=== 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: 007813510527d819d1aa75b16315fe3ed2c20304 ################# 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/007813510527d819d1aa75b16315fe3ed2c20304_performance_tests/01 job id: 42431355 --- 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/007813510527d819d1aa75b16315fe3ed2c20304_performance_tests/02 job id: 42431356 --- 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/007813510527d819d1aa75b16315fe3ed2c20304_performance_tests/03 job id: 42431358 --- 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/007813510527d819d1aa75b16315fe3ed2c20304_performance_tests/04 job id: 42431359 --- 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/007813510527d819d1aa75b16315fe3ed2c20304_performance_tests/05 job id: 42431360 --- 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/007813510527d819d1aa75b16315fe3ed2c20304_performance_tests/06 job id: 42431361 --- 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/007813510527d819d1aa75b16315fe3ed2c20304_performance_tests/07 job id: 42431362 --- 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/007813510527d819d1aa75b16315fe3ed2c20304_performance_tests/08 job id: 42431363 --- 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/007813510527d819d1aa75b16315fe3ed2c20304_performance_tests/09 job id: 42431364 --- 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/007813510527d819d1aa75b16315fe3ed2c20304_performance_tests/10 job id: 42431365 --- 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/007813510527d819d1aa75b16315fe3ed2c20304_performance_tests/11 job id: 42431366 --- 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/007813510527d819d1aa75b16315fe3ed2c20304_performance_tests/12 job id: 42431367 --- 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/007813510527d819d1aa75b16315fe3ed2c20304_performance_tests/13 job id: 42431368 --- 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/007813510527d819d1aa75b16315fe3ed2c20304_performance_tests/14 job id: 42431369 --- 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/007813510527d819d1aa75b16315fe3ed2c20304_performance_tests/15 job id: 42431370 --- 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/007813510527d819d1aa75b16315fe3ed2c20304_performance_tests/16 job id: 42431371 --- 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/007813510527d819d1aa75b16315fe3ed2c20304_performance_tests/17 job id: 42431373 --- 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/007813510527d819d1aa75b16315fe3ed2c20304_performance_tests/18 job id: 42431374 --- 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/007813510527d819d1aa75b16315fe3ed2c20304_performance_tests/19 job id: 42431375 --- 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/007813510527d819d1aa75b16315fe3ed2c20304_performance_tests/20 job id: 42431376 --- 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/007813510527d819d1aa75b16315fe3ed2c20304_performance_tests/21 job id: 42431377 --- 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/007813510527d819d1aa75b16315fe3ed2c20304_performance_tests/22 job id: 42431378 --- 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/007813510527d819d1aa75b16315fe3ed2c20304_performance_tests/23 job id: 42431379 --- 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/007813510527d819d1aa75b16315fe3ed2c20304_performance_tests/24 job id: 42431380 --- 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/007813510527d819d1aa75b16315fe3ed2c20304_performance_tests/25 job id: 42431381 --- 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/007813510527d819d1aa75b16315fe3ed2c20304_performance_tests/26 job id: 42431382 --- 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/007813510527d819d1aa75b16315fe3ed2c20304_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 344 9. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.022 0.035 132.375 132.376 farming_run 1 2.0 131.803 131.805 132.343 132.347 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.453449E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 228 1113141. MP_Allreduce 489 2263609. MP_Sync 27 MP_Alltoall 38 9316958. MP_SendRecv 30 829726. MP_ISendRecv 135 235435. MP_Wait 281 MP_comm_split 8 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.032 115.152 115.152 qs_energies 1 2.0 0.000 0.000 114.902 114.903 mp2_main 1 3.0 0.000 0.000 112.943 112.944 mp2_gpw_main 1 4.0 0.019 0.026 112.096 112.097 mp2_ri_gpw_compute_in 1 5.0 0.171 0.172 93.196 93.608 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 55.524 55.937 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.169 41.844 47.204 get_2c_integrals 1 6.0 0.000 0.000 36.700 37.500 integrate_v_rspace 273 8.0 0.436 0.449 25.189 30.295 pw_transfer 6555 10.6 0.374 0.384 27.423 28.011 fft_wrap_pw1pw2 5465 11.4 0.045 0.047 26.101 26.678 grid_integrate_task_list 273 9.0 20.985 26.595 20.985 26.595 fft_wrap_pw1pw2_100 2178 12.4 1.167 1.253 23.662 24.241 compute_2c_integrals 1 7.0 0.002 0.002 19.352 19.353 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.882 19.073 mp2_eri_2c_integrate_gpw 1 9.0 2.394 2.437 18.878 19.069 rpa_ri_compute_en 1 5.0 0.000 0.000 18.790 19.026 cp_fm_cholesky_decompose 12 8.2 17.306 18.101 17.306 18.101 cholesky_decomp 1 7.0 0.000 0.000 16.196 16.999 fft3d_s 5443 13.4 16.191 16.676 16.213 16.697 ao_to_mo_and_store_B_mult_1 272 7.0 10.851 15.567 10.851 15.567 calculate_wavefunction 272 8.0 5.401 5.550 12.563 13.136 rpa_num_int 1 6.0 0.000 0.000 10.624 10.633 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.577 10.594 calc_mat_Q 8 8.0 0.000 0.000 9.418 9.496 contract_S_to_Q 8 9.0 0.000 0.000 8.839 8.919 calc_potential_gpw 544 9.5 0.005 0.005 8.254 8.616 parallel_gemm_fm 14 9.1 0.000 0.000 8.429 8.503 parallel_gemm_fm_cosma 14 10.1 8.429 8.503 8.429 8.503 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.208 8.419 potential_pw2rs 545 10.0 0.107 0.110 7.708 8.281 create_integ_mat 1 6.0 0.014 0.026 7.660 7.669 collocate_single_gaussian 272 10.0 0.041 0.043 7.461 7.665 array2fm 1 7.0 0.000 0.000 6.689 7.081 pw_scatter_s 2720 13.7 4.440 4.584 4.440 4.584 pw_gather_s 2722 13.2 3.933 4.205 3.933 4.205 array2fm_buffer_send 1 8.0 2.932 3.155 2.932 3.155 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.096195, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2724.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/007813510527d819d1aa75b16315fe3ed2c20304_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 0.034 0.061 395.482 395.484 farming_run 1 2.0 394.382 394.386 395.421 395.423 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.224024E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 703 408373. MP_Allreduce 1825 23678. MP_Sync 38 MP_Alltoall 77 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 0.013 0.048 209.172 209.173 qs_energies 1 2.0 0.001 0.010 208.957 208.963 scf_env_do_scf 1 3.0 0.000 0.000 106.065 106.065 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.209 105.216 rebuild_ks_matrix 4 6.0 0.000 0.000 105.208 105.215 qs_ks_build_kohn_sham_matrix 4 7.0 0.056 0.063 105.208 105.215 hfx_ks_matrix 4 8.0 0.001 0.001 104.833 104.836 integrate_four_center 4 9.0 0.144 0.456 104.832 104.836 mp2_main 1 3.0 0.000 0.003 102.611 102.619 mp2_gpw_main 1 4.0 0.033 0.055 101.761 101.772 integrate_four_center_main 4 10.0 0.083 0.471 96.529 98.965 integrate_four_center_bin 262 11.0 96.446 98.898 96.446 98.898 init_scf_loop 1 4.0 0.000 0.000 91.862 91.863 mp2_ri_gpw_compute_in 1 5.0 0.067 0.090 74.855 75.913 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.003 54.454 55.514 mp2_eri_3c_integrate_gpw 91 7.0 0.145 0.161 42.185 47.152 integrate_v_rspace 95 8.0 0.397 0.573 28.541 33.350 pw_transfer 2240 10.6 0.143 0.167 29.917 30.454 fft_wrap_pw1pw2 1868 11.4 0.017 0.021 28.918 29.481 grid_integrate_task_list 95 9.0 23.829 28.848 23.829 28.848 mp2_ri_gpw_compute_en 1 5.0 0.055 0.063 26.731 28.325 ao_to_mo_and_store_B_mult_1 91 7.0 10.589 27.261 10.589 27.261 fft_wrap_pw1pw2_100 730 12.4 1.294 1.459 26.655 27.167 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.829 1.886 25.041 25.051 get_2c_integrals 1 6.0 0.002 0.020 20.307 20.332 compute_2c_integrals 1 7.0 0.004 0.013 19.282 19.290 compute_2c_integrals_loop_lm 1 8.0 0.007 0.044 18.917 19.155 mp2_eri_2c_integrate_gpw 1 9.0 1.744 1.913 18.911 19.154 fft3d_s 1823 13.4 18.408 18.694 18.421 18.708 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.201 14.201 calculate_wavefunction 91 8.0 2.020 2.054 9.756 10.012 mp2_ri_gpw_compute_en_expansio 172 7.0 0.558 0.583 8.812 9.603 potential_pw2rs 186 10.0 0.034 0.035 8.633 9.198 local_gemm 172 8.0 8.253 9.021 8.253 9.021 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.008 8.270 8.655 mp2_ri_gpw_compute_en_comm 22 7.0 0.497 0.514 8.015 8.638 calc_potential_gpw 182 9.5 0.002 0.002 7.949 8.156 collocate_single_gaussian 91 10.0 0.019 0.052 7.900 8.141 mp_sendrecv_dm3 2068 8.0 6.052 6.664 6.052 6.664 mp2_ri_gpw_compute_en_ener 172 7.0 6.341 6.454 6.341 6.454 pw_gather_s 912 13.2 4.905 5.393 4.905 5.393 mp_sync 38 10.4 2.949 4.915 2.949 4.915 pw_scatter_s 910 13.7 3.944 4.194 3.944 4.194 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.757189, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1511.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/007813510527d819d1aa75b16315fe3ed2c20304_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 451.895296E+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 592243. 