 56.596 qs_energies 11 4.9 0.001 0.001 52.871 52.969 scf_env_do_scf 11 5.9 0.000 0.001 44.027 44.029 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 35.705 35.705 velocity_verlet 10 3.0 0.001 0.001 31.616 31.619 dbcsr_multiply_generic 2055 12.4 0.106 0.108 23.631 23.803 qs_scf_new_mos 99 7.5 0.001 0.001 21.324 21.391 qs_scf_loop_do_ot 99 8.5 0.001 0.001 21.323 21.391 ot_scf_mini 99 9.5 0.002 0.002 20.065 20.121 multiply_cannon 2055 13.4 0.245 0.259 17.903 19.057 multiply_cannon_loop 2055 14.4 0.323 0.335 16.520 16.858 rebuild_ks_matrix 110 8.3 0.000 0.000 11.751 11.789 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.750 11.788 ot_mini 99 10.5 0.001 0.001 10.823 10.870 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.476 10.508 multiply_cannon_multrec 8220 15.4 3.263 4.557 7.739 8.957 mp_waitall_1 106626 16.5 6.599 8.562 6.599 8.562 init_scf_loop 11 6.9 0.000 0.000 8.269 8.270 qs_ot_get_derivative 99 11.5 0.001 0.001 6.926 6.981 prepare_preconditioner 11 7.9 0.000 0.000 6.494 6.502 make_preconditioner 11 8.9 0.000 0.000 6.494 6.502 sum_up_and_integrate 110 10.3 0.080 0.081 6.326 6.339 integrate_v_rspace 110 11.3 0.003 0.003 6.246 6.258 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.067 6.148 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.741 5.754 calculate_rho_elec 110 8.6 0.115 0.115 5.741 5.754 init_scf_run 11 5.9 0.000 0.001 5.419 5.419 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.419 5.419 dbcsr_mm_accdrv_process 17442 15.9 2.894 4.164 4.346 5.347 qs_ot_get_p 110 10.4 0.001 0.001 5.258 5.325 multiply_cannon_metrocomm3 8220 15.4 0.017 0.018 3.161 4.778 make_m2s 4110 13.4 0.039 0.040 4.397 4.680 make_images 4110 14.4 0.633 0.692 4.266 4.551 ot_diis_step 99 11.5 0.012 0.012 3.872 3.873 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.776 3.813 apply_single 110 13.6 0.000 0.000 3.776 3.812 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.795 3.800 grid_integrate_task_list 110 12.3 3.362 3.473 3.362 3.473 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.460 3.461 pw_transfer 1331 11.6 0.065 0.069 3.379 3.388 cp_fm_cholesky_invert 11 10.9 3.285 3.289 3.285 3.289 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.273 3.282 qs_energies_init_hamiltonians 11 5.9 0.009 0.015 2.945 3.038 multiply_cannon_sync_h2d 8220 15.4 2.887 3.015 2.887 3.015 make_images_data 4110 15.4 0.038 0.043 2.533 3.004 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.899 2.901 density_rs2pw 110 9.6 0.004 0.004 2.743 2.867 cp_fm_redistribute_end 48 14.0 0.734 2.862 0.740 2.863 hybrid_alltoall_any 4261 16.3 0.200 0.862 2.434 2.832 cp_fm_diag_elpa_base 48 14.0 1.929 2.655 2.111 2.820 wfi_extrapolate 11 7.9 0.001 0.001 2.803 2.803 fft_wrap_pw1pw2_140 451 13.1 0.211 0.215 2.776 2.788 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.663 2.664 rs_pw_transfer 902 11.9 0.010 0.010 2.449 2.648 calculate_dm_sparse 110 9.5 0.001 0.001 2.574 2.615 fft3d_ps 1111 14.6 1.115 1.143 2.532 2.542 calculate_first_density_matrix 1 7.0 0.000 0.000 2.529 2.530 grid_collocate_task_list 110 9.6 2.332 2.445 2.332 2.445 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.291 2.305 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.034 2.064 potential_pw2rs 110 12.3 0.015 0.015 2.018 2.022 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.766 1.990 jit_kernel_multiply 10 15.7 1.144 1.975 1.144 1.975 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.928 1.957 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.752 1.927 mp_alltoall_d11v 2046 13.8 1.787 1.875 1.787 1.875 cp_fm_cholesky_decompose 22 10.9 1.771 1.782 1.771 1.782 mp_allgather_i34 2055 14.4 0.572 1.735 0.572 1.735 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.607 1.613 dbcsr_complete_redistribute 325 12.2 0.554 0.575 1.483 1.575 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.523 1.532 mp_waitany 9240 13.8 1.256 1.486 1.256 1.486 multiply_cannon_metrocomm4 6165 15.4 0.017 0.018 0.478 1.396 qs_create_task_list 11 7.9 0.010 0.036 1.258 1.356 mp_irecv_dv 24056 15.7 0.454 1.354 0.454 1.354 generate_qs_task_list 11 8.9 0.378 0.446 1.248 1.339 multiply_cannon_metrocomm1 8220 15.4 0.020 0.021 0.878 1.315 mp_alltoall_z22v 1111 16.6 1.183 1.207 1.183 1.207 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.157 1.181 rs_pw_transfer_RS2PW_140 121 11.5 0.172 0.199 0.969 1.168 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=57.515000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=783.000000, yerr=10.304456 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a007b1b0cc940e469a6f710d029fae2a5232a07b_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 1.333236E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67104. MP_Allreduce 9672 819. MP_Sync 52 MP_Alltoall 1474 16505187. MP_ISendRecv 4620 360267. MP_Wait 7524 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.093 0.180 90.129 90.129 qs_mol_dyn_low 1 2.0 0.003 0.005 89.030 89.038 qs_forces 11 3.9 0.002 0.002 88.950 88.962 qs_energies 11 4.9 0.001 0.001 84.884 84.898 scf_env_do_scf 11 5.9 0.001 0.001 72.894 72.895 velocity_verlet 10 3.0 0.001 0.001 56.083 56.090 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 43.719 43.720 init_scf_loop 11 6.9 0.000 0.000 29.099 29.100 dbcsr_multiply_generic 2055 12.4 0.120 0.122 28.978 29.074 prepare_preconditioner 11 7.9 0.000 0.000 27.133 27.144 make_preconditioner 11 8.9 0.000 0.000 27.133 27.144 qs_scf_new_mos 99 7.5 0.001 0.001 26.936 27.027 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.935 27.027 make_full_inverse_cholesky 11 9.9 0.000 0.000 21.289 26.600 ot_scf_mini 99 9.5 0.002 0.002 25.154 25.218 multiply_cannon 2055 13.4 0.354 0.395 21.702 22.526 multiply_cannon_loop 2055 14.4 0.342 0.350 19.854 20.187 cp_fm_upper_to_full 70 14.2 12.833 18.385 12.833 18.385 ot_mini 99 10.5 0.001 0.001 13.953 14.016 rebuild_ks_matrix 110 8.3 0.001 0.001 13.382 13.459 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 13.382 13.459 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.183 12.251 dbcsr_complete_redistribute 325 12.2 1.063 1.091 7.512 10.871 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.448 9.801 multiply_cannon_multrec 8220 15.4 4.376 4.620 9.541 9.640 mp_waitall_1 87304 16.6 8.226 9.346 8.226 9.346 qs_ot_get_derivative 99 11.5 0.001 0.001 9.229 9.294 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.829 9.156 mp_alltoall_i22 605 13.7 5.458 8.787 5.458 8.787 init_scf_run 11 5.9 0.000 0.001 7.540 7.540 scf_env_initial_rho_setup 11 6.9 0.008 0.015 7.540 7.540 sum_up_and_integrate 110 10.3 0.151 0.152 6.786 6.799 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.638 6.673 calculate_rho_elec 110 8.6 0.226 0.227 6.638 6.673 integrate_v_rspace 110 11.3 0.003 0.003 6.635 6.649 qs_ot_get_p 110 10.4 0.001 0.001 5.962 6.035 make_m2s 4110 13.4 0.043 0.044 5.449 5.990 make_images 4110 14.4 0.873 0.924 5.260 5.799 cp_fm_cholesky_invert 11 10.9 5.761 5.766 5.761 5.766 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.801 5.302 apply_single 110 13.6 0.000 0.000 4.801 5.302 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 4.962 5.300 dbcsr_mm_accdrv_process 