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 0.023 0.064 48.056 48.057 qs_mol_dyn_low 1 2.0 0.005 0.018 47.640 47.648 qs_forces 11 3.9 0.002 0.003 47.559 47.562 qs_energies 11 4.9 0.001 0.002 46.107 46.116 scf_env_do_scf 11 5.9 0.000 0.001 40.287 40.288 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 38.242 38.243 dbcsr_multiply_generic 2286 12.5 0.093 0.099 30.081 30.497 qs_scf_new_mos 108 7.5 0.000 0.001 28.266 28.557 qs_scf_loop_do_ot 108 8.5 0.000 0.001 28.265 28.556 ot_scf_mini 108 9.5 0.002 0.002 26.717 26.891 multiply_cannon 2286 13.5 0.182 0.189 24.394 25.752 multiply_cannon_loop 2286 14.5 1.457 1.531 23.684 25.077 velocity_verlet 10 3.0 0.001 0.002 23.179 23.180 ot_mini 108 10.5 0.001 0.001 16.703 16.941 qs_ot_get_derivative 108 11.5 0.001 0.001 13.926 14.103 mp_waitall_1 267858 16.1 7.074 12.853 7.074 12.853 multiply_cannon_metrocomm3 54864 15.5 0.069 0.072 4.898 11.253 multiply_cannon_multrec 54864 15.5 4.450 6.791 7.591 11.020 rebuild_ks_matrix 119 8.3 0.000 0.000 7.902 8.025 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.012 7.901 8.025 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.976 7.098 multiply_cannon_sync_h2d 54864 15.5 6.193 6.786 6.193 6.786 qs_ot_get_p 119 10.4 0.001 0.001 5.711 5.980 mp_sum_l 7207 12.9 3.683 5.297 3.683 5.297 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 4.760 4.856 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 4.416 4.846 init_scf_run 11 5.9 0.000 0.001 4.580 4.580 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.580 4.580 sum_up_and_integrate 119 10.3 0.012 0.014 4.544 4.550 integrate_v_rspace 119 11.3 0.002 0.002 4.532 4.539 dbcsr_mm_accdrv_process 76910 16.1 1.063 1.673 3.063 4.281 qs_rho_update_rho_low 119 7.7 0.001 0.002 4.055 4.115 calculate_rho_elec 119 8.7 0.011 0.017 4.054 4.115 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.326 3.356 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.852 3.048 apply_single 119 13.6 0.000 0.000 2.851 3.048 rs_pw_transfer 974 11.9 0.012 0.013 2.887 2.964 calculate_dm_sparse 119 9.5 0.000 0.000 2.783 2.935 cp_dbcsr_syevd 50 12.0 0.002 0.003 2.615 2.615 ot_diis_step 108 11.5 0.006 0.006 2.594 2.595 jit_kernel_multiply 13 15.8 1.942 2.518 1.942 2.518 calculate_first_density_matrix 1 7.0 0.000 0.000 2.508 2.513 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.394 2.394 cp_fm_redistribute_end 50 14.0 2.172 2.372 2.176 2.373 density_rs2pw 119 9.7 0.004 0.004 2.239 2.348 cp_fm_diag_elpa_base 50 14.0 0.195 2.337 0.195 2.346 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.092 2.142 grid_integrate_task_list 119 12.3 2.018 2.124 2.018 2.124 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.111 2.113 multiply_cannon_metrocomm1 54864 15.5 0.054 0.059 1.196 2.051 acc_transpose_blocks 54864 15.5 0.210 0.230 1.629 2.049 init_scf_loop 11 6.9 0.000 0.000 2.024 2.025 wfi_extrapolate 11 7.9 0.001 0.001 2.011 2.011 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.905 1.946 potential_pw2rs 119 12.3 0.004 0.004 1.917 1.933 make_m2s 4572 13.5 0.053 0.055 1.696 1.743 pw_transfer 1439 11.6 0.051 0.055 1.665 1.742 mp_sum_d 4129 12.0 1.106 1.736 1.106 1.736 fft_wrap_pw1pw2 1201 12.6 0.006 0.007 1.591 1.669 make_images 4572 14.5 0.132 0.138 1.615 1.662 mp_waitany 12084 13.8 1.279 1.443 1.279 1.443 fft3d_ps 1201 14.6 0.357 0.463 1.370 1.440 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.393 1.413 mp_alltoall_d11v 2130 13.8 1.236 1.382 1.236 1.382 grid_collocate_task_list 119 9.7 1.288 1.346 1.288 1.346 fft_wrap_pw1pw2_140 487 13.2 0.079 0.093 1.227 1.303 parallel_gemm_fm 81 9.0 0.000 0.000 1.010 1.014 parallel_gemm_fm_cosma 81 10.0 1.010 1.014 1.010 1.014 prepare_preconditioner 11 7.9 0.000 0.000 0.955 0.988 make_preconditioner 11 8.9 0.000 0.000 0.955 0.988 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.970 0.987 acc_transpose_blocks_kernels 54864 16.5 0.236 0.361 0.727 0.980 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=48.057000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=430.272727, yerr=0.962091 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/007813510527d819d1aa75b16315fe3ed2c20304_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 487.989248E+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 1032117. 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 0.029 0.055 39.031 39.032 qs_mol_dyn_low 1 2.0 0.003 0.003 38.707 38.715 qs_forces 11 3.9 0.002 0.002 38.649 38.650 qs_energies 11 4.9 0.003 0.010 36.835 36.839 scf_env_do_scf 11 5.9 0.001 0.001 31.615 31.615 scf_env_do_scf_inner_loop 108 6.5 0.002 0.008 29.054 29.055 dbcsr_multiply_generic 2286 12.5 0.101 0.104 21.559 21.953 qs_scf_new_mos 108 7.5 0.001 0.001 19.935 20.184 qs_scf_loop_do_ot 108 8.5 0.001 0.001 19.935 20.183 ot_scf_mini 108 9.5 0.003 0.003 19.037 19.214 velocity_verlet 10 3.0 0.001 0.002 18.468 18.469 multiply_cannon 2286 13.5 0.208 0.217 16.396 17.925 multiply_cannon_loop 2286 14.5 0.900 0.979 15.249 16.769 ot_mini 108 10.5 0.001 0.001 11.893 12.129 mp_waitall_1 217478 16.2 6.354 11.602 6.354 11.602 multiply_cannon_metrocomm3 27432 15.5 0.068 0.070 4.207 9.694 qs_ot_get_derivative 108 11.5 0.001 0.001 9.422 9.598 multiply_cannon_multrec 27432 15.5 1.969 4.488 5.851 8.590 rebuild_ks_matrix 119 8.3 0.000 0.000 7.399 7.539 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 7.398 7.538 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.437 6.564 dbcsr_mm_accdrv_process 47894 16.0 2.809 4.976 3.812 5.652 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.628 4.479 qs_ot_get_p 119 10.4 0.001 0.001 4.233 4.463 sum_up_and_integrate 119 10.3 0.024 0.027 4.416 4.437 integrate_v_rspace 119 11.3 0.002 0.002 4.392 4.416 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.056 4.142 apply_single 119 13.6 0.000 0.000 3.056 4.142 mp_sum_l 7207 12.9 2.109 4.075 2.109 4.075 init_scf_run 11 5.9 0.000 0.001 3.968 3.968 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.968 3.968 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.661 3.703 calculate_rho_elec 119 8.7 0.021 0.024 3.660 3.702 make_m2s 4572 13.5 0.052 0.054 2.606 2.850 rs_pw_transfer 974 11.9 0.010 0.011 2.759 2.845 multiply_cannon_sync_h2d 27432 15.5 2.182 2.830 2.182 2.830 make_images 4572 14.5 0.201 0.238 2.517 2.759 qs_ot_p2m_diag 50 11.0 0.008 0.012 2.711 2.731 init_scf_loop 11 6.9 0.001 0.003 2.539 2.541 calculate_first_density_matrix 1 7.0 0.001 0.009 2.424 2.427 ot_diis_step 108 11.5 0.011 0.011 2.418 2.419 qs_ot_get_derivative_diag 49 12.0 0.001 0.002 2.194 2.284 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.272 2.273 potential_pw2rs 119 12.3 0.006 0.006 2.159 2.188 density_rs2pw 119 9.7 0.004 0.004 2.027 2.114 jit_kernel_multiply 11 16.1 0.951 2.112 0.951 2.112 calculate_dm_sparse 119 9.5 0.000 0.001 2.036 2.110 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.018 2.022 pw_transfer 1439 11.6 0.063 0.066 1.889 1.934 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.914 1.914 grid_integrate_task_list 119 12.3 1.833 1.914 1.833 1.914 cp_fm_redistribute_end 50 14.0 1.575 1.889 1.578 1.890 cp_fm_diag_elpa_base 50 14.0 0.303 1.830 0.310 1.862 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 1.800 1.846 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.786 1.823 prepare_preconditioner 11 7.9 0.000 0.000 1.595 1.623 make_preconditioner 11 8.9 0.000 0.001 1.595 1.623 make_images_data 4572 15.5 0.045 0.052 1.176 1.611 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.495 1.555 fft3d_ps 1201 14.6 0.496 0.548 1.507 1.546 hybrid_alltoall_any 4725 16.4 0.051 0.112 1.024 1.506 wfi_extrapolate 11 7.9 0.001 0.001 1.485 1.485 acc_transpose_blocks 27432 15.5 0.108 0.112 1.187 1.477 fft_wrap_pw1pw2_140 487 13.2 0.076 0.084 1.390 1.437 mp_allgather_i34 2286 14.5 0.589 1.362 0.589 1.362 mp_alltoall_d11v 2130 13.8 1.220 1.327 1.220 1.327 grid_collocate_task_list 119 9.7 1.241 1.322 1.241 1.322 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.248 1.256 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.198 1.243 mp_sum_d 4129 12.0 0.606 1.049 0.606 1.049 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.018 1.031 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.966 0.967 rs_pw_transfer_PW2RS_50 119 14.3 0.589 0.609 0.829 0.884 acc_transpose_blocks_kernels 27432 16.5 0.183 0.272 0.655 0.871 make_images_sizes 4572 15.5 0.005 0.005 0.610 0.865 mp_alltoall_z22v 1201 16.6 0.782 0.862 0.782 0.862 mp_alltoall_i44 4572 16.5 0.605 0.861 0.605 0.861 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=39.032000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=465.090909, yerr=1.975051 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/007813510527d819d1aa75b16315fe3ed2c20304_performance_tests/05/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 59.051995E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3143552 0.0% 0.0% 100.0% average stack size 0.0 0.0 46.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 521.166848E+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 0.039 0.069 33.341 33.342 qs_mol_dyn_low 1 2.0 0.003 0.003 32.921 32.931 qs_forces 11 3.9 0.002 0.008 32.530 32.534 qs_energies 11 4.9 0.004 0.009 30.923 30.929 scf_env_do_scf 11 5.9 0.001 0.002 26.066 26.067 scf_env_do_scf_inner_loop 108 6.5 0.003 0.008 23.400 23.401 dbcsr_multiply_generic 2286 12.5 0.095 0.110 16.682 16.813 velocity_verlet 10 3.0 0.002 0.004 15.941 15.953 qs_scf_new_mos 108 7.5 0.001 0.001 15.236 15.250 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.235 15.250 ot_scf_mini 108 9.5 0.003 0.003 14.493 14.503 multiply_cannon 2286 13.5 0.195 0.201 13.085 13.809 multiply_cannon_loop 2286 14.5 0.638 0.664 12.248 13.002 ot_mini 108 10.5 0.001 0.001 8.965 8.981 qs_ot_get_derivative 108 11.5 0.001 0.001 7.414 7.435 multiply_cannon_multrec 18288 15.5 1.921 2.870 6.758 7.120 rebuild_ks_matrix 119 8.3 0.000 0.000 6.474 6.496 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.015 6.473 6.496 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.705 5.725 dbcsr_mm_accdrv_process 38222 16.0 4.058 5.270 4.753 5.553 sum_up_and_integrate 119 10.3 0.029 0.030 4.059 4.068 integrate_v_rspace 119 11.3 0.002 0.003 4.030 4.040 mp_waitall_1 169478 16.3 3.185 4.035 3.185 4.035 init_scf_run 11 5.9 0.001 0.004 3.588 3.588 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.587 3.588 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.767 3.385 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.358 3.366 calculate_rho_elec 119 8.7 0.030 0.031 3.357 3.365 qs_ot_get_p 119 10.4 0.001 0.001 3.316 3.333 init_scf_loop 11 6.9 0.001 0.003 2.645 2.651 rs_pw_transfer 974 11.9 0.009 0.010 2.444 2.574 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.063 2.362 apply_single 119 13.6 0.000 0.000 2.063 2.362 calculate_first_density_matrix 1 7.0 0.001 0.003 2.343 2.344 multiply_cannon_metrocomm3 18288 15.5 0.045 0.047 1.432 2.172 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.163 2.171 make_m2s 4572 13.5 0.044 0.045 1.984 2.145 density_rs2pw 119 9.7 0.004 0.004 1.959 2.068 make_images 4572 14.5 0.192 0.204 1.899 2.060 grid_integrate_task_list 119 12.3 1.805 1.895 1.805 1.895 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.886 1.886 calculate_dm_sparse 119 9.5 0.000 0.000 1.861 1.870 jit_kernel_multiply 10 16.1 0.643 1.867 0.643 1.867 pw_transfer 1439 11.6 0.063 0.066 1.837 1.855 potential_pw2rs 119 12.3 0.007 0.008 1.840 1.851 prepare_preconditioner 11 7.9 0.000 0.000 1.818 1.821 make_preconditioner 11 8.9 0.000 0.001 1.818 1.821 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.782 1.784 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 1.747 1.765 make_full_inverse_cholesky 11 9.9 0.000 0.001 1.676 1.755 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.635 1.641 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.623 1.625 cp_fm_redistribute_end 50 14.0 1.201 1.598 1.203 1.598 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.568 1.578 cp_fm_diag_elpa_base 50 14.0 0.379 1.536 0.393 1.573 multiply_cannon_sync_h2d 18288 15.5 1.368 1.570 1.368 1.570 mp_sum_l 7207 12.9 1.228 1.532 1.228 1.532 ot_diis_step 108 11.5 0.011 0.011 1.508 1.508 fft3d_ps 1201 14.6 0.506 0.524 1.410 1.451 fft_wrap_pw1pw2_140 487 13.2 0.110 0.121 1.340 1.359 grid_collocate_task_list 119 9.7 1.202 1.288 1.202 1.288 acc_transpose_blocks 18288 15.5 0.075 0.076 1.225 1.245 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.215 1.221 wfi_extrapolate 11 7.9 0.001 0.001 1.197 1.197 make_images_data 4572 15.5 0.045 0.049 0.861 1.020 qs_energies_init_hamiltonians 11 5.9 0.003 0.005 1.012 1.016 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.930 0.949 hybrid_alltoall_any 4725 16.4 0.055 0.112 0.738 0.920 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.839 0.841 mp_alltoall_d11v 2130 13.8 0.724 0.815 0.724 0.815 acc_transpose_blocks_kernels 18288 16.5 0.211 0.220 0.785 0.798 cp_fm_cholesky_invert 11 10.9 0.792 0.797 0.792 0.797 mp_alltoall_z22v 1201 16.6 0.734 0.796 0.734 0.796 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.643 0.730 mp_waitany 9880 13.7 0.551 0.683 0.551 0.683 rs_pw_transfer_RS2PW_140 130 11.5 0.119 0.122 0.533 0.680 mp_allgather_i34 2286 14.5 0.297 0.675 0.297 0.675 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=33.342000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.000000, yerr=1.595448 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/007813510527d819d1aa75b16315fe3ed2c20304_performance_tests/06/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 114.044384E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3805952 0.0% 0.0% 100.0% average stack size 0.0 0.0 38.6 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 557.121536E+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 0.101 0.126 34.694 34.701 qs_mol_dyn_low 1 2.0 0.003 0.003 34.282 34.302 qs_forces 11 3.9 0.002 0.002 34.224 34.225 qs_energies 11 4.9 0.004 0.007 32.529 32.538 scf_env_do_scf 11 5.9 0.001 0.002 27.387 27.389 scf_env_do_scf_inner_loop 108 6.5 0.003 0.009 23.883 23.884 dbcsr_multiply_generic 2286 12.5 0.101 0.105 17.966 18.093 velocity_verlet 10 3.0 0.001 0.002 17.546 17.548 qs_scf_new_mos 108 7.5 0.001 0.001 15.788 15.849 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.788 15.849 ot_scf_mini 108 9.5 0.003 0.004 14.851 14.903 multiply_cannon 2286 13.5 0.227 0.259 14.301 14.740 multiply_cannon_loop 2286 14.5 0.933 0.968 13.363 13.762 ot_mini 108 10.5 0.001 0.001 9.083 9.154 multiply_cannon_multrec 27432 15.5 2.327 2.987 8.577 8.939 dbcsr_mm_accdrv_process 47916 15.9 5.254 6.983 6.156 7.376 qs_ot_get_derivative 108 11.5 0.001 0.001 7.288 7.344 rebuild_ks_matrix 119 8.3 0.000 0.000 6.432 6.483 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 6.432 6.482 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.721 5.765 sum_up_and_integrate 119 10.3 0.034 0.037 3.716 3.724 init_scf_run 11 5.9 0.000 0.001 3.697 3.697 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.696 3.697 integrate_v_rspace 119 11.3 0.002 0.003 3.682 3.690 init_scf_loop 11 6.9 0.002 0.006 3.485 3.486 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.266 3.295 calculate_rho_elec 119 8.7 0.040 0.046 3.266 3.294 qs_ot_get_p 119 10.4 0.001 0.001 3.201 3.275 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.692 3.140 mp_waitall_1 145218 16.4 2.030 2.676 2.030 2.676 prepare_preconditioner 11 7.9 0.000 0.000 2.603 2.610 make_preconditioner 11 8.9 0.000 0.001 2.602 2.610 make_full_inverse_cholesky 11 9.9 0.000 0.001 2.202 2.538 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.077 2.518 apply_single 119 13.6 0.000 0.000 2.077 2.518 make_m2s 4572 13.5 0.054 0.056 2.242 2.368 calculate_first_density_matrix 1 7.0 0.001 0.002 2.362 2.363 make_images 4572 14.5 0.275 0.334 2.135 2.258 rs_pw_transfer 974 11.9 0.009 0.009 2.020 2.163 calculate_dm_sparse 119 9.5 0.000 0.000 2.031 2.082 qs_ot_p2m_diag 50 11.0 0.015 0.023 1.916 1.925 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.895 1.923 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.912 1.913 grid_integrate_task_list 119 12.3 1.829 1.884 1.829 1.884 density_rs2pw 119 9.7 0.003 0.004 1.749 1.842 jit_kernel_multiply 10 15.9 0.841 1.830 0.841 1.830 pw_transfer 1439 11.6 0.062 0.065 1.738 1.769 ot_diis_step 108 11.5 0.012 0.012 1.756 1.757 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.711 1.723 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 1.649 1.683 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.602 1.603 mp_sum_l 7207 12.9 1.021 1.581 1.021 1.581 potential_pw2rs 119 12.3 0.008 0.009 1.502 1.505 acc_transpose_blocks 27432 15.5 0.112 0.116 1.444 1.472 fft_wrap_pw1pw2_140 487 13.2 0.085 0.092 1.351 1.386 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.364 1.366 fft3d_ps 1201 14.6 0.535 0.587 1.330 1.357 cp_fm_redistribute_end 50 14.0 0.898 1.343 0.900 1.344 grid_collocate_task_list 119 9.7 1.222 1.330 1.222 1.330 cp_fm_diag_elpa_base 50 14.0 0.421 1.283 0.441 1.325 wfi_extrapolate 11 7.9 0.001 0.001 1.288 1.288 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.686 1.228 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.213 1.224 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.189 1.190 cp_fm_upper_to_full 72 13.5 0.827 1.146 0.827 1.146 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.094 1.112 dbcsr_complete_redistribute 329 12.2 0.129 0.151 0.810 1.087 multiply_cannon_sync_h2d 27432 15.5 0.995 1.067 0.995 1.067 make_images_data 4572 15.5 0.045 0.050 0.878 1.037 hybrid_alltoall_any 4725 16.4 0.062 0.152 0.754 0.970 copy_fm_to_dbcsr 176 11.2 0.001 0.002 0.606 0.880 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.793 0.873 mp_alltoall_d11v 2130 13.8 0.735 0.833 0.735 0.833 acc_transpose_blocks_kernels 27432 16.5 0.268 0.275 0.819 0.828 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.802 0.806 cp_fm_cholesky_invert 11 10.9 0.755 0.759 0.755 0.759 mp_alltoall_i22 627 13.8 0.433 0.724 0.433 0.724 qs_env_update_s_mstruct 11 6.9 0.000 0.001 0.685 0.723 mp_alltoall_z22v 1201 16.6 0.677 0.706 0.677 0.706 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=34.701000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=525.090909, yerr=5.035412 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/007813510527d819d1aa75b16315fe3ed2c20304_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 596.230144E+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 0.012 0.030 28.690 28.690 qs_mol_dyn_low 1 2.0 0.003 0.003 28.520 28.529 qs_forces 11 3.9 0.002 0.002 28.462 28.464 qs_energies 11 4.9 0.001 0.001 26.760 26.765 scf_env_do_scf 11 5.9 0.000 0.001 21.665 21.665 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 19.105 19.106 velocity_verlet 10 3.0 0.002 0.003 14.547 14.549 dbcsr_multiply_generic 2286 12.5 0.113 0.116 12.665 12.751 qs_scf_new_mos 108 7.5 0.001 0.001 11.359 11.385 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.358 11.384 ot_scf_mini 108 9.5 0.002 0.002 10.666 10.691 multiply_cannon 2286 13.5 0.231 0.240 9.909 10.350 multiply_cannon_loop 2286 14.5 0.330 0.342 8.954 9.151 multiply_cannon_multrec 9144 15.5 1.661 1.899 6.002 6.221 ot_mini 108 10.5 0.001 0.001 5.960 5.992 rebuild_ks_matrix 119 8.3 0.000 0.000 5.876 5.899 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.875 5.898 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.245 5.266 qs_ot_get_derivative 108 11.5 0.001 0.001 4.642 4.670 dbcsr_mm_accdrv_process 12550 15.8 3.139 4.036 4.239 4.308 init_scf_run 11 5.9 0.000 0.001 3.655 3.655 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.655 3.655 sum_up_and_integrate 119 10.3 0.037 0.040 3.619 3.626 integrate_v_rspace 119 11.3 0.003 0.003 3.582 3.589 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.330 3.338 calculate_rho_elec 119 8.7 0.060 0.061 3.329 3.338 qs_ot_get_p 119 10.4 0.001 0.001 2.906 2.945 init_scf_loop 11 6.9 0.000 0.000 2.539 2.540 calculate_first_density_matrix 1 7.0 0.000 0.000 2.494 2.495 mp_waitall_1 121218 16.5 1.812 2.330 1.812 2.330 make_m2s 4572 13.5 0.034 0.035 1.886 2.044 make_images 4572 14.5 0.268 0.301 1.797 1.953 jit_kernel_multiply 10 15.9 1.062 1.938 1.062 1.938 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.928 1.931 grid_integrate_task_list 119 12.3 1.852 1.919 1.852 1.919 calculate_dm_sparse 119 9.5 0.000 0.000 1.833 1.854 rs_pw_transfer 974 11.9 0.008 0.008 1.761 1.842 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.818 1.819 prepare_preconditioner 11 7.9 0.000 0.000 1.792 1.796 make_preconditioner 11 8.9 0.000 0.000 1.791 1.796 density_rs2pw 119 9.7 0.004 0.004 1.709 1.787 pw_transfer 1439 11.6 0.063 0.065 1.773 1.780 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.718 1.718 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.674 1.704 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.683 1.689 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.637 1.650 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.535 1.543 potential_pw2rs 119 12.3 0.010 0.011 1.398 1.403 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.400 1.400 cp_fm_redistribute_end 50 14.0 0.691 1.371 0.692 1.371 cp_fm_diag_elpa_base 50 14.0 0.634 1.301 0.678 1.357 fft_wrap_pw1pw2_140 487 13.2 0.083 0.087 1.347 1.356 grid_collocate_task_list 119 9.7 1.272 1.355 1.272 1.355 fft3d_ps 1201 14.6 0.540 0.553 1.348 1.355 ot_diis_step 108 11.5 0.012 0.013 1.298 1.299 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.239 1.245 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.220 1.222 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.181 1.194 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.166 1.188 apply_single 119 13.6 0.000 0.000 1.166 1.188 wfi_extrapolate 11 7.9 0.001 0.001 1.112 1.112 hybrid_alltoall_any 4725 16.4 0.062 0.175 0.820 1.075 make_images_data 4572 15.5 0.039 0.042 0.851 1.055 acc_transpose_blocks 9144 15.5 0.039 0.040 0.990 0.998 mp_alltoall_d11v 2130 13.8 0.878 0.991 0.878 0.991 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.868 0.922 cp_fm_cholesky_invert 11 10.9 0.906 0.910 0.906 0.910 multiply_cannon_sync_h2d 9144 15.5 0.704 0.783 0.704 0.783 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.750 0.758 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.753 0.755 acc_transpose_blocks_kernels 9144 16.5 0.117 0.121 