11614 15.7 3.300 3.642 5.023 5.254 ot_diis_step 99 11.5 0.015 0.016 4.701 4.702 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.328 4.336 calculate_first_density_matrix 1 7.0 0.000 0.000 4.145 4.152 multiply_cannon_sync_h2d 8220 15.4 3.947 3.954 3.947 3.954 pw_transfer 1331 11.6 0.073 0.073 3.912 3.918 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.873 3.874 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.829 3.840 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 3.796 3.803 hybrid_alltoall_any 4261 16.3 0.256 0.551 2.975 3.739 grid_integrate_task_list 110 12.3 3.669 3.733 3.669 3.733 make_images_data 4110 15.4 0.041 0.044 3.017 3.715 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.945 3.409 fft_wrap_pw1pw2_140 451 13.1 0.214 0.216 3.282 3.291 wfi_extrapolate 11 7.9 0.001 0.001 3.284 3.284 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.228 3.228 cp_fm_diag_elpa_base 48 14.0 2.687 2.889 3.226 3.227 calculate_dm_sparse 110 9.5 0.001 0.001 3.144 3.166 density_rs2pw 110 9.6 0.004 0.004 3.096 3.119 fft3d_ps 1111 14.6 1.268 1.282 3.007 3.014 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.926 2.931 make_basis_sm 11 9.8 0.000 0.000 2.683 2.686 grid_collocate_task_list 110 9.6 2.636 2.666 2.636 2.666 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.477 2.509 parallel_gemm_fm 81 9.0 0.000 0.000 2.403 2.407 parallel_gemm_fm_cosma 81 10.0 2.403 2.407 2.403 2.407 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.348 2.376 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.278 2.321 rs_pw_transfer 902 11.9 0.010 0.010 2.246 2.308 mp_alltoall_d11v 2046 13.8 2.203 2.242 2.203 2.242 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.092 2.183 cp_fm_cholesky_decompose 22 10.9 2.165 2.180 2.165 2.180 potential_pw2rs 110 12.3 0.021 0.022 2.151 2.155 qs_create_task_list 11 7.9 0.010 0.021 1.918 1.960 generate_qs_task_list 11 8.9 0.738 0.789 1.908 1.940 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.815 1.851 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.803 1.811 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=90.129000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1194.363636, yerr=59.550243 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a007b1b0cc940e469a6f710d029fae2a5232a07b_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 627.064832E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175955383376 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4085 56760. MP_Allreduce 11253 785. MP_Sync 170 MP_Alltoall 2226 3068946. MP_ISendRecv 48640 18752. MP_Wait 66796 MP_comm_split 83 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 0.023 0.092 206.262 206.264 qs_mol_dyn_low 1 2.0 0.004 0.009 205.747 205.762 qs_forces 11 3.9 0.003 0.003 205.642 205.643 qs_energies 11 4.9 0.001 0.002 200.123 200.142 scf_env_do_scf 11 5.9 0.001 0.001 182.444 182.448 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 161.019 161.021 dbcsr_multiply_generic 2507 12.6 0.179 0.187 124.622 125.260 velocity_verlet 10 3.0 0.001 0.001 123.725 123.726 qs_scf_new_mos 117 7.6 0.001 0.001 121.941 122.243 qs_scf_loop_do_ot 117 8.6 0.001 0.001 121.941 122.242 ot_scf_mini 117 9.6 0.003 0.003 115.313 115.549 multiply_cannon 2507 13.6 0.235 0.244 100.870 103.566 multiply_cannon_loop 2507 14.6 2.149 2.225 98.562 101.044 ot_mini 117 10.6 0.001 0.001 66.286 66.552 multiply_cannon_multrec 60168 15.6 33.027 35.316 41.634 44.055 qs_ot_get_derivative 117 11.6 0.001 0.001 41.480 41.720 rebuild_ks_matrix 128 8.3 0.001 0.001 33.523 33.818 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 33.523 33.817 mp_waitall_1 291448 16.2 29.023 31.622 29.023 31.622 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.175 30.444 multiply_cannon_sync_h2d 60168 15.6 27.193 28.683 27.193 28.683 qs_ot_get_p 128 10.4 0.001 0.001 27.171 27.475 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.210 25.005 apply_single 128 13.6 0.001 0.001 24.210 25.005 ot_diis_step 117 11.6 0.008 0.008 24.448 24.449 init_scf_loop 11 6.9 0.000 0.000 21.333 21.335 qs_ot_p2m_diag 83 11.4 0.077 0.096 20.287 20.379 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.277 19.457 cp_dbcsr_syevd 83 12.4 0.004 0.005 17.729 17.730 multiply_cannon_metrocomm3 60168 15.6 0.115 0.122 15.512 17.488 prepare_preconditioner 11 7.9 0.000 0.000 16.594 16.636 make_preconditioner 11 8.9 0.000 0.000 16.594 16.636 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.814 15.995 cp_fm_diag_elpa 83 13.4 0.000 0.001 14.756 14.764 cp_fm_redistribute_end 83 14.4 11.620 14.672 11.634 14.677 cp_fm_diag_elpa_base 83 14.4 2.999 14.432 3.030 14.551 make_m2s 5014 13.6 0.103 0.111 14.073 14.409 make_images 5014 14.6 0.399 0.417 13.891 14.240 sum_up_and_integrate 128 10.3 0.089 0.107 14.190 14.210 integrate_v_rspace 128 11.3 0.003 0.004 14.101 14.122 init_scf_run 11 5.9 0.000 0.001 13.512 13.512 scf_env_initial_rho_setup 11 6.9 0.000 0.001 13.512 13.512 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.115 13.208 calculate_rho_elec 128 8.7 0.044 0.063 13.115 13.207 mp_sum_l 7870 13.0 8.472 9.612 8.472 9.612 cp_fm_cholesky_invert 11 10.9 9.537 9.545 9.537 9.545 wfi_extrapolate 11 7.9 0.001 0.001 9.201 9.201 multiply_cannon_metrocomm1 60168 15.6 0.083 0.088 6.109 8.701 dbcsr_mm_accdrv_process 124484 16.2 3.335 3.491 8.171 8.669 calculate_dm_sparse 128 9.5 0.001 0.001 8.437 8.580 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.950 8.086 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.864 7.955 make_images_data 5014 15.6 0.067 0.073 6.917 7.794 pw_transfer 1547 11.6 0.074 0.106 7.208 7.532 grid_integrate_task_list 128 12.3 7.121 7.490 7.121 7.490 density_rs2pw 128 9.7 0.006 0.007 6.914 7.340 fft_wrap_pw1pw2 1291 12.7 0.010 0.013 7.004 7.306 hybrid_alltoall_any 5200 16.5 0.292 2.252 6.048 7.173 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.004 6.692 6.702 rs_pw_transfer 1046 11.9 0.016 0.018 5.943 6.428 fft3d_ps 1291 14.7 2.092 2.571 5.799 6.044 fft_wrap_pw1pw2_140 523 13.2 0.443 0.510 5.771 5.966 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.818 5.935 mp_alltoall_d11v 2415 14.1 4.289 5.610 4.289 5.610 grid_collocate_task_list 128 9.7 4.726 5.064 4.726 5.064 cp_fm_cholesky_decompose 22 10.9 4.853 4.866 4.853 4.866 potential_pw2rs 128 12.3 0.009 0.010 4.667 4.711 mp_sum_d 4460 12.2 3.694 4.493 3.694 4.493 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=206.264000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=592.454545, yerr=6.035706 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a007b1b0cc940e469a6f710d029fae2a5232a07b_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410023282688 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444707676160 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796579E+12 0.0% 0.0% 100.0% flops max/rank 2.166472E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500928 0.0% 0.0% 100.0% number of processed stacks 5925696 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.6 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 827.133952E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2385600 MPI messages size (bytes): total size 4.069300E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.705776E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70188 2295595008 32768 < size <= 131072 716032 54973693952 131072 < size <= 4194304 1363760 1386318135296 4194304 < size <= 16777216 153648 1453843137296 16777216 < size 67056 1171888537600 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4079 57283. MP_Allreduce 11237 947. MP_Sync 168 MP_Alltoall 1955 5777448. MP_ISendRecv 23876 47072. MP_Wait 37656 MP_comm_split 82 MP_ISend 11660 212488. MP_IRecv 11660 212488. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.043 0.092 194.554 194.554 qs_mol_dyn_low 1 2.0 0.003 0.007 194.017 194.030 qs_forces 11 3.9 0.005 0.007 193.929 193.932 qs_energies 11 4.9 0.005 0.016 187.163 187.176 scf_env_do_scf 11 5.9 0.002 0.009 169.425 169.434 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 134.461 134.464 velocity_verlet 10 3.0 0.001 0.001 120.364 120.365 dbcsr_multiply_generic 2485 12.5 0.188 0.196 97.577 98.854 qs_scf_new_mos 116 7.6 0.001 0.001 95.612 96.102 qs_scf_loop_do_ot 116 8.6 0.001 0.002 95.612 96.101 ot_scf_mini 116 9.6 0.004 0.005 90.824 91.382 multiply_cannon 2485 13.5 0.473 0.525 77.077 81.448 multiply_cannon_loop 2485 14.5 1.243 1.288 73.603 76.338 ot_mini 116 10.6 0.001 0.001 50.030 50.540 mp_waitall_1 224796 16.4 25.807 39.341 25.807 39.341 multiply_cannon_multrec 29820 15.5 21.653 26.212 31.247 36.283 init_scf_loop 11 6.9 0.010 0.075 34.865 34.866 rebuild_ks_matrix 127 8.3 0.001 0.001 33.736 34.287 qs_ks_build_kohn_sham_matrix 127 9.3 0.031 0.131 33.735 34.286 qs_ks_update_qs_env 127 7.6 0.001 0.001 30.463 30.969 prepare_preconditioner 11 7.9 0.000 0.000 29.585 29.707 make_preconditioner 11 8.9 0.003 0.023 29.585 29.707 multiply_cannon_metrocomm3 29820 15.5 0.091 0.097 15.999 29.152 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.258 28.831 qs_ot_get_derivative 116 11.6 0.001 0.002 28.132 28.699 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 21.912 23.199 apply_single 127 13.6 0.001 0.001 21.912 23.198 qs_ot_get_p 127 10.4 0.001 0.001 22.547 23.136 ot_diis_step 116 11.6 0.014 0.015 21.716 21.717 multiply_cannon_sync_h2d 29820 15.5 18.925 21.597 18.925 21.597 qs_ot_p2m_diag 82 11.4 0.185 0.213 17.666 17.709 cp_fm_cholesky_invert 11 10.9 17.274 17.291 17.274 17.291 cp_dbcsr_syevd 82 12.4 0.005 0.005 16.487 16.489 make_m2s 4970 13.5 0.090 0.097 14.602 16.288 make_images 4970 14.5 1.140 1.328 14.392 16.078 sum_up_and_integrate 127 10.3 0.115 0.132 14.679 14.735 integrate_v_rspace 127 11.3 0.003 0.003 14.564 14.623 cp_fm_diag_elpa 82 13.4 0.000 0.001 13.287 13.298 cp_fm_redistribute_end 82 14.4 7.723 13.213 7.735 13.215 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.159 13.191 calculate_rho_elec 127 8.7 0.087 0.104 13.159 13.190 cp_fm_diag_elpa_base 82 14.4 5.245 12.751 5.458 13.106 init_scf_run 11 5.9 0.000 0.001 12.884 12.885 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.883 12.885 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 11.094 11.509 make_images_data 4970 15.5 0.064 0.071 8.963 10.911 multiply_cannon_metrocomm4 27335 15.5 0.095 0.112 3.741 10.626 mp_irecv_dv 68888 16.3 3.548 10.227 3.548 10.227 hybrid_alltoall_any 5155 16.4 0.340 1.507 7.550 10.176 dbcsr_mm_accdrv_process 61726 16.2 4.541 5.337 9.067 9.629 wfi_extrapolate 11 7.9 0.001 0.001 8.532 8.532 pw_transfer 1535 11.6 0.084 0.096 8.100 8.185 fft_wrap_pw1pw2 1281 12.7 0.010 0.010 7.878 7.959 grid_integrate_task_list 127 12.3 7.120 7.522 7.120 7.522 density_rs2pw 127 9.7 0.006 0.006 6.946 7.395 cp_fm_cholesky_decompose 22 10.9 7.244 7.330 7.244 7.330 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.293 7.026 fft_wrap_pw1pw2_140 519 13.2 0.464 0.513 6.803 6.875 calculate_dm_sparse 127 9.5 0.001 0.001 6.481 6.639 rs_pw_transfer 1038 11.9 0.014 0.016 5.857 6.367 fft3d_ps 1281 14.7 2.757 2.890 6.267 6.327 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.269 6.279 mp_sum_l 7804 13.0 4.138 5.939 4.138 5.939 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.336 5.503 qs_ot_get_orbitals 116 10.6 0.001 0.001 5.374 5.454 mp_allgather_i34 2485 14.5 2.064 5.381 2.064 5.381 grid_collocate_task_list 127 9.7 4.876 5.244 4.876 5.244 potential_pw2rs 127 12.3 0.014 0.015 5.159 5.207 mp_alltoall_d11v 2401 14.1 4.295 4.821 4.295 4.821 mp_sum_d 4439 12.1 3.299 4.707 3.299 4.707 calculate_first_density_matrix 1 7.0 0.000 0.002 4.176 4.181 dbcsr_complete_redistribute 393 12.7 0.773 0.864 3.239 4.109 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=194.554000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=787.000000, yerr=5.116817 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a007b1b0cc940e469a6f710d029fae2a5232a07b_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456039424 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613072052224 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176077312 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176077312 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.233020E+12 0.0% 0.0% 100.0% flops max/rank 2.951022E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383904 0.0% 0.0% 100.0% number of processed stacks 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 940.978176E+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 937891119088 16777216 < size 28896 757491302400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4127 56629. MP_Allreduce 11356 981. MP_Sync 172 MP_Alltoall 1724 9394234. MP_ISendRecv 15996 75008. MP_Wait 29984 MP_comm_split 84 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 0.030 0.058 184.485 184.487 qs_mol_dyn_low 1 2.0 0.003 0.004 183.909 183.924 qs_forces 11 3.9 0.008 0.038 183.799 183.806 qs_energies 11 4.9 0.009 0.065 177.226 177.292 scf_env_do_scf 11 5.9 0.001 0.001 159.996 160.001 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 123.051 123.052 velocity_verlet 10 3.0 0.001 0.001 116.984 116.986 qs_scf_new_mos 118 7.6 0.001 0.001 85.176 85.551 qs_scf_loop_do_ot 118 8.6 0.001 0.001 85.175 85.550 dbcsr_multiply_generic 2529 12.6 0.185 0.189 83.356 84.434 ot_scf_mini 118 9.6 0.004 0.004 80.950 81.341 multiply_cannon 2529 13.6 0.499 0.520 62.762 67.420 multiply_cannon_loop 2529 14.6 0.867 0.906 59.664 62.159 ot_mini 118 10.6 0.001 0.001 43.982 44.377 init_scf_loop 11 6.9 0.019 0.152 36.838 36.841 mp_waitall_1 180004 16.5 27.086 36.620 27.086 36.620 prepare_preconditioner 11 7.9 0.000 0.000 32.397 32.443 make_preconditioner 11 8.9 0.000 0.002 32.397 32.443 rebuild_ks_matrix 129 8.3 0.001 0.001 31.139 31.658 qs_ks_build_kohn_sham_matrix 129 9.3 0.016 0.020 31.139 31.657 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.982 31.376 qs_ks_update_qs_env 129 7.6 0.001 0.001 28.260 28.746 multiply_cannon_metrocomm3 20232 15.6 0.060 0.065 16.084 25.884 multiply_cannon_multrec 20232 15.6 13.401 16.523 22.276 25.319 qs_ot_get_derivative 118 11.6 0.001 0.002 23.857 24.240 qs_ot_get_p 129 10.4 0.001 0.001 22.107 22.614 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 20.258 21.193 apply_single 129 13.6 0.001 0.001 20.257 21.193 ot_diis_step 118 11.6 0.018 0.019 20.016 20.017 