0.749 0.753 qs_env_update_s_mstruct 11 6.9 0.001 0.001 0.677 0.729 mp_alltoall_z22v 1201 16.6 0.681 0.714 0.681 0.714 dbcsr_complete_redistribute 329 12.2 0.231 0.275 0.679 0.706 mp_allgather_i34 2286 14.5 0.283 0.705 0.283 0.705 jit_kernel_transpose 5 15.6 0.632 0.634 0.632 0.634 multiply_cannon_metrocomm3 9144 15.5 0.018 0.019 0.327 0.629 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=28.690000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=565.000000, yerr=6.324555 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/007813510527d819d1aa75b16315fe3ed2c20304_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 744.079360E+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 0.015 0.035 40.286 40.287 qs_mol_dyn_low 1 2.0 0.003 0.003 40.039 40.046 qs_forces 11 3.9 0.001 0.002 39.978 39.979 qs_energies 11 4.9 0.002 0.002 38.055 38.058 scf_env_do_scf 11 5.9 0.001 0.001 32.220 32.220 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 24.292 24.294 velocity_verlet 10 3.0 0.002 0.002 22.782 22.788 dbcsr_multiply_generic 2286 12.5 0.107 0.109 17.387 17.626 qs_scf_new_mos 108 7.5 0.001 0.001 15.540 15.637 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.539 15.637 multiply_cannon 2286 13.5 0.299 0.304 13.632 14.593 ot_scf_mini 108 9.5 0.002 0.002 14.469 14.567 multiply_cannon_loop 2286 14.5 0.344 0.351 12.390 13.386 ot_mini 108 10.5 0.001 0.001 8.641 8.755 multiply_cannon_multrec 9144 15.5 3.335 4.664 8.617 8.743 init_scf_loop 11 6.9 0.000 0.000 7.902 7.906 prepare_preconditioner 11 7.9 0.000 0.000 6.991 7.004 make_preconditioner 11 8.9 0.000 0.000 6.991 7.004 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.484 6.882 rebuild_ks_matrix 119 8.3 0.000 0.000 6.739 6.878 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 6.738 6.878 qs_ot_get_derivative 108 11.5 0.001 0.001 6.647 6.746 dbcsr_mm_accdrv_process 12550 15.8 4.095 5.356 5.158 6.489 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.089 6.216 cp_fm_upper_to_full 72 14.2 3.287 4.738 3.287 4.738 init_scf_run 11 5.9 0.000 0.001 3.817 3.817 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.817 3.817 sum_up_and_integrate 119 10.3 0.064 0.065 3.624 3.631 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.617 3.625 calculate_rho_elec 119 8.7 0.118 0.121 3.617 3.625 integrate_v_rspace 119 11.3 0.003 0.003 3.559 3.567 mp_waitall_1 97218 16.6 2.359 3.328 2.359 3.328 qs_ot_get_p 119 10.4 0.001 0.001 3.138 3.272 dbcsr_complete_redistribute 329 12.2 0.302 0.305 2.060 2.943 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.443 2.848 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.750 2.611 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.149 2.430 apply_single 119 13.6 0.000 0.000 2.149 2.429 calculate_first_density_matrix 1 7.0 0.000 0.000 2.421 2.423 make_m2s 4572 13.5 0.037 0.038 2.239 2.387 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.502 2.355 mp_alltoall_i22 627 13.8 1.489 2.340 1.489 2.340 make_images 4572 14.5 0.357 0.390 2.119 2.267 calculate_dm_sparse 119 9.5 0.000 0.000 2.239 2.262 multiply_cannon_metrocomm3 9144 15.5 0.019 0.020 1.222 2.107 grid_integrate_task_list 119 12.3 1.995 2.006 1.995 2.006 pw_transfer 1439 11.6 0.066 0.067 1.997 2.002 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.997 1.998 ot_diis_step 108 11.5 0.014 0.014 1.970 1.971 fft_wrap_pw1pw2 1201 12.6 0.008 0.009 1.902 1.907 qs_ot_p2m_diag 50 11.0 0.043 0.044 1.882 1.884 density_rs2pw 119 9.7 0.003 0.004 1.740 1.755 qs_energies_init_hamiltonians 11 5.9 0.004 0.009 1.751 1.752 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.718 1.748 mp_sum_l 7207 12.9 0.972 1.702 0.972 1.702 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.649 1.699 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.611 1.612 fft_wrap_pw1pw2_140 487 13.2 0.089 0.091 1.592 1.595 jit_kernel_multiply 8 15.6 1.036 1.543 1.036 1.543 fft3d_ps 1201 14.6 0.570 0.580 1.534 1.541 grid_collocate_task_list 119 9.7 1.440 1.457 1.440 1.457 cp_fm_cholesky_invert 11 10.9 1.429 1.432 1.429 1.432 rs_pw_transfer 974 11.9 0.009 0.009 1.349 1.372 wfi_extrapolate 11 7.9 0.001 0.001 1.338 1.339 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.315 1.325 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.321 1.322 cp_fm_diag_elpa_base 50 14.0 1.173 1.229 1.320 1.320 hybrid_alltoall_any 4725 16.4 0.088 0.150 1.097 1.298 make_images_data 4572 15.5 0.043 0.046 1.017 1.213 potential_pw2rs 119 12.3 0.014 0.014 1.188 1.190 mp_alltoall_d11v 2130 13.8 1.127 1.140 1.127 1.140 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.121 1.138 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.081 1.096 multiply_cannon_sync_h2d 9144 15.5 1.040 1.044 1.040 1.044 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.986 1.028 acc_transpose_blocks 9144 15.5 0.038 0.039 1.012 1.015 qs_create_task_list 11 7.9 0.001 0.001 0.940 0.953 generate_qs_task_list 11 8.9 0.372 0.392 0.940 0.953 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.860 0.873 mp_alltoall_z22v 1201 16.6 0.828 0.841 0.828 0.841 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=40.287000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=698.363636, yerr=15.691762 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/007813510527d819d1aa75b16315fe3ed2c20304_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 500.862976E+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 3325326. 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 0.031 0.050 82.709 82.710 qs_mol_dyn_low 1 2.0 0.003 0.004 82.403 82.413 qs_forces 11 3.9 0.002 0.002 82.336 82.337 qs_energies 11 4.9 0.013 0.033 79.539 79.553 scf_env_do_scf 11 5.9 0.000 0.001 70.685 70.687 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 65.214 65.214 dbcsr_multiply_generic 2055 12.4 0.107 0.110 52.684 52.974 qs_scf_new_mos 99 7.5 0.000 0.001 48.548 48.646 qs_scf_loop_do_ot 99 8.5 0.001 0.001 48.548 48.645 ot_scf_mini 99 9.5 0.002 0.002 46.157 46.265 multiply_cannon 2055 13.4 0.177 0.184 42.902 43.840 multiply_cannon_loop 2055 14.4 1.565 1.607 41.939 42.895 velocity_verlet 10 3.0 0.001 0.002 41.517 41.518 ot_mini 99 10.5 0.001 0.001 28.236 28.345 qs_ot_get_derivative 99 11.5 0.001 0.001 21.425 21.507 multiply_cannon_multrec 49320 15.4 12.028 12.699 17.058 17.916 rebuild_ks_matrix 110 8.3 0.000 0.001 14.255 14.367 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.254 14.366 mp_waitall_1 241148 16.1 12.178 13.374 12.178 13.374 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.526 12.622 multiply_cannon_sync_h2d 49320 15.4 9.885 10.518 9.885 10.518 qs_ot_get_p 110 10.4 0.001 0.001 9.392 9.511 multiply_cannon_metrocomm3 49320 15.4 0.076 0.080 7.322 8.713 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.072 7.599 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.145 7.580 apply_single 110 13.6 0.000 0.001 7.145 7.580 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 7.246 7.313 sum_up_and_integrate 110 10.3 0.037 0.042 6.886 6.898 integrate_v_rspace 110 11.3 0.003 0.003 6.850 6.870 init_scf_run 11 5.9 0.000 0.001 6.711 6.711 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.710 6.711 ot_diis_step 99 11.5 0.006 0.006 6.519 6.520 mp_sum_l 6514 12.8 5.520 6.340 5.520 6.340 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.090 6.250 calculate_rho_elec 110 8.6 0.020 0.024 6.089 6.249 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.100 6.149 init_scf_loop 11 6.9 0.000 0.000 5.444 5.444 dbcsr_mm_accdrv_process 87628 16.1 2.084 2.184 4.909 5.208 cp_dbcsr_syevd 48 12.0 0.002 0.002 5.047 5.047 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.579 4.580 cp_fm_redistribute_end 48 14.0 3.985 4.555 3.989 4.556 cp_fm_diag_elpa_base 48 14.0 0.561 4.493 0.564 4.518 multiply_cannon_metrocomm1 49320 15.4 0.060 0.063 2.855 4.044 wfi_extrapolate 11 7.9 0.001 0.001 3.972 3.972 make_m2s 4110 13.4 0.060 0.065 3.802 3.919 calculate_dm_sparse 110 9.5 0.000 0.001 3.753 3.887 rs_pw_transfer 902 11.9 0.012 0.013 3.647 3.867 make_images 4110 14.4 0.177 0.191 3.708 3.827 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.544 3.548 grid_integrate_task_list 110 12.3 3.255 3.378 3.255 3.378 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.297 3.348 prepare_preconditioner 11 7.9 0.000 0.000 3.301 3.320 make_preconditioner 11 8.9 0.000 0.000 3.301 3.320 density_rs2pw 110 9.6 0.004 0.004 3.154 3.319 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.193 3.245 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.074 3.111 pw_transfer 1331 11.6 0.053 0.062 2.736 2.794 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 2.649 2.711 calculate_first_density_matrix 1 7.0 0.000 0.000 2.653 2.659 jit_kernel_multiply 13 15.9 2.539 2.564 2.539 2.564 potential_pw2rs 110 12.3 0.006 0.007 2.499 2.522 mp_alltoall_d11v 2046 13.8 2.098 2.477 2.098 2.477 fft_wrap_pw1pw2_140 451 13.1 0.169 0.189 2.233 2.296 acc_transpose_blocks 49320 15.4 0.216 0.225 2.179 2.260 fft3d_ps 1111 14.6 0.751 0.835 2.186 2.236 grid_collocate_task_list 110 9.6 2.087 2.195 2.087 2.195 mp_waitany 14300 13.8 1.811 2.125 1.811 2.125 mp_sum_d 3883 11.9 1.461 1.979 1.461 1.979 make_images_data 4110 15.4 0.043 0.046 1.752 1.884 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.866 1.883 hybrid_alltoall_any 4261 16.3 0.082 0.475 1.527 1.809 cp_fm_cholesky_invert 11 10.9 1.802 1.805 1.802 1.805 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.643 1.666 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=82.710000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=475.909091, yerr=2.234219 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/007813510527d819d1aa75b16315fe3ed2c20304_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 586.194944E+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 2075986. 