qs_ot_p2m_diag 84 11.4 0.268 0.275 17.709 17.731 cp_dbcsr_syevd 84 12.4 0.005 0.005 16.628 16.631 multiply_cannon_sync_h2d 20232 15.6 14.155 16.227 14.155 16.227 make_m2s 5058 13.6 0.082 0.086 15.322 16.078 make_images 5058 14.6 1.195 1.295 15.086 15.846 cp_fm_cholesky_invert 11 10.9 15.380 15.393 15.380 15.393 sum_up_and_integrate 129 10.3 0.134 0.145 14.786 14.810 integrate_v_rspace 129 11.3 0.003 0.004 14.652 14.681 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.880 13.913 calculate_rho_elec 129 8.7 0.133 0.148 13.879 13.913 cp_fm_diag_elpa 84 13.4 0.000 0.001 13.343 13.346 cp_fm_redistribute_end 84 14.4 5.019 13.265 5.037 13.268 cp_fm_diag_elpa_base 84 14.4 7.782 12.639 8.204 13.142 init_scf_run 11 5.9 0.000 0.001 10.790 10.790 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.789 10.790 make_images_data 5058 15.6 0.060 0.066 9.524 10.784 hybrid_alltoall_any 5245 16.5 0.432 1.986 8.279 10.081 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.332 9.613 multiply_cannon_metrocomm4 17703 15.6 0.060 0.070 3.551 9.596 mp_irecv_dv 50659 16.2 3.430 9.346 3.430 9.346 dbcsr_mm_accdrv_process 41846 16.2 4.470 5.414 8.339 8.524 pw_transfer 1559 11.6 0.085 0.104 8.342 8.459 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 8.116 8.238 grid_integrate_task_list 129 12.3 7.305 8.063 7.305 8.063 cp_fm_cholesky_decompose 22 10.9 7.723 7.759 7.723 7.759 density_rs2pw 129 9.7 0.006 0.006 7.188 7.657 wfi_extrapolate 11 7.9 0.001 0.001 7.579 7.580 cp_fm_upper_to_full 106 14.5 5.824 7.510 5.824 7.510 rs_pw_transfer 1054 12.0 0.013 0.014 6.781 7.374 fft_wrap_pw1pw2_140 527 13.2 0.479 0.535 7.038 7.168 dbcsr_complete_redistribute 397 12.7 1.168 1.194 4.786 6.588 fft3d_ps 1301 14.7 2.714 2.923 6.374 6.447 calculate_dm_sparse 129 9.5 0.001 0.001 5.836 5.922 grid_collocate_task_list 129 9.7 5.144 5.780 5.144 5.780 mp_alltoall_d11v 2429 14.1 4.661 5.688 4.661 5.688 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.475 5.479 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.605 5.297 copy_fm_to_dbcsr 210 11.7 0.002 0.002 3.487 5.281 potential_pw2rs 129 12.3 0.020 0.022 4.997 5.023 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 5.001 5.008 mp_sum_l 7936 13.1 3.453 4.836 3.453 4.836 mp_allgather_i34 2529 14.6 1.623 4.737 1.623 4.737 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.496 4.634 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.164 4.193 mp_waitany 11836 13.9 3.554 4.161 3.554 4.161 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.395 4.143 mp_alltoall_i22 720 14.1 2.032 3.979 2.032 3.979 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=184.487000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=889.272727, yerr=8.113859 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a007b1b0cc940e469a6f710d029fae2a5232a07b_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430454546432 0.0% 0.0% 100.0% flops 32 x 32 x 32 1973537472512 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992006770688 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613065416704 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239182565376 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239182565376 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.243766E+12 0.0% 0.0% 100.0% flops max/rank 4.387745E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806547424 0.0% 0.0% 100.0% number of processed stacks 6032640 0.0% 0.0% 100.0% average stack size 0.0 0.0 1128.3 marketing flops 145.662591E+12 ------------------------------------------------------------------------------- # multiplications 2534 max memory usage/rank 1.157837E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1155504 MPI messages size (bytes): total size 2.039825E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.765312E+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 323336 36533174272 131072 < size <= 4194304 723016 792311627776 4194304 < size <= 16777216 70800 669922351760 16777216 < size 30960 541065216000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4127 56626. MP_Allreduce 11356 1063. MP_Sync 172 MP_Alltoall 1724 12509574. MP_ISendRecv 11868 75008. MP_Wait 28546 MP_comm_split 84 MP_ISend 15064 244788. MP_IRecv 15064 244788. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.053 199.838 199.839 qs_mol_dyn_low 1 2.0 0.003 0.004 199.451 199.465 qs_forces 11 3.9 0.003 0.004 199.337 199.342 qs_energies 11 4.9 0.015 0.059 192.209 192.217 scf_env_do_scf 11 5.9 0.001 0.002 172.811 172.819 velocity_verlet 10 3.0 0.001 0.001 131.047 131.050 scf_env_do_scf_inner_loop 118 6.6 0.003 0.010 123.760 123.762 qs_scf_new_mos 118 7.6 0.001 0.001 86.868 87.158 qs_scf_loop_do_ot 118 8.6 0.001 0.001 86.868 87.157 ot_scf_mini 118 9.6 0.004 0.004 82.296 82.560 dbcsr_multiply_generic 2534 12.6 0.193 0.199 80.887 81.710 multiply_cannon 2534 13.6 0.556 0.589 55.278 58.100 multiply_cannon_loop 2534 14.6 1.202 1.228 51.365 53.289 init_scf_loop 11 6.9 0.001 0.005 48.899 48.900 prepare_preconditioner 11 7.9 0.000 0.000 44.698 44.742 make_preconditioner 11 8.9 0.000 0.001 44.698 44.741 ot_mini 118 10.6 0.001 0.001 43.472 43.717 make_full_inverse_cholesky 11 9.9 0.000 0.000 38.264 43.215 multiply_cannon_multrec 30408 15.6 14.273 19.374 26.337 30.937 rebuild_ks_matrix 129 8.3 0.001 0.001 29.906 30.152 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.019 29.906 30.151 mp_waitall_1 155386 16.5 18.501 27.789 18.501 27.789 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.074 27.281 qs_ot_get_p 129 10.4 0.001 0.001 24.792 25.048 qs_ot_get_derivative 118 11.6 0.001 0.002 23.583 23.845 make_m2s 5068 13.6 0.096 0.100 21.033 22.063 make_images 5068 14.6 1.960 2.264 20.721 21.749 qs_ot_p2m_diag 84 11.4 0.346 0.394 20.428 20.482 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 19.272 19.868 apply_single 129 13.6 0.001 0.001 19.272 19.868 ot_diis_step 118 11.6 0.018 0.018 19.753 19.755 cp_dbcsr_syevd 84 12.4 0.005 0.005 19.075 19.077 cp_fm_cholesky_invert 11 10.9 17.831 17.842 17.831 17.842 cp_fm_upper_to_full 106 14.7 11.052 16.311 11.052 16.311 cp_fm_diag_elpa 84 13.4 0.000 0.001 15.631 15.636 cp_fm_redistribute_end 84 14.4 2.678 15.507 2.710 15.517 cp_fm_diag_elpa_base 84 14.4 12.158 14.821 12.762 15.325 multiply_cannon_metrocomm3 30408 15.6 0.047 0.050 6.525 15.174 sum_up_and_integrate 129 10.3 0.141 0.152 14.706 14.731 integrate_v_rspace 129 11.3 0.003 0.003 14.564 14.595 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.776 13.808 calculate_rho_elec 129 8.7 0.176 0.191 13.775 13.807 dbcsr_complete_redistribute 397 12.7 1.504 1.623 9.330 13.111 make_images_data 5068 15.6 0.064 0.067 11.393 12.881 multiply_cannon_sync_h2d 30408 15.6 11.794 12.447 11.794 12.447 init_scf_run 11 5.9 0.000 0.001 12.402 12.403 scf_env_initial_rho_setup 11 6.9 0.001 0.003 12.401 12.403 dbcsr_mm_accdrv_process 62840 16.2 7.440 8.150 11.632 12.144 hybrid_alltoall_any 5255 16.5 0.530 2.227 10.041 11.920 copy_fm_to_dbcsr 210 11.7 0.002 0.002 7.803 11.557 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.411 10.042 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.807 9.994 mp_alltoall_i22 720 14.1 5.561 9.360 5.561 9.360 pw_transfer 1559 11.6 0.085 0.101 8.383 8.476 cp_fm_cholesky_decompose 22 10.9 8.315 8.397 8.315 8.397 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 8.157 8.253 grid_integrate_task_list 129 12.3 7.584 8.002 7.584 8.002 wfi_extrapolate 11 7.9 0.001 0.001 7.834 7.834 density_rs2pw 129 9.7 0.005 0.006 6.733 7.251 fft_wrap_pw1pw2_140 527 13.2 0.483 0.493 7.086 7.174 multiply_cannon_metrocomm4 25340 15.6 0.076 0.085 2.749 6.703 fft3d_ps 1301 14.7 2.818 2.884 6.379 6.458 mp_irecv_dv 76866 16.2 2.605 6.439 2.605 6.439 calculate_dm_sparse 129 9.5 0.001 0.001 6.256 6.318 rs_pw_transfer 1054 12.0 0.013 0.014 5.658 6.194 mp_alltoall_d11v 2429 14.1 5.481 6.098 5.481 6.098 grid_collocate_task_list 129 9.7 5.245 5.665 5.245 5.665 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.587 5.639 qs_energies_init_hamiltonians 11 5.9 0.012 0.054 5.402 5.415 potential_pw2rs 129 12.3 0.022 0.022 4.694 4.722 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.595 4.671 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.420 4.480 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.320 4.374 calculate_first_density_matrix 1 7.0 0.000 0.002 4.362 4.369 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=199.839000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1077.545455, yerr=25.868070 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a007b1b0cc940e469a6f710d029fae2a5232a07b_performance_tests/19/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 5.909858E+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 1976192 0.0% 0.0% 100.0% average stack size 0.0 0.0 3444.2 marketing flops 145.646636E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 1.524892E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 242592 MPI messages size (bytes): total size 1.341705E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.530704E+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 114816 60196651008 4194304 < size <= 16777216 105840 554906419200 16777216 < size 20352 726591885696 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 9003 51. MP_Alltoall 9708 795382. MP_ISend 40388 2101966. MP_IRecv 40388 2101070. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4024 57898. MP_Allreduce 11057 1172. MP_Sync 87 MP_Alltoall 1724 18848449. MP_ISendRecv 7740 122880. MP_Wait 20114 MP_ISend 10760 423501. MP_IRecv 10760 423501. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.037 178.722 178.722 qs_mol_dyn_low 1 2.0 0.003 0.004 178.344 178.356 qs_forces 11 3.9 0.003 0.003 178.149 178.156 qs_energies 11 4.9 0.017 0.045 170.801 170.809 scf_env_do_scf 11 5.9 0.001 0.001 151.014 151.022 velocity_verlet 10 3.0 0.001 0.001 116.251 116.260 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 113.799 113.801 qs_scf_new_mos 118 7.6 0.001 0.001 76.920 77.028 qs_scf_loop_do_ot 118 8.6 0.001 0.001 76.920 77.027 dbcsr_multiply_generic 2527 12.6 0.193 0.206 75.603 76.083 ot_scf_mini 118 9.6 0.003 0.004 72.384 72.479 multiply_cannon 2527 13.6 0.584 0.619 54.995 59.929 multiply_cannon_loop 2527 14.6 0.451 0.461 50.303 51.098 ot_mini 118 10.6 0.001 0.001 40.435 40.527 init_scf_loop 11 6.9 0.000 0.000 37.051 37.053 mp_waitall_1 130650 16.6 27.119 34.618 27.119 34.618 prepare_preconditioner 11 7.9 0.000 0.000 33.117 33.136 make_preconditioner 11 8.9 0.000 0.000 33.117 33.136 make_full_inverse_cholesky 11 9.9 0.000 0.000 30.808 31.122 rebuild_ks_matrix 129 8.3 0.001 0.001 29.332 29.408 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.018 29.331 29.408 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.796 26.871 multiply_cannon_multrec 10108 15.6 10.465 14.209 17.825 20.771 qs_ot_get_derivative 118 11.6 0.002 0.002 20.321 20.422 ot_diis_step 118 11.6 0.020 0.021 20.048 20.049 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.642 19.919 apply_single 129 13.6 0.001 0.001 19.641 19.919 multiply_cannon_metrocomm3 10108 15.6 0.023 0.024 12.304 19.361 make_m2s 5054 13.6 0.067 0.074 16.965 19.316 make_images 5054 14.6 2.338 2.703 16.656 19.002 qs_ot_get_p 129 10.4 0.001 0.001 18.817 18.883 cp_fm_cholesky_invert 11 10.9 18.839 18.845 18.839 18.845 qs_ot_p2m_diag 83 11.4 0.496 0.505 14.936 14.953 sum_up_and_integrate 129 10.3 0.183 0.193 14.538 14.583 integrate_v_rspace 129 11.3 0.003 0.004 14.355 14.407 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.800 13.853 calculate_rho_elec 129 8.7 0.260 0.270 13.799 13.852 cp_dbcsr_syevd 83 12.4 0.005 0.005 13.752 13.753 make_images_data 5054 15.6 0.053 0.061 10.395 13.019 init_scf_run 11 5.9 0.000 0.001 12.782 12.782 scf_env_initial_rho_setup 11 6.9 0.011 0.021 12.781 12.782 hybrid_alltoall_any 5240 16.5 0.826 3.679 10.142 12.572 multiply_cannon_sync_h2d 10108 15.6 11.744 12.050 11.744 12.050 cp_fm_diag_elpa 83 13.4 0.000 0.000 10.636 10.637 cp_fm_diag_elpa_base 83 14.4 10.371 10.462 10.625 10.625 cp_fm_cholesky_decompose 22 10.9 8.338 8.454 8.338 8.454 pw_transfer 1559 11.6 0.083 0.091 8.365 8.398 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 8.149 8.213 mp_allgather_i34 2527 14.6 2.902 8.170 2.902 8.170 fft_wrap_pw1pw2 1301 12.7 0.010 0.010 8.142 8.169 grid_integrate_task_list 129 12.3 7.858 8.160 7.858 8.160 dbcsr_mm_accdrv_process 20926 16.1 2.663 3.684 6.984 7.618 multiply_cannon_metrocomm1 10108 15.6 0.028 0.029 4.767 7.550 wfi_extrapolate 11 7.9 0.001 0.001 7.482 7.482 fft_wrap_pw1pw2_140 527 13.2 0.508 0.534 7.090 7.114 density_rs2pw 129 9.7 0.005 0.006 6.402 6.686 fft3d_ps 1301 14.7 2.749 2.817 6.282 6.343 calculate_dm_sparse 129 9.5 0.001 0.001 6.212 6.313 dbcsr_complete_redistribute 395 12.7 2.171 2.279 5.639 6.088 grid_collocate_task_list 129 9.7 5.534 5.820 5.534 5.820 mp_alltoall_d11v 2423 14.1 5.320 5.722 5.320 5.722 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.335 5.342 qs_energies_init_hamiltonians 11 5.9 0.003 0.008 5.279 5.281 calculate_first_density_matrix 1 7.0 0.000 0.000 5.077 5.091 rs_pw_transfer 1054 12.0 0.012 0.013 4.348 4.626 potential_pw2rs 129 12.3 0.027 0.028 4.341 4.355 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.118 4.146 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.749 4.101 multiply_cannon_metrocomm4 7581 15.6 0.022 0.025 1.874 3.911 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.561 3.851 mp_irecv_dv 29086 15.9 1.840 3.848 1.840 3.848 parallel_gemm_fm 81 9.0 0.000 0.000 3.812 3.833 parallel_gemm_fm_cosma 81 10.0 3.811 3.833 3.811 3.833 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.757 3.801 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.678 3.786 make_basis_sm 11 9.8 0.000 0.000 3.756 3.761 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 3.629 3.662 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=178.722000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1430.181818, yerr=55.407282 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a007b1b0cc940e469a6f710d029fae2a5232a07b_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430454546432 0.0% 0.0% 100.0% flops 32 x 32 x 32 1975684956160 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992006770688 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613065416704 