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 0.012 0.030 67.629 67.630 qs_mol_dyn_low 1 2.0 0.003 0.003 67.421 67.431 qs_forces 11 3.9 0.002 0.002 67.341 67.342 qs_energies 11 4.9 0.001 0.002 64.057 64.062 scf_env_do_scf 11 5.9 0.000 0.001 55.695 55.698 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 48.132 48.133 dbcsr_multiply_generic 2055 12.4 0.114 0.117 37.601 37.753 velocity_verlet 10 3.0 0.001 0.002 35.718 35.720 qs_scf_new_mos 99 7.5 0.001 0.001 32.527 32.645 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.526 32.645 multiply_cannon 2055 13.4 0.222 0.245 30.969 32.097 ot_scf_mini 99 9.5 0.003 0.003 30.867 30.992 multiply_cannon_loop 2055 14.4 0.929 0.949 29.688 30.696 ot_mini 99 10.5 0.001 0.001 18.247 18.386 multiply_cannon_multrec 24660 15.4 7.634 9.197 13.841 15.649 rebuild_ks_matrix 110 8.3 0.000 0.000 13.453 13.573 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 13.453 13.573 qs_ot_get_derivative 99 11.5 0.001 0.001 12.411 12.545 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.865 11.974 mp_waitall_1 186928 16.3 7.751 10.332 7.751 10.332 multiply_cannon_sync_h2d 24660 15.4 7.025 8.224 7.025 8.224 multiply_cannon_metrocomm3 24660 15.4 0.071 0.073 5.160 7.819 init_scf_loop 11 6.9 0.000 0.000 7.527 7.527 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.565 7.305 apply_single 110 13.6 0.000 0.001 6.565 7.305 sum_up_and_integrate 110 10.3 0.052 0.058 6.401 6.412 integrate_v_rspace 110 11.3 0.002 0.003 6.349 6.361 dbcsr_mm_accdrv_process 52282 16.1 4.560 5.285 6.046 6.317 qs_ot_get_p 110 10.4 0.001 0.001 6.004 6.160 init_scf_run 11 5.9 0.000 0.001 6.013 6.014 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.013 6.013 ot_diis_step 99 11.5 0.010 0.010 5.787 5.787 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.646 5.655 calculate_rho_elec 110 8.6 0.039 0.047 5.645 5.655 prepare_preconditioner 11 7.9 0.000 0.000 5.529 5.550 make_preconditioner 11 8.9 0.000 0.000 5.528 5.550 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.646 5.335 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.116 5.272 make_m2s 4110 13.4 0.056 0.060 4.167 4.646 make_images 4110 14.4 0.400 0.446 4.059 4.533 qs_ot_p2m_diag 48 11.0 0.028 0.044 4.125 4.145 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.689 3.689 wfi_extrapolate 11 7.9 0.001 0.001 3.480 3.480 pw_transfer 1331 11.6 0.065 0.070 3.248 3.382 grid_integrate_task_list 110 12.3 3.161 3.321 3.161 3.321 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.142 3.274 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.191 3.257 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.229 3.230 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.143 3.145 cp_fm_redistribute_end 48 14.0 2.342 3.117 2.344 3.119 density_rs2pw 110 9.6 0.004 0.004 2.960 3.097 cp_fm_diag_elpa_base 48 14.0 0.740 3.002 0.771 3.088 calculate_dm_sparse 110 9.5 0.001 0.001 2.955 2.984 rs_pw_transfer 902 11.9 0.012 0.013 2.805 2.974 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.885 2.937 make_images_data 4110 15.4 0.047 0.051 2.353 2.838 hybrid_alltoall_any 4261 16.3 0.102 0.443 2.036 2.802 fft_wrap_pw1pw2_140 451 13.1 0.201 0.218 2.649 2.781 fft3d_ps 1111 14.6 1.074 1.267 2.498 2.617 cp_fm_cholesky_invert 11 10.9 2.575 2.583 2.575 2.583 mp_sum_l 6514 12.8 1.837 2.494 1.837 2.494 jit_kernel_multiply 13 16.3 1.133 2.476 1.133 2.476 calculate_first_density_matrix 1 7.0 0.000 0.000 2.449 2.452 potential_pw2rs 110 12.3 0.008 0.009 2.292 2.302 grid_collocate_task_list 110 9.6 2.120 2.271 2.120 2.271 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.954 1.975 mp_alltoall_d11v 2046 13.8 1.707 1.877 1.707 1.877 multiply_cannon_metrocomm4 22605 15.4 0.075 0.078 0.786 1.848 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.834 1.835 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.751 1.765 mp_irecv_dv 57340 16.2 0.661 1.628 0.661 1.628 mp_allgather_i34 2055 14.4 0.610 1.609 0.610 1.609 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.527 1.537 acc_transpose_blocks 24660 15.4 0.107 0.110 1.509 1.525 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.371 1.483 dbcsr_complete_redistribute 325 12.2 0.242 0.308 1.103 1.369 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=67.630000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=554.545455, yerr=7.139692 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/007813510527d819d1aa75b16315fe3ed2c20304_performance_tests/11/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 404.681598E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 3346752 0.0% 0.0% 100.0% average stack size 0.0 0.0 294.1 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 660.967424E+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 0.027 0.059 59.290 59.291 qs_mol_dyn_low 1 2.0 0.003 0.004 58.955 58.964 qs_forces 11 3.9 0.002 0.004 58.890 58.890 qs_energies 11 4.9 0.007 0.049 55.772 55.776 scf_env_do_scf 11 5.9 0.001 0.001 47.750 47.750 scf_env_do_scf_inner_loop 99 6.5 0.002 0.008 39.209 39.209 velocity_verlet 10 3.0 0.001 0.001 32.039 32.040 dbcsr_multiply_generic 2055 12.4 0.108 0.112 28.209 28.462 qs_scf_new_mos 99 7.5 0.001 0.001 24.825 24.932 qs_scf_loop_do_ot 99 8.5 0.001 0.001 24.824 24.932 ot_scf_mini 99 9.5 0.002 0.003 23.604 23.722 multiply_cannon 2055 13.4 0.211 0.220 21.955 23.135 multiply_cannon_loop 2055 14.4 0.619 0.631 20.830 21.770 ot_mini 99 10.5 0.001 0.001 13.455 13.577 rebuild_ks_matrix 110 8.3 0.000 0.000 12.040 12.178 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.040 12.178 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.634 10.759 multiply_cannon_multrec 16440 15.4 3.900 4.990 9.635 10.627 mp_waitall_1 146766 16.3 7.102 10.451 7.102 10.451 qs_ot_get_derivative 99 11.5 0.001 0.001 9.007 9.129 init_scf_loop 11 6.9 0.000 0.002 8.507 8.507 multiply_cannon_metrocomm3 16440 15.4 0.043 0.044 4.298 7.498 prepare_preconditioner 11 7.9 0.000 0.000 6.770 6.788 make_preconditioner 11 8.9 0.000 0.002 6.770 6.788 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.065 6.409 sum_up_and_integrate 110 10.3 0.060 0.061 6.333 6.351 integrate_v_rspace 110 11.3 0.002 0.003 6.274 6.292 dbcsr_mm_accdrv_process 34862 16.1 4.623 5.339 5.589 5.748 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.508 5.519 calculate_rho_elec 110 8.6 0.058 0.058 5.507 5.518 qs_ot_get_p 110 10.4 0.001 0.001 5.363 5.516 init_scf_run 11 5.9 0.000 0.001 5.507 5.507 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.506 5.507 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.970 5.415 apply_single 110 13.6 0.000 0.000 4.969 5.415 make_m2s 4110 13.4 0.050 0.052 4.156 4.473 ot_diis_step 99 11.5 0.010 0.011 4.420 4.420 make_images 4110 14.4 0.393 0.507 4.041 4.359 multiply_cannon_sync_h2d 16440 15.4 3.694 4.327 3.694 4.327 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.767 3.772 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.060 3.700 grid_integrate_task_list 110 12.3 3.203 3.476 3.203 3.476 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.430 3.431 pw_transfer 1331 11.6 0.064 0.071 3.170 3.176 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.064 3.071 density_rs2pw 110 9.6 0.004 0.004 2.798 2.961 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.921 2.923 wfi_extrapolate 11 7.9 0.001 0.001 2.903 2.903 cp_fm_redistribute_end 48 14.0 1.813 2.896 1.815 2.896 cp_fm_diag_elpa_base 48 14.0 1.019 2.765 1.076 2.874 make_images_data 4110 15.4 0.044 0.048 2.433 2.847 rs_pw_transfer 902 11.9 0.010 0.011 2.628 2.789 hybrid_alltoall_any 4261 16.3 0.105 0.375 2.137 2.757 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.721 2.722 cp_fm_cholesky_invert 11 10.9 2.657 2.662 2.657 2.662 fft_wrap_pw1pw2_140 451 13.1 0.211 0.215 2.623 2.632 calculate_dm_sparse 110 9.5 0.001 0.001 2.522 2.549 calculate_first_density_matrix 1 7.0 0.001 0.004 2.523 2.524 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.406 2.457 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.363 2.434 multiply_cannon_metrocomm4 14385 15.4 0.045 0.049 0.886 2.406 fft3d_ps 1111 14.6 1.066 1.079 2.378 2.390 grid_collocate_task_list 110 9.6 2.171 2.360 2.171 2.360 mp_irecv_dv 48980 15.7 0.816 2.281 0.816 2.281 potential_pw2rs 110 12.3 0.011 0.011 2.174 2.184 mp_alltoall_d11v 2046 13.8 1.759 2.094 1.759 2.094 mp_sum_l 6514 12.8 1.496 2.071 1.496 2.071 qs_energies_init_hamiltonians 11 5.9 0.001 0.005 1.959 1.959 dbcsr_complete_redistribute 325 12.2 0.330 0.374 1.475 1.931 cp_fm_upper_to_full 70 13.6 1.367 1.813 1.367 1.813 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.606 1.620 cp_fm_cholesky_decompose 22 10.9 1.575 1.592 1.575 1.592 mp_allgather_i34 2055 14.4 0.450 1.551 0.450 1.551 jit_kernel_multiply 8 16.6 0.576 1.505 0.576 1.505 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.484 1.494 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.996 1.444 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.354 1.437 mp_waitany 17072 13.8 1.141 1.293 1.141 1.293 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.260 1.267 acc_transpose_blocks 16440 15.4 0.073 0.075 1.211 1.237 rs_gather_matrices 110 12.3 0.137 0.150 0.848 1.189 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=59.291000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=624.818182, yerr=8.144359 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/007813510527d819d1aa75b16315fe3ed2c20304_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 732.655616E+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 0.014 0.030 64.703 64.703 qs_mol_dyn_low 1 2.0 0.003 0.004 64.382 64.392 qs_forces 11 3.9 0.002 0.002 64.318 64.318 qs_energies 11 4.9 0.001 0.001 60.961 60.964 scf_env_do_scf 11 5.9 0.000 0.001 52.623 52.626 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 40.673 40.674 velocity_verlet 10 3.0 0.002 0.002 36.840 36.842 dbcsr_multiply_generic 2055 12.4 0.116 0.118 29.579 29.786 qs_scf_new_mos 99 7.5 0.001 0.001 26.339 26.431 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.338 26.430 ot_scf_mini 99 9.5 0.003 0.003 24.716 24.824 multiply_cannon 2055 13.4 0.243 0.265 22.645 23.889 multiply_cannon_loop 2055 14.4 0.879 0.905 21.095 21.704 ot_mini 99 10.5 0.001 0.001 14.081 14.187 multiply_cannon_multrec 24660 15.4 4.191 6.597 12.714 13.830 rebuild_ks_matrix 110 8.3 0.000 0.000 11.891 12.021 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.890 12.021 init_scf_loop 11 6.9 0.000 0.000 11.904 11.904 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.532 10.645 prepare_preconditioner 11 7.9 0.000 0.000 10.177 10.190 make_preconditioner 11 8.9 0.000 0.000 10.177 10.190 qs_ot_get_derivative 99 11.5 0.001 0.001 9.928 10.037 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.342 9.856 dbcsr_mm_accdrv_process 52304 16.0 6.775 8.251 8.377 9.261 mp_waitall_1 126806 16.4 4.380 6.345 4.380 6.345 sum_up_and_integrate 110 10.3 0.067 0.070 6.309 6.322 integrate_v_rspace 110 11.3 0.003 0.003 6.241 6.253 make_m2s 4110 13.4 0.059 0.061 5.327 5.759 qs_ot_get_p 110 10.4 0.001 0.001 5.623 5.756 init_scf_run 11 5.9 0.000 0.001 5.617 5.617 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.616 5.617 make_images 4110 14.4 0.576 0.700 5.187 5.615 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.569 5.578 calculate_rho_elec 110 8.6 0.077 0.081 5.568 5.577 cp_fm_upper_to_full 70 13.8 3.373 4.810 3.373 4.810 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.024 