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239182565376 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239182565376 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.245913E+12 0.0% 0.0% 100.0% flops max/rank 11.787674E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806580192 0.0% 0.0% 100.0% number of processed stacks 1982496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3433.3 marketing flops 145.663816E+12 ------------------------------------------------------------------------------- # multiplications 2535 max memory usage/rank 3.029078E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101400 MPI messages size (bytes): total size 1.145171E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.293599E+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 45888 35634806784 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592540656 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4041 58592. MP_Allreduce 11099 1511. MP_Sync 88 MP_Alltoall 1724 36993632. MP_ISendRecv 3612 218624. MP_Wait 11682 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 0.028 0.051 296.680 296.680 qs_mol_dyn_low 1 2.0 0.003 0.004 296.133 296.145 qs_forces 11 3.9 0.003 0.003 296.045 296.049 qs_energies 11 4.9 0.002 0.002 287.398 287.411 scf_env_do_scf 11 5.9 0.001 0.001 264.944 264.957 velocity_verlet 10 3.0 0.001 0.001 212.810 212.817 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 139.241 139.242 init_scf_loop 11 6.9 0.000 0.000 125.430 125.432 prepare_preconditioner 11 7.9 0.000 0.000 120.158 120.183 make_preconditioner 11 8.9 0.000 0.000 120.158 120.183 make_full_inverse_cholesky 11 9.9 0.000 0.000 97.317 117.270 qs_scf_new_mos 118 7.6 0.001 0.001 96.155 96.366 qs_scf_loop_do_ot 118 8.6 0.001 0.001 96.154 96.365 ot_scf_mini 118 9.6 0.004 0.004 91.296 91.391 dbcsr_multiply_generic 2535 12.6 0.217 0.224 85.424 86.129 cp_fm_upper_to_full 106 14.8 52.872 74.570 52.872 74.570 multiply_cannon 2535 13.6 0.702 0.749 60.206 61.010 multiply_cannon_loop 2535 14.6 0.478 0.492 56.549 58.190 ot_mini 118 10.6 0.001 0.001 46.020 46.142 dbcsr_complete_redistribute 397 12.7 3.970 4.011 28.629 40.715 copy_fm_to_dbcsr 210 11.7 0.002 0.002 25.184 37.223 cp_fm_cholesky_invert 11 10.9 34.948 34.955 34.948 34.955 rebuild_ks_matrix 129 8.3 0.001 0.001 34.768 34.883 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.018 34.768 34.882 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 22.795 34.567 mp_waitall_1 106626 16.7 29.895 33.435 29.895 33.435 mp_alltoall_i22 720 14.1 20.596 32.715 20.596 32.715 qs_ks_update_qs_env 129 7.6 0.001 0.001 32.466 32.583 qs_ot_get_p 129 10.4 0.001 0.001 29.383 29.551 qs_ot_get_derivative 118 11.6 0.002 0.002 25.383 25.476 qs_ot_p2m_diag 84 11.4 0.889 0.897 25.068 25.104 cp_dbcsr_syevd 84 12.4 0.006 0.006 23.250 23.253 multiply_cannon_metrocomm3 10140 15.6 0.024 0.024 19.997 21.474 make_m2s 5070 13.6 0.077 0.079 20.221 21.115 make_images 5070 14.6 3.802 3.998 19.737 20.634 ot_diis_step 118 11.6 0.022 0.023 20.580 20.581 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.748 20.119 apply_single 129 13.6 0.001 0.001 19.748 20.119 cp_fm_diag_elpa 84 13.4 0.000 0.000 19.840 19.840 cp_fm_diag_elpa_base 84 14.4 15.439 16.959 19.835 19.835 multiply_cannon_multrec 10140 15.6 10.561 12.369 18.114 18.303 sum_up_and_integrate 129 10.3 0.326 0.329 16.105 16.220 integrate_v_rspace 129 11.3 0.004 0.004 15.778 15.895 multiply_cannon_sync_h2d 10140 15.6 15.777 15.784 15.777 15.784 qs_rho_update_rho_low 129 7.7 0.001 0.001 15.449 15.503 calculate_rho_elec 129 8.7 0.486 0.487 15.449 15.503 init_scf_run 11 5.9 0.000 0.001 12.522 12.523 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.522 12.522 hybrid_alltoall_any 5257 16.5 1.306 3.053 10.806 12.476 make_images_data 5070 15.6 0.060 0.064 10.758 12.419 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.944 10.036 cp_fm_cholesky_decompose 22 10.9 9.311 9.339 9.311 9.339 dbcsr_mm_accdrv_process 20958 16.1 3.970 5.838 7.315 9.142 wfi_extrapolate 11 7.9 0.001 0.001 8.833 8.833 pw_transfer 1559 11.6 0.091 0.091 8.783 8.810 grid_integrate_task_list 129 12.3 8.601 8.776 8.601 8.776 fft_wrap_pw1pw2 1301 12.7 0.011 0.011 8.543 8.567 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 7.927 7.928 fft_wrap_pw1pw2_140 527 13.2 0.541 0.544 7.503 7.532 mp_alltoall_d11v 2429 14.1 7.249 7.428 7.249 7.428 calculate_dm_sparse 129 9.5 0.001 0.001 6.775 6.913 fft3d_ps 1301 14.7 2.761 2.783 6.562 6.580 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.351 6.434 grid_collocate_task_list 129 9.7 6.389 6.425 6.389 6.425 density_rs2pw 129 9.7 0.005 0.005 6.358 6.386 copy_dbcsr_to_fm 187 11.8 0.004 0.004 6.157 6.243 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=296.680000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2691.363636, yerr=173.236188 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a007b1b0cc940e469a6f710d029fae2a5232a07b_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 1.261466E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255669. MP_Allreduce 3059 6274. MP_Sync 4 MP_Alltoall 54 MP_ISendRecv 570 19200. MP_Wait 1302 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.038 0.059 87.223 87.227 qs_energies 1 2.0 0.000 0.000 85.602 85.615 ls_scf 1 3.0 0.000 0.001 84.669 84.682 dbcsr_multiply_generic 111 6.7 0.014 0.015 73.372 73.596 multiply_cannon 111 7.7 0.017 0.021 56.071 57.401 multiply_cannon_loop 111 8.7 0.209 0.224 52.583 53.879 ls_scf_main 1 4.0 0.000 0.005 52.979 52.982 density_matrix_trs4 2 5.0 0.002 0.003 47.272 47.344 ls_scf_init_scf 1 4.0 0.000 0.002 28.583 28.588 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.318 27.381 mp_waitall_1 11316 10.9 23.147 26.346 23.147 26.346 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.224 25.237 multiply_cannon_multrec 2664 9.7 8.177 8.860 15.498 17.111 multiply_cannon_sync_h2d 2664 9.7 13.522 14.959 13.522 14.959 make_m2s 222 7.7 0.008 0.010 13.620 14.126 make_images 222 8.7 0.098 0.108 13.599 14.107 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 9.804 12.757 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.566 9.474 make_images_data 222 9.7 0.004 0.005 8.082 8.797 hybrid_alltoall_any 227 10.6 0.215 1.837 6.793 8.691 dbcsr_mm_accdrv_process 4760 10.4 0.505 0.610 6.942 7.861 dbcsr_mm_accdrv_process_sort 4760 11.4 6.235 7.098 6.235 7.098 calculate_norms 4752 9.8 5.517 6.197 5.517 6.197 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.172 5.294 mp_sum_l 807 5.4 3.166 4.701 3.166 4.701 make_images_sizes 222 9.7 0.000 0.000 1.030 3.825 mp_alltoall_i44 222 10.7 1.029 3.825 1.029 3.825 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.355 3.416 arnoldi_extremal 4 6.8 0.000 0.000 3.315 3.340 arnoldi_normal_ev 4 7.8 0.001 0.004 3.315 3.340 multiply_cannon_metrocomm4 2442 9.7 0.012 0.014 2.071 3.296 mp_irecv_dv 6231 10.9 2.054 3.264 2.054 3.264 build_subspace 16 8.4 0.009 0.012 3.218 3.220 ls_scf_post 1 4.0 0.001 0.005 3.107 3.120 ls_scf_store_result 1 5.0 0.000 0.000 2.921 2.967 dbcsr_special_finalize 555 9.7 0.004 0.005 2.161 2.640 make_images_pack 222 9.7 2.204 2.637 2.206 2.638 dbcsr_merge_single_wm 555 10.7 0.454 0.593 2.154 2.633 dbcsr_matrix_vector_mult 304 9.0 0.003 0.010 2.342 2.580 dbcsr_matrix_vector_mult_local 304 10.0 2.060 2.448 2.062 2.450 ls_scf_dm_to_ks 2 5.0 0.000 0.008 2.331 2.442 dbcsr_sort_data 658 11.4 1.957 2.410 1.957 2.410 buffer_matrices_ensure_size 222 8.7 1.764 2.068 1.764 2.068 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.909 1.914 rebuild_ks_matrix 3 7.3 0.000 0.000 1.899 1.905 qs_ks_build_kohn_sham_matrix 3 8.3 0.007 0.023 1.899 1.905 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=87.227000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1132.