4.114 apply_single 110 13.6 0.000 0.000 4.024 4.114 ot_diis_step 99 11.5 0.011 0.011 4.112 4.113 dbcsr_complete_redistribute 325 12.2 0.459 0.518 2.781 3.946 qs_ot_p2m_diag 48 11.0 0.055 0.063 3.818 3.833 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.244 3.398 grid_integrate_task_list 110 12.3 3.270 3.377 3.270 3.377 multiply_cannon_sync_h2d 24660 15.4 3.179 3.373 3.179 3.373 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.302 3.349 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.336 3.337 multiply_cannon_metrocomm3 24660 15.4 0.036 0.037 1.464 3.303 pw_transfer 1331 11.6 0.064 0.070 3.229 3.254 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.124 3.152 make_images_data 4110 15.4 0.046 0.050 2.665 3.099 hybrid_alltoall_any 4261 16.3 0.120 0.457 2.317 3.086 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.026 3.085 calculate_dm_sparse 110 9.5 0.001 0.001 2.967 2.995 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.827 2.969 wfi_extrapolate 11 7.9 0.001 0.001 2.958 2.958 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.926 2.928 mp_alltoall_i22 605 13.7 1.701 2.908 1.701 2.908 density_rs2pw 110 9.6 0.004 0.004 2.740 2.890 cp_fm_cholesky_invert 11 10.9 2.813 2.821 2.813 2.821 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.779 2.781 cp_fm_redistribute_end 48 14.0 1.382 2.752 1.384 2.753 cp_fm_diag_elpa_base 48 14.0 1.286 2.624 1.366 2.734 fft_wrap_pw1pw2_140 451 13.1 0.203 0.216 2.636 2.665 calculate_first_density_matrix 1 7.0 0.000 0.000 2.573 2.575 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.522 2.565 rs_pw_transfer 902 11.9 0.010 0.010 2.381 2.553 fft3d_ps 1111 14.6 1.063 1.091 2.426 2.441 grid_collocate_task_list 110 9.6 2.220 2.375 2.220 2.375 jit_kernel_multiply 12 15.6 1.270 2.308 1.270 2.308 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.219 2.219 mp_alltoall_d11v 2046 13.8 1.811 2.179 1.811 2.179 potential_pw2rs 110 12.3 0.012 0.013 2.084 2.099 cp_fm_cholesky_decompose 22 10.9 1.704 1.751 1.704 1.751 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.699 1.731 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.599 1.704 mp_allgather_i34 2055 14.4 0.648 1.666 0.648 1.666 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.604 1.618 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.583 1.597 acc_transpose_blocks 24660 15.4 0.104 0.106 1.550 1.574 mp_sum_l 6514 12.8 0.923 1.512 0.923 1.512 multiply_cannon_metrocomm4 20550 15.4 0.057 0.061 0.857 1.397 mp_irecv_dv 62702 16.1 0.759 1.324 0.759 1.324 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=64.703000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=693.000000, yerr=11.497035 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/007813510527d819d1aa75b16315fe3ed2c20304_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 858.394624E+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 0.022 0.049 54.725 54.725 qs_mol_dyn_low 1 2.0 0.003 0.003 54.450 54.486 qs_forces 11 3.9 0.002 0.002 54.384 54.385 qs_energies 11 4.9 0.001 0.001 50.786 50.791 scf_env_do_scf 11 5.9 0.000 0.001 42.490 42.490 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 34.672 34.673 velocity_verlet 10 3.0 0.002 0.002 30.563 30.566 dbcsr_multiply_generic 2055 12.4 0.105 0.106 22.927 23.069 qs_scf_new_mos 99 7.5 0.001 0.001 20.651 20.701 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.650 20.701 ot_scf_mini 99 9.5 0.002 0.002 19.419 19.458 multiply_cannon 2055 13.4 0.245 0.261 17.504 18.761 multiply_cannon_loop 2055 14.4 0.321 0.334 16.169 16.497 rebuild_ks_matrix 110 8.3 0.000 0.000 11.388 11.408 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.388 11.407 ot_mini 99 10.5 0.001 0.001 10.495 10.529 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.139 10.156 multiply_cannon_multrec 8220 15.4 3.293 4.679 7.677 8.642 init_scf_loop 11 6.9 0.000 0.000 7.770 7.772 mp_waitall_1 106626 16.5 6.081 7.702 6.081 7.702 qs_ot_get_derivative 99 11.5 0.001 0.001 6.718 6.757 sum_up_and_integrate 110 10.3 0.080 0.081 6.175 6.187 prepare_preconditioner 11 7.9 0.000 0.000 6.169 6.172 make_preconditioner 11 8.9 0.000 0.000 6.169 6.172 integrate_v_rspace 110 11.3 0.003 0.003 6.095 6.106 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.757 5.828 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.572 5.585 calculate_rho_elec 110 8.6 0.115 0.115 5.571 5.585 dbcsr_mm_accdrv_process 17442 15.9 2.990 4.674 4.254 5.323 init_scf_run 11 5.9 0.000 0.001 5.166 5.166 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.166 5.166 qs_ot_get_p 110 10.4 0.001 0.001 5.032 5.077 make_m2s 4110 13.4 0.038 0.040 4.151 4.411 multiply_cannon_metrocomm3 8220 15.4 0.017 0.017 2.990 4.369 make_images 4110 14.4 0.635 0.689 4.021 4.281 ot_diis_step 99 11.5 0.012 0.012 3.757 3.757 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.673 3.731 apply_single 110 13.6 0.000 0.000 3.673 3.731 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.651 3.656 grid_integrate_task_list 110 12.3 3.376 3.446 3.376 3.446 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.344 3.345 pw_transfer 1331 11.6 0.064 0.069 3.202 3.215 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.095 3.111 multiply_cannon_sync_h2d 8220 15.4 2.917 3.086 2.917 3.086 cp_fm_cholesky_invert 11 10.9 3.036 3.039 3.036 3.039 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.813 2.814 cp_fm_redistribute_end 48 14.0 0.710 2.784 0.714 2.784 make_images_data 4110 15.4 0.038 0.043 2.348 2.768 cp_fm_diag_elpa_base 48 14.0 1.885 2.594 2.062 2.753 density_rs2pw 110 9.6 0.004 0.004 2.647 2.744 wfi_extrapolate 11 7.9 0.001 0.001 2.713 2.713 hybrid_alltoall_any 4261 16.3 0.200 0.858 2.274 2.666 fft_wrap_pw1pw2_140 451 13.1 0.212 0.215 2.643 2.663 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.645 2.647 calculate_dm_sparse 110 9.5 0.001 0.001 2.559 2.595 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.547 2.548 grid_collocate_task_list 110 9.6 2.328 2.474 2.328 2.474 fft3d_ps 1111 14.6 1.116 1.146 2.352 2.376 calculate_first_density_matrix 1 7.0 0.000 0.000 2.366 2.367 rs_pw_transfer 902 11.9 0.010 0.010 2.213 2.331 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.188 2.198 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.765 1.986 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.963 1.976 potential_pw2rs 110 12.3 0.015 0.016 1.942 1.950 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.807 1.833 mp_alltoall_d11v 2046 13.8 1.590 1.752 1.590 1.752 cp_fm_cholesky_decompose 22 10.9 1.686 1.700 1.686 1.700 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.489 1.605 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.581 1.586 jit_kernel_multiply 8 15.8 0.956 1.579 0.956 1.579 mp_allgather_i34 2055 14.4 0.518 1.553 0.518 1.553 dbcsr_complete_redistribute 325 12.2 0.564 0.589 1.452 1.548 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.534 1.545 qs_create_task_list 11 7.9 0.000 0.001 1.221 1.319 generate_qs_task_list 11 8.9 0.378 0.446 1.221 1.319 mp_waitany 9240 13.8 1.063 1.187 1.063 1.187 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.144 1.169 multiply_cannon_metrocomm1 8220 15.4 0.021 0.022 0.751 1.169 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=54.725000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=806.818182, yerr=12.925086 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/007813510527d819d1aa75b16315fe3ed2c20304_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 1.378763E+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 0.040 0.061 85.391 85.392 qs_mol_dyn_low 1 2.0 0.003 0.004 85.022 85.039 qs_forces 11 3.9 0.002 0.002 84.955 84.957 qs_energies 11 4.9 0.001 0.001 80.926 80.929 scf_env_do_scf 11 5.9 0.001 0.001 71.085 71.085 velocity_verlet 10 3.0 0.002 0.002 55.051 55.057 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 42.099 42.101 init_scf_loop 11 6.9 0.000 0.000 28.913 28.916 dbcsr_multiply_generic 2055 12.4 0.120 0.122 28.335 28.414 prepare_preconditioner 11 7.9 0.000 0.000 26.934 26.940 make_preconditioner 11 8.9 0.000 0.000 26.934 26.940 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.871 26.412 qs_scf_new_mos 99 7.5 0.001 0.001 25.866 25.910 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.865 25.909 ot_scf_mini 99 9.5 0.002 0.002 24.104 24.137 multiply_cannon 2055 13.4 0.346 0.381 21.498 22.252 multiply_cannon_loop 2055 14.4 0.341 0.343 19.730 20.012 cp_fm_upper_to_full 70 14.2 13.076 18.865 13.076 18.865 ot_mini 99 10.5 0.001 0.001 13.379 13.413 rebuild_ks_matrix 110 8.3 0.000 0.001 13.028 13.072 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 13.027 13.072 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.846 11.886 dbcsr_complete_redistribute 325 12.2 1.019 1.060 7.625 11.017 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.611 9.990 multiply_cannon_multrec 8220 15.4 4.364 4.583 9.596 9.678 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.047 9.388 mp_alltoall_i22 605 13.7 5.668 9.065 5.668 9.065 qs_ot_get_derivative 99 11.5 0.001 0.001 8.839 8.875 mp_waitall_1 87304 16.6 7.644 8.535 7.644 8.535 sum_up_and_integrate 110 10.3 0.150 0.151 6.562 6.575 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.413 6.452 calculate_rho_elec 110 8.6 0.227 0.227 6.412 6.451 integrate_v_rspace 110 11.3 0.004 0.004 6.412 6.426 make_m2s 4110 13.4 0.043 0.043 5.142 5.634 qs_ot_get_p 110 10.4 0.001 0.001 5.594 5.633 init_scf_run 11 5.9 0.000 0.001 5.627 5.627 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.626 5.627 make_images 4110 14.4 0.880 0.934 4.955 5.446 dbcsr_mm_accdrv_process 11614 15.7 3.309 3.615 5.089 5.357 cp_fm_cholesky_invert 11 10.9 5.311 5.316 5.311 5.316 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.643 5.111 apply_single 110 13.6 0.000 0.000 4.643 5.111 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 4.755 5.073 ot_diis_step 99 11.5 0.015 0.016 4.519 4.519 qs_ot_p2m_diag 48 11.0 0.151 0.155 4.018 4.024 multiply_cannon_sync_h2d 8220 15.4 3.948 3.965 3.948 3.965 grid_integrate_task_list 110 12.3 3.653 3.710 3.653 3.710 pw_transfer 1331 11.6 0.074 0.074 3.703 3.707 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.625 3.627 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 3.587 3.591 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.572 3.573 hybrid_alltoall_any 4261 16.3 0.257 0.555 2.846 3.506 make_images_data 4110 15.4 0.041 0.044 2.781 3.424 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.826 3.272 calculate_dm_sparse 110 9.5 0.001 0.001 3.219 3.245 wfi_extrapolate 11 7.9 0.001 0.001 3.212 3.212 fft_wrap_pw1pw2_140 451 13.1 0.216 0.219 