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a007b1b0cc940e469a6f710d029fae2a5232a07b_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 2.101006E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 10339. MP_Sync 4 MP_Alltoall 47 15335933. MP_ISendRecv 282 57600. MP_Wait 828 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.103 0.162 96.408 96.415 qs_energies 1 2.0 0.000 0.000 95.076 95.081 ls_scf 1 3.0 0.000 0.002 93.518 93.523 dbcsr_multiply_generic 111 6.7 0.037 0.088 78.553 78.917 multiply_cannon 111 7.7 0.027 0.043 54.847 60.676 ls_scf_main 1 4.0 0.000 0.002 57.359 57.363 multiply_cannon_loop 111 8.7 0.115 0.124 51.383 56.200 density_matrix_trs4 2 5.0 0.002 0.003 51.461 51.703 ls_scf_init_scf 1 4.0 0.001 0.002 32.587 32.589 mp_waitall_1 9246 10.9 23.619 31.691 23.619 31.691 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.353 31.486 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.909 28.926 multiply_cannon_multrec 1332 9.7 12.911 15.965 22.118 26.263 multiply_cannon_metrocomm3 1332 9.7 0.006 0.008 13.352 22.152 make_m2s 222 7.7 0.006 0.007 16.064 16.708 make_images 222 8.7 1.572 1.956 16.034 16.677 dbcsr_mm_accdrv_process 4041 10.4 0.293 0.476 8.808 10.372 make_images_data 222 9.7 0.004 0.004 9.400 10.312 dbcsr_mm_accdrv_process_sort 4041 11.4 8.394 9.921 8.394 9.921 hybrid_alltoall_any 227 10.6 0.521 2.468 8.676 9.627 mp_sum_l 807 5.4 6.320 9.542 6.320 9.542 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.244 8.061 mp_irecv_dv 3311 11.0 3.225 8.006 3.225 8.006 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.939 7.796 calculate_norms 2376 9.8 5.944 6.563 5.944 6.563 multiply_cannon_sync_h2d 1332 9.7 4.633 5.972 4.633 5.972 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.272 5.495 arnoldi_extremal 4 6.8 0.000 0.000 4.770 4.791 arnoldi_normal_ev 4 7.8 0.001 0.005 4.770 4.791 build_subspace 16 8.4 0.014 0.021 4.508 4.511 ls_scf_post 1 4.0 0.000 0.001 3.571 3.576 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.182 3.395 ls_scf_store_result 1 5.0 0.000 0.000 3.317 3.394 dbcsr_matrix_vector_mult_local 304 10.0 2.735 3.212 2.737 3.214 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.483 3.141 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.595 2.722 mp_allgather_i34 111 8.7 1.096 2.610 1.096 2.610 make_images_pack 222 9.7 2.020 2.414 2.022 2.416 dbcsr_sort_data 436 11.2 1.790 2.021 1.790 2.021 compute_matrix_preconditioner 1 6.0 0.000 0.001 1.923 1.928 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=96.415000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1689.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a007b1b0cc940e469a6f710d029fae2a5232a07b_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 2.698572E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265470. MP_Allreduce 3058 11181. MP_Sync 4 MP_Alltoall 47 23526250. MP_ISendRecv 186 57600. MP_Wait 732 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.043 0.093 95.574 95.577 qs_energies 1 2.0 0.000 0.000 94.879 94.893 ls_scf 1 3.0 0.000 0.001 93.471 93.484 dbcsr_multiply_generic 111 6.7 0.016 0.021 77.915 78.199 ls_scf_main 1 4.0 0.000 0.004 58.552 58.560 multiply_cannon 111 7.7 0.040 0.114 53.634 58.144 multiply_cannon_loop 111 8.7 0.100 0.106 49.745 53.254 density_matrix_trs4 2 5.0 0.002 0.002 52.499 52.644 mp_waitall_1 7374 11.0 24.871 34.690 24.871 34.690 ls_scf_init_scf 1 4.0 0.000 0.001 31.310 31.314 ls_scf_init_matrix_S 1 5.0 0.000 0.000 30.093 30.166 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.481 27.492 multiply_cannon_multrec 888 9.7 12.567 15.269 21.111 24.347 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.421 23.053 make_m2s 222 7.7 0.006 0.007 17.490 18.998 make_images 222 8.7 1.978 2.295 17.452 18.960 make_images_data 222 9.7 0.004 0.009 10.167 11.401 hybrid_alltoall_any 227 10.6 0.622 2.850 9.763 11.033 dbcsr_mm_accdrv_process 3754 10.4 0.252 0.414 8.064 9.281 dbcsr_mm_accdrv_process_sort 3754 11.4 7.694 8.866 7.694 8.866 mp_sum_l 807 5.4 5.259 8.831 5.259 8.831 multiply_cannon_sync_h2d 888 9.7 5.986 7.125 5.986 7.125 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.455 6.969 mp_irecv_dv 2335 11.1 2.440 6.929 2.440 6.929 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.087 6.878 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.892 6.867 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.228 5.437 arnoldi_extremal 4 6.8 0.000 0.000 5.223 5.238 arnoldi_normal_ev 4 7.8 0.001 0.005 5.223 5.238 build_subspace 16 8.4 0.014 0.020 4.915 4.921 calculate_norms 1584 9.8 4.263 4.554 4.263 4.554 mp_allgather_i34 111 8.7 1.729 4.319 1.729 4.319 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.488 3.866 ls_scf_post 1 4.0 0.000 0.001 3.608 3.624 dbcsr_matrix_vector_mult_local 304 10.0 3.013 3.581 3.015 3.583 ls_scf_store_result 1 5.0 0.000 0.000 3.350 3.425 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.831 2.930 make_images_sizes 222 9.7 0.000 0.000 0.976 2.216 mp_alltoall_i44 222 10.7 0.976 2.216 0.976 2.216 make_images_pack 222 9.7 1.838 2.146 1.841 2.148 dbcsr_sort_data 325 11.1 1.880 2.128 1.880 2.128 dbcsr_data_release 9322 10.9 1.318 2.012 1.318 2.012 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.919 1.923 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=95.577000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2163.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a007b1b0cc940e469a6f710d029fae2a5232a07b_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 3.344036E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 13371. MP_Sync 4 MP_Alltoall 47 30278988. MP_ISendRecv 138 86400. MP_Wait 600 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.037 0.049 97.908 97.910 qs_energies 1 2.0 0.000 0.000 97.236 97.242 ls_scf 1 3.0 0.000 0.001 95.583 95.590 dbcsr_multiply_generic 111 6.7 0.016 0.018 79.001 79.282 ls_scf_main 1 4.0 0.000 0.001 59.332 59.333 multiply_cannon 111 7.7 0.049 0.121 51.809 56.423 density_matrix_trs4 2 5.0 0.002 0.003 53.229 53.403 multiply_cannon_loop 111 8.7 0.114 0.127 46.573 50.504 ls_scf_init_scf 1 4.0 0.008 0.020 33.041 33.044 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.857 31.939 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.225 29.239 mp_waitall_1 6438 11.0 22.846 28.391 22.846 28.391 multiply_cannon_multrec 1332 9.7 14.240 17.319 22.102 24.932 make_m2s 222 7.7 0.007 0.008 21.229 22.703 make_images 222 8.7 3.129 3.602 21.179 22.654 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.194 16.694 make_images_data 222 9.7 0.004 0.004 11.905 