3.116 3.120 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.980 2.981 cp_fm_diag_elpa_base 48 14.0 2.438 2.642 2.979 2.979 density_rs2pw 110 9.6 0.004 0.004 2.909 2.924 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.830 2.834 fft3d_ps 1111 14.6 1.264 1.279 2.795 2.800 grid_collocate_task_list 110 9.6 2.624 2.649 2.624 2.649 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.350 2.369 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.328 2.345 calculate_first_density_matrix 1 7.0 0.000 0.000 2.311 2.312 qs_env_update_s_mstruct 11 6.9 0.001 0.001 2.161 2.228 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.097 2.180 rs_pw_transfer 902 11.9 0.011 0.011 2.070 2.115 mp_alltoall_d11v 2046 13.8 2.004 2.050 2.004 2.050 potential_pw2rs 110 12.3 0.021 0.022 2.021 2.024 cp_fm_cholesky_decompose 22 10.9 1.997 2.019 1.997 2.019 qs_create_task_list 11 7.9 0.000 0.001 1.898 1.942 generate_qs_task_list 11 8.9 0.737 0.792 1.897 1.941 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.784 1.789 jit_kernel_multiply 10 15.3 1.578 1.780 1.578 1.780 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.751 1.765 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.731 1.762 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=85.392000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1227.727273, yerr=67.013012 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/007813510527d819d1aa75b16315fe3ed2c20304_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430458527744 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613083000832 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228655E+12 0.0% 0.0% 100.0% flops max/rank 1.103326E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806314816 0.0% 0.0% 100.0% number of processed stacks 12044928 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.1 marketing flops 145.647559E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 628.891648E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10431456 MPI messages size (bytes): total size 4.526012E+12 min size 0.000000E+00 max size 4.537280E+06 average size 433.881094E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3605352 96403587072 32768 < size <= 131072 1305088 74666999808 131072 < size <= 4194304 5190114 3200148065288 4194304 < size <= 16777216 263934 1154745839544 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4113 56379. MP_Allreduce 11323 782. MP_Sync 170 MP_Alltoall 2242 1902552. MP_ISendRecv 49020 18752. MP_Wait 67312 MP_comm_split 83 MP_ISend 16140 108019. MP_IRecv 16140 108019. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.029 202.776 202.778 qs_mol_dyn_low 1 2.0 0.004 0.006 202.418 202.431 qs_forces 11 3.9 0.003 0.005 202.330 202.332 qs_energies 11 4.9 0.001 0.002 196.827 196.840 scf_env_do_scf 11 5.9 0.001 0.001 180.551 180.555 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 159.818 159.820 dbcsr_multiply_generic 2527 12.6 0.179 0.183 124.250 125.151 velocity_verlet 10 3.0 0.001 0.002 122.960 122.961 qs_scf_new_mos 118 7.6 0.001 0.001 120.935 121.244 qs_scf_loop_do_ot 118 8.6 0.001 0.001 120.934 121.243 ot_scf_mini 118 9.6 0.003 0.004 114.346 114.631 multiply_cannon 2527 13.6 0.239 0.247 101.027 102.988 multiply_cannon_loop 2527 14.6 2.127 2.196 98.887 100.532 ot_mini 118 10.6 0.001 0.002 65.838 66.142 multiply_cannon_multrec 60648 15.6 33.373 35.409 41.771 43.727 qs_ot_get_derivative 118 11.6 0.001 0.003 40.920 41.213 rebuild_ks_matrix 129 8.3 0.001 0.001 33.361 33.731 qs_ks_build_kohn_sham_matrix 129 9.3 0.020 0.053 33.360 33.731 mp_waitall_1 293764 16.2 28.470 30.920 28.470 30.920 qs_ks_update_qs_env 129 7.6 0.001 0.001 30.028 30.366 multiply_cannon_sync_h2d 60648 15.6 27.580 29.733 27.580 29.733 qs_ot_get_p 129 10.4 0.001 0.001 26.590 26.909 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 24.358 25.107 apply_single 129 13.6 0.001 0.001 24.358 25.107 ot_diis_step 118 11.6 0.008 0.012 24.609 24.611 init_scf_loop 11 6.9 0.000 0.000 20.659 20.660 qs_ot_p2m_diag 83 11.4 0.077 0.091 19.893 19.964 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 18.854 19.045 multiply_cannon_metrocomm3 60648 15.6 0.111 0.117 15.474 17.621 cp_dbcsr_syevd 83 12.4 0.004 0.005 17.448 17.449 prepare_preconditioner 11 7.9 0.000 0.000 16.134 16.180 make_preconditioner 11 8.9 0.000 0.000 16.134 16.180 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.376 15.547 cp_fm_diag_elpa 83 13.4 0.000 0.001 14.627 14.632 cp_fm_redistribute_end 83 14.4 11.528 14.556 11.539 14.559 cp_fm_diag_elpa_base 83 14.4 2.977 14.330 3.008 14.450 make_m2s 5054 13.6 0.104 0.111 13.917 14.245 make_images 5054 14.6 0.405 0.425 13.735 14.079 sum_up_and_integrate 129 10.3 0.090 0.108 13.977 13.994 integrate_v_rspace 129 11.3 0.003 0.004 13.886 13.908 qs_rho_update_rho_low 129 7.7 0.001 0.002 12.685 12.772 calculate_rho_elec 129 8.7 0.045 0.064 12.685 12.771 init_scf_run 11 5.9 0.000 0.001 12.175 12.176 scf_env_initial_rho_setup 11 6.9 0.001 0.002 12.175 12.176 mp_sum_l 7930 13.1 8.029 9.312 8.029 9.312 cp_fm_cholesky_invert 11 10.9 9.293 9.300 9.293 9.300 wfi_extrapolate 11 7.9 0.001 0.004 9.020 9.020 dbcsr_mm_accdrv_process 125468 16.2 3.337 3.490 7.949 8.490 calculate_dm_sparse 129 9.5 0.001 0.001 8.378 8.479 multiply_cannon_metrocomm1 60648 15.6 0.087 0.092 5.973 8.361 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 8.168 8.312 qs_ot_get_orbitals 118 10.6 0.001 0.001 7.763 7.849 make_images_data 5054 15.6 0.070 0.076 6.849 7.756 grid_integrate_task_list 129 12.3 7.187 7.643 7.187 7.643 hybrid_alltoall_any 5240 16.5 0.293 2.268 5.989 7.188 pw_transfer 1559 11.6 0.074 0.106 6.787 7.064 density_rs2pw 129 9.7 0.006 0.007 6.483 6.989 fft_wrap_pw1pw2 1301 12.7 0.010 0.013 6.582 6.837 cp_dbcsr_sm_fm_multiply 37 9.5 0.003 0.003 6.617 6.625 rs_pw_transfer 1054 12.0 0.017 0.020 5.700 6.175 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.778 5.871 fft_wrap_pw1pw2_140 527 13.2 0.449 0.512 5.659 5.846 fft3d_ps 1301 14.7 2.107 2.592 5.363 5.553 grid_collocate_task_list 129 9.7 4.758 5.242 4.758 5.242 mp_alltoall_d11v 2423 14.1 4.037 4.970 4.037 4.970 cp_fm_cholesky_decompose 22 10.9 4.654 4.666 4.654 4.666 potential_pw2rs 129 12.3 0.009 0.010 4.459 4.481 mp_sum_d 4496 12.2 3.554 4.381 3.554 4.381 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=202.778000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=594.181818, yerr=7.068730 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/007813510527d819d1aa75b16315fe3ed2c20304_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 2.199914E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 6022464 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.2 marketing flops 145.647559E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 830.062592E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2425920 MPI messages size (bytes): total size 4.132350E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703416E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 71436 2336489472 32768 < size <= 131072 728832 55956209664 131072 < size <= 4194304 1386864 1409906900992 4194304 < size <= 16777216 155760 1473826772352 16777216 < size 68112 1190343475200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4113 56823. MP_Allreduce 11322 944. MP_Sync 170 MP_Alltoall 1983 6038489. MP_ISendRecv 24252 47072. MP_Wait 38240 MP_comm_split 83 MP_ISend 11836 212447. MP_IRecv 11836 212447. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.022 0.049 189.441 189.442 qs_mol_dyn_low 1 2.0 0.003 0.005 189.026 189.041 qs_forces 11 3.9 0.004 0.014 188.907 188.913 qs_energies 11 4.9 0.001 0.003 182.193 182.243 scf_env_do_scf 11 5.9 0.001 0.001 165.987 165.996 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 133.131 133.133 velocity_verlet 10 3.0 0.001 0.002 119.452 119.455 dbcsr_multiply_generic 2527 12.6 0.188 0.194 97.415 98.728 qs_scf_new_mos 118 7.6 0.001 0.001 94.522 95.050 qs_scf_loop_do_ot 118 8.6 0.001 0.001 94.521 95.049 ot_scf_mini 118 9.6 0.004 0.005 89.691 90.253 multiply_cannon 2527 13.6 0.478 0.531 77.419 81.662 multiply_cannon_loop 2527 14.6 1.265 1.307 74.421 76.887 ot_mini 118 10.6 0.001 0.002 49.931 50.575 mp_waitall_1 228564 16.4 24.857 38.083 24.857 38.083 multiply_cannon_multrec 30324 15.6 22.070 26.259 31.736 36.112 rebuild_ks_matrix 129 8.3 0.001 0.001 32.693 33.136 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.024 32.692 33.135 init_scf_loop 11 6.9 0.000 0.000 32.766 32.768 qs_ks_update_qs_env 129 7.6 0.001 0.001 29.497 29.914 multiply_cannon_metrocomm3 30324 15.6 0.096 0.102 15.785 28.714 prepare_preconditioner 11 7.9 0.000 0.000 28.421 28.502 make_preconditioner 11 8.9 0.000 0.000 28.421 28.502 qs_ot_get_derivative 118 11.6 0.001 0.002 27.851 28.424 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.112 27.633 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 22.111 23.359 apply_single 129 13.6 0.001 0.001 22.111 23.358 qs_ot_get_p 129 10.4 0.001 0.001 21.520 22.060 ot_diis_step 118 11.6 0.015 0.018 21.907 21.909 multiply_cannon_sync_h2d 30324 15.6 19.354 21.573 19.354 21.573 qs_ot_p2m_diag 83 11.4 0.188 0.216 16.697 16.732 cp_fm_cholesky_invert 11 10.9 16.674 16.686 16.674 16.686 make_m2s 5054 13.6 0.090 0.095 14.271 15.591 cp_dbcsr_syevd 83 12.4 0.005 0.006 15.557 15.558 make_images 5054 14.6 1.163 1.347 14.059 15.376 sum_up_and_integrate 129 10.3 0.115 0.133 14.377 14.407 integrate_v_rspace 129 11.3 0.003 0.004 14.261 14.297 qs_rho_update_rho_low 129 7.7 0.001 0.002 12.925 12.956 calculate_rho_elec 129 8.7 0.088 0.106 12.924 12.956 cp_fm_diag_elpa 83 13.4 0.000 0.001 12.585 12.596 cp_fm_redistribute_end 83 14.4 7.316 12.522 7.326 12.524 cp_fm_diag_elpa_base 83 14.4 4.961 12.065 5.177 12.433 init_scf_run 11 5.9 0.000 0.001 11.379 11.380 scf_env_initial_rho_setup 11 6.9 0.001 0.003 11.379 11.380 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 10.932 11.365 multiply_cannon_metrocomm4 27797 15.6 0.098 0.116 3.836 10.733 mp_irecv_dv 70031 16.3 3.639 10.343 3.639 10.343 make_images_data 5054 15.6 0.067 0.074 8.644 10.210 dbcsr_mm_accdrv_process 62734 16.2 4.543 5.403 9.123 9.721 hybrid_alltoall_any 5240 16.5 0.344 1.522 7.323 9.383 wfi_extrapolate 11 7.9 0.004 0.028 8.292 8.292 pw_transfer 1559 11.6 0.085 0.105 7.756 7.823 grid_integrate_task_list 129 12.3 7.232 7.659 7.232 7.659 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 7.531 7.590 density_rs2pw 129 9.7 0.006 0.006 6.695 7.179 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 6.264 7.009 cp_fm_cholesky_decompose 22 10.9 6.898 6.985 6.898 6.985 fft_wrap_pw1pw2_140 527 13.2 0.474 0.521 6.599 6.663 calculate_dm_sparse 129 9.5 0.001 0.001 6.510 6.646 rs_pw_transfer 1054 12.0 0.015 0.016 5.527 6.151 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.060 6.068 mp_sum_l 7930 13.1 3.938 5.929 3.938 5.929 fft3d_ps 1301 14.7 2.792 2.941 5.889 5.927 qs_ot_get_orbitals 118 10.6 0.001 0.001 5.371 5.436 grid_collocate_task_list 129 9.7 4.946 5.385 4.946 5.385 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.216 5.355 potential_pw2rs 129 12.3 0.015 0.017 4.751 4.772 mp_alltoall_d11v 2423 14.1 4.004 4.536 4.004 4.536 mp_allgather_i34 2527 14.6 1.565 4.514 1.565 4.514 mp_sum_d 4496 12.2 2.689 4.025 2.689 4.025 dbcsr_complete_redistribute 395 12.7 0.768 0.852 2.971 3.821 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=189.442000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=789.909091, yerr=5.230236 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/007813510527d819d1aa75b16315fe3ed2c20304_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420241154048 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528896499712 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514757E+12 0.0% 0.0% 100.0% flops max/rank 2.928533E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755939872 0.0% 0.0% 100.0% number of processed stacks 3984192 0.0% 0.0% 100.0% average stack size 0.0 0.0 1695.