13.574 hybrid_alltoall_any 227 10.6 0.799 3.787 11.252 13.184 dbcsr_mm_accdrv_process 3641 10.4 0.211 0.410 7.500 9.044 dbcsr_mm_accdrv_process_sort 3641 11.4 7.128 8.628 7.128 8.628 mp_sum_l 807 5.4 4.389 8.370 4.389 8.370 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.414 6.643 multiply_cannon_sync_h2d 1332 9.7 5.517 6.118 5.517 6.118 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.057 5.996 mp_irecv_dv 3229 10.9 2.034 5.927 2.034 5.927 arnoldi_extremal 4 6.8 0.000 0.000 5.441 5.455 arnoldi_normal_ev 4 7.8 0.101 0.123 5.441 5.455 build_subspace 16 8.4 0.014 0.021 4.996 5.006 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.568 4.973 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.576 4.753 mp_allgather_i34 111 8.7 2.218 4.693 2.218 4.693 calculate_norms 2376 9.8 4.190 4.516 4.190 4.516 dbcsr_matrix_vector_mult 304 9.0 0.006 0.016 3.664 3.936 dbcsr_matrix_vector_mult_local 304 10.0 3.183 3.682 3.185 3.683 dbcsr_sort_data 658 11.4 3.018 3.502 3.018 3.502 dbcsr_special_finalize 555 9.7 0.006 0.007 2.754 3.221 ls_scf_post 1 4.0 0.000 0.001 3.209 3.215 dbcsr_merge_single_wm 555 10.7 0.530 0.657 2.746 3.213 ls_scf_store_result 1 5.0 0.000 0.000 2.958 3.014 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.929 2.987 dbcsr_data_release 10477 10.7 1.595 2.411 1.595 2.411 dbcsr_finalize 304 7.8 0.049 0.061 1.800 2.002 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.910000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2696.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a007b1b0cc940e469a6f710d029fae2a5232a07b_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 4.638638E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265558. MP_Allreduce 3049 15663. MP_Sync 4 MP_Alltoall 47 46208988. MP_ISendRecv 90 115200. MP_Wait 573 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.171 0.367 101.218 101.218 qs_energies 1 2.0 0.000 0.000 99.358 99.363 ls_scf 1 3.0 0.000 0.000 97.432 97.436 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.701 78.890 ls_scf_main 1 4.0 0.000 0.000 62.493 62.493 multiply_cannon 111 7.7 0.076 0.137 55.847 61.400 density_matrix_trs4 2 5.0 0.002 0.003 55.469 55.581 multiply_cannon_loop 111 8.7 0.069 0.078 51.262 52.954 mp_waitall_1 5481 11.0 27.324 32.970 27.324 32.970 ls_scf_init_scf 1 4.0 0.000 0.001 31.298 31.304 ls_scf_init_matrix_S 1 5.0 0.000 0.000 30.077 30.109 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.889 27.902 multiply_cannon_multrec 444 9.7 13.972 16.044 21.011 23.964 make_m2s 222 7.7 0.004 0.005 17.935 20.503 make_images 222 8.7 3.714 4.427 17.873 20.442 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.728 16.729 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.172 14.397 make_images_data 222 9.7 0.003 0.004 10.216 12.681 hybrid_alltoall_any 227 10.6 0.787 3.773 9.945 12.409 multiply_cannon_sync_h2d 444 9.7 6.537 8.803 6.537 8.803 dbcsr_mm_accdrv_process 3003 10.4 0.164 0.343 6.747 7.891 dbcsr_mm_accdrv_process_sort 3003 11.4 6.432 7.545 6.432 7.545 mp_allgather_i34 111 8.7 2.814 7.082 2.814 7.082 arnoldi_extremal 4 6.8 0.000 0.000 6.010 6.026 arnoldi_normal_ev 4 7.8 0.005 0.013 6.010 6.026 build_subspace 16 8.4 0.015 0.019 5.613 5.625 mp_sum_l 807 5.4 2.958 5.076 2.958 5.076 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.630 4.767 dbcsr_matrix_vector_mult 304 9.0 0.007 0.016 4.241 4.484 multiply_cannon_metrocomm4 333 9.7 0.001 0.001 1.601 4.333 mp_irecv_dv 1241 11.2 1.583 4.310 1.583 4.310 dbcsr_matrix_vector_mult_local 304 10.0 3.696 4.171 3.698 4.173 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.029 3.782 calculate_norms 792 9.8 3.542 3.725 3.542 3.725 make_images_sizes 222 9.7 0.000 0.000 1.079 3.655 mp_alltoall_i44 222 10.7 1.079 3.655 1.079 3.655 ls_scf_post 1 4.0 0.000 0.000 3.641 3.645 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.397 3.496 ls_scf_store_result 1 5.0 0.000 0.000 3.430 3.466 dbcsr_finalize 304 7.8 0.062 0.078 2.200 2.330 dbcsr_merge_all 275 8.9 0.473 0.534 2.051 2.164 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=101.218000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3624.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a007b1b0cc940e469a6f710d029fae2a5232a07b_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 8.722792E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284111. MP_Allreduce 3043 21950. MP_Sync 4 MP_Alltoall 47 88727262. MP_ISendRecv 84 732600. MP_Wait 309 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.101 0.184 109.567 109.611 qs_energies 1 2.0 0.000 0.000 108.041 108.054 ls_scf 1 3.0 0.000 0.000 105.077 105.089 dbcsr_multiply_generic 111 6.7 0.023 0.026 78.892 79.019 ls_scf_main 1 4.0 0.000 0.000 66.700 66.702 density_matrix_trs4 2 5.0 0.002 0.003 57.765 57.815 multiply_cannon 111 7.7 0.135 0.203 51.033 53.109 multiply_cannon_loop 111 8.7 0.067 0.070 47.697 48.709 ls_scf_init_scf 1 4.0 0.002 0.003 34.637 34.638 ls_scf_init_matrix_S 1 5.0 0.000 0.000 33.296 33.314 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 30.281 30.296 mp_waitall_1 4569 11.1 23.594 27.089 23.594 27.089 make_m2s 222 7.7 0.005 0.005 24.088 24.937 make_images 222 8.7 4.600 4.995 23.982 24.828 multiply_cannon_multrec 444 9.7 17.871 18.757 22.519 23.383 hybrid_alltoall_any 227 10.6 1.667 3.642 13.114 15.864 make_images_data 222 9.7 0.003 0.003 13.286 15.601 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 11.563 12.441 multiply_cannon_sync_h2d 444 9.7 8.843 8.897 8.843 8.897 arnoldi_extremal 4 6.8 0.000 0.000 7.638 7.657 arnoldi_normal_ev 4 7.8 0.003 0.009 7.638 7.657 build_subspace 16 8.4 0.025 0.036 7.091 7.101 dbcsr_matrix_vector_mult 304 9.0 0.009 0.025 5.643 5.806 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.263 5.484 dbcsr_matrix_vector_mult_local 304 10.0 5.052 5.381 5.055 5.383 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.832 4.925 dbcsr_mm_accdrv_process 1814 10.4 0.288 0.476 4.478 4.567 dbcsr_mm_accdrv_process_sort 1814 11.4 4.140 4.272 4.140 4.272 ls_scf_post 1 4.0 0.000 0.000 3.739 3.752 make_images_sizes 222 9.7 0.000 0.000 1.462 3.546 mp_alltoall_i44 222 10.7 1.462 3.545 1.462 3.545 ls_scf_store_result 1 5.0 0.000 0.000 3.473 3.485 mp_allgather_i34 111 8.7 0.993 3.325 0.993 3.325 calculate_norms 792 9.8 3.245 3.293 3.245 3.293 dbcsr_finalize 304 7.8 0.082 0.089 3.099 3.148 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.934 2.934 dbcsr_merge_all 275 8.9 0.893 0.915 2.886 2.925 dbcsr_complete_redistribute 5 7.6 1.426 1.470 2.788 2.897 dbcsr_data_release 12724 10.6 2.335 2.886 2.335 2.886 matrix_ls_to_qs 2 6.0 0.000 0.000 2.427 2.553 dbcsr_sort_data 325 11.1 2.452 2.515 2.452 2.515 dbcsr_new_transposed 4 7.5 0.244 0.252 2.343 2.363 dbcsr_frobenius_norm 74 6.6 2.059 2.135 2.196 2.232 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.226 2.228 dbcsr_add_d 103 6.2 0.000 0.000 2.133 2.203 dbcsr_add_anytype 103 7.2 0.859 0.892 2.132 2.202 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=109.611000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6820.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: a007b1b0cc940e469a6f710d029fae2a5232a07b Summary: empty Status: OK