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 933.142528E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1042912 MPI messages size (bytes): total size 2.716210E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.604448E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 281856 36943429632 131072 < size <= 4194304 660064 996105256960 4194304 < size <= 16777216 65632 931531083728 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 57335. MP_Allreduce 11226 986. MP_Sync 170 MP_Alltoall 1712 9388896. MP_ISendRecv 15872 75008. MP_Wait 29756 MP_comm_split 83 MP_ISend 11748 275205. MP_IRecv 11748 275205. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.072 0.134 174.046 174.047 qs_mol_dyn_low 1 2.0 0.003 0.003 173.463 173.477 qs_forces 11 3.9 0.004 0.009 173.357 173.365 qs_energies 11 4.9 0.009 0.023 166.904 166.914 scf_env_do_scf 11 5.9 0.001 0.001 151.499 151.500 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 116.969 116.970 velocity_verlet 10 3.0 0.001 0.002 111.346 111.348 dbcsr_multiply_generic 2507 12.6 0.183 0.189 81.124 82.289 qs_scf_new_mos 117 7.6 0.001 0.001 80.713 81.108 qs_scf_loop_do_ot 117 8.6 0.001 0.002 80.713 81.107 ot_scf_mini 117 9.6 0.004 0.005 76.587 77.017 multiply_cannon 2507 13.6 0.496 0.512 61.382 65.793 multiply_cannon_loop 2507 14.6 0.859 0.886 58.539 61.044 ot_mini 117 10.6 0.001 0.001 42.339 42.758 mp_waitall_1 178456 16.5 25.472 35.028 25.472 35.028 init_scf_loop 11 6.9 0.001 0.004 34.431 34.432 prepare_preconditioner 11 7.9 0.000 0.000 30.444 30.489 make_preconditioner 11 8.9 0.000 0.001 30.444 30.489 rebuild_ks_matrix 128 8.3 0.001 0.001 29.904 30.385 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.018 29.904 30.384 make_full_inverse_cholesky 11 9.9 0.000 0.001 28.151 29.513 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.022 27.468 multiply_cannon_multrec 20056 15.6 13.420 16.486 22.058 25.066 multiply_cannon_metrocomm3 20056 15.6 0.058 0.061 15.406 24.991 qs_ot_get_derivative 117 11.6 0.001 0.002 22.527 22.952 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.878 20.750 apply_single 128 13.6 0.001 0.001 19.878 20.750 qs_ot_get_p 128 10.4 0.001 0.001 19.719 20.181 ot_diis_step 117 11.6 0.018 0.019 19.709 19.710 multiply_cannon_sync_h2d 20056 15.6 14.204 16.191 14.204 16.191 make_m2s 5014 13.6 0.080 0.084 14.867 15.602 qs_ot_p2m_diag 83 11.4 0.266 0.272 15.550 15.560 make_images 5014 14.6 1.174 1.264 14.634 15.374 cp_dbcsr_syevd 83 12.4 0.005 0.005 14.547 14.548 cp_fm_cholesky_invert 11 10.9 14.194 14.202 14.194 14.202 sum_up_and_integrate 128 10.3 0.132 0.143 14.072 14.095 integrate_v_rspace 128 11.3 0.003 0.004 13.940 13.966 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.863 12.907 calculate_rho_elec 128 8.7 0.132 0.146 12.863 12.906 cp_fm_diag_elpa 83 13.4 0.000 0.001 11.455 11.456 cp_fm_redistribute_end 83 14.4 4.302 11.401 4.315 11.403 cp_fm_diag_elpa_base 83 14.4 6.662 10.830 7.069 11.322 make_images_data 5014 15.6 0.061 0.067 9.266 10.460 init_scf_run 11 5.9 0.000 0.001 10.306 10.306 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.305 10.306 hybrid_alltoall_any 5200 16.5 0.434 1.993 8.062 9.432 multiply_cannon_metrocomm4 17549 15.6 0.063 0.074 3.482 9.432 mp_irecv_dv 50230 16.2 3.358 9.181 3.358 9.181 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 8.838 9.159 dbcsr_mm_accdrv_process 41502 16.2 4.413 5.279 8.094 8.237 grid_integrate_task_list 128 12.3 7.318 7.758 7.318 7.758 pw_transfer 1547 11.6 0.084 0.103 7.651 7.751 fft_wrap_pw1pw2 1291 12.7 0.010 0.010 7.428 7.533 cp_fm_upper_to_full 105 14.5 5.677 7.379 5.677 7.379 wfi_extrapolate 11 7.9 0.001 0.001 7.321 7.321 cp_fm_cholesky_decompose 22 10.9 7.212 7.241 7.212 7.241 density_rs2pw 128 9.7 0.006 0.006 6.360 6.781 fft_wrap_pw1pw2_140 523 13.2 0.480 0.531 6.561 6.675 dbcsr_complete_redistribute 395 12.7 1.170 1.201 4.500 6.240 calculate_dm_sparse 128 9.5 0.001 0.001 5.737 5.823 fft3d_ps 1291 14.7 2.691 2.893 5.694 5.754 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.428 5.434 rs_pw_transfer 1046 11.9 0.014 0.014 4.976 5.411 grid_collocate_task_list 128 9.7 5.065 5.377 5.065 5.377 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.500 5.185 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.337 5.071 mp_alltoall_d11v 2415 14.1 4.227 4.841 4.227 4.841 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 4.527 4.672 mp_sum_l 7870 13.0 3.083 4.489 3.083 4.489 potential_pw2rs 128 12.3 0.020 0.022 4.441 4.456 mp_allgather_i34 2507 14.6 1.367 4.422 1.367 4.422 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.008 4.037 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.274 3.954 qs_energies_init_hamiltonians 11 5.9 0.003 0.007 3.771 3.773 mp_alltoall_i22 716 14.1 1.885 3.724 1.885 3.724 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=174.047000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=885.363636, yerr=8.689311 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/007813510527d819d1aa75b16315fe3ed2c20304_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420241154048 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528896499712 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514757E+12 0.0% 0.0% 100.0% flops max/rank 4.353788E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755939872 0.0% 0.0% 100.0% number of processed stacks 5977344 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.138991E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1143192 MPI messages size (bytes): total size 2.023815E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.770319E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 319024 36042702848 131072 < size <= 4194304 715736 785529176064 4194304 < size <= 16777216 70320 665379345520 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 57332. MP_Allreduce 11226 1068. MP_Sync 170 MP_Alltoall 1712 12503107. MP_ISendRecv 11776 75008. MP_Wait 28330 MP_comm_split 83 MP_ISend 14952 244818. MP_IRecv 14952 244818. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.022 0.047 187.563 187.564 qs_mol_dyn_low 1 2.0 0.003 0.003 187.167 187.180 qs_forces 11 3.9 0.005 0.018 187.046 187.055 qs_energies 11 4.9 0.003 0.011 180.012 180.026 scf_env_do_scf 11 5.9 0.001 0.004 163.200 163.210 velocity_verlet 10 3.0 0.002 0.002 124.523 124.527 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 116.199 116.200 qs_scf_new_mos 117 7.6 0.001 0.001 80.612 80.895 qs_scf_loop_do_ot 117 8.6 0.001 0.001 80.611 80.894 dbcsr_multiply_generic 2507 12.6 0.192 0.198 79.563 80.578 ot_scf_mini 117 9.6 0.003 0.004 76.120 76.429 multiply_cannon 2507 13.6 0.557 0.599 54.685 57.230 multiply_cannon_loop 2507 14.6 1.174 1.200 50.817 53.040 init_scf_loop 11 6.9 0.001 0.007 46.877 46.878 prepare_preconditioner 11 7.9 0.000 0.000 42.778 42.796 make_preconditioner 11 8.9 0.000 0.002 42.778 42.796 ot_mini 117 10.6 0.001 0.001 42.360 42.675 make_full_inverse_cholesky 11 9.9 0.000 0.001 36.354 41.430 multiply_cannon_multrec 30084 15.6 14.285 19.679 26.169 30.973 rebuild_ks_matrix 128 8.3 0.001 0.001 29.111 29.379 qs_ks_build_kohn_sham_matrix 128 9.3 0.019 0.036 29.110 29.378 mp_waitall_1 153770 16.5 17.634 28.842 17.634 28.842 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.334 26.581 qs_ot_get_derivative 117 11.6 0.001 0.002 22.766 23.076 make_m2s 5014 13.6 0.094 0.098 20.424 21.608 make_images 5014 14.6 1.951 2.313 20.119 21.304 qs_ot_get_p 128 10.4 0.001 0.001 19.920 20.280 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 18.949 19.599 apply_single 128 13.6 0.001 0.001 18.949 19.599 ot_diis_step 117 11.6 0.017 0.018 19.459 19.461 cp_fm_upper_to_full 105 14.7 11.207 16.641 11.207 16.641 cp_fm_cholesky_invert 11 10.9 16.422 16.431 16.422 16.431 multiply_cannon_metrocomm3 30084 15.6 0.047 0.049 6.266 16.227 qs_ot_p2m_diag 83 11.4 0.343 0.390 15.789 15.842 cp_dbcsr_syevd 83 12.4 0.005 0.005 14.507 14.508 sum_up_and_integrate 128 10.3 0.141 0.153 14.065 14.088 integrate_v_rspace 128 11.3 0.003 0.004 13.924 13.952 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.106 13.163 calculate_rho_elec 128 8.7 0.175 0.190 13.105 13.162 dbcsr_complete_redistribute 395 12.7 1.507 1.627 9.090 12.925 multiply_cannon_sync_h2d 30084 15.6 11.705 12.864 11.705 12.864 make_images_data 5014 15.6 0.063 0.068 11.068 12.838 dbcsr_mm_accdrv_process 62264 16.2 7.318 8.392 11.460 11.951 hybrid_alltoall_any 5200 16.5 0.525 2.189 9.684 11.520 copy_fm_to_dbcsr 209 11.7 0.002 0.003 7.701 11.504 cp_fm_diag_elpa 83 13.4 0.000 0.001 11.373 11.377 cp_fm_redistribute_end 83 14.4 1.937 11.306 1.950 11.311 cp_fm_diag_elpa_base 83 14.4 8.738 10.708 9.329 11.205 init_scf_run 11 5.9 0.000 0.001 10.876 10.877 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.875 10.877 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.401 10.115 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.426 9.645 mp_alltoall_i22 716 14.1 5.634 9.412 5.634 9.412 pw_transfer 1547 11.6 0.084 0.097 7.875 7.953 grid_integrate_task_list 128 12.3 7.470 7.902 7.470 7.902 cp_fm_cholesky_decompose 22 10.9 7.669 7.762 7.669 7.762 fft_wrap_pw1pw2 1291 12.7 0.010 0.010 7.652 7.734 wfi_extrapolate 11 7.9 0.001 0.001 7.674 7.674 multiply_cannon_metrocomm4 25070 15.6 0.076 0.086 2.734 6.967 fft_wrap_pw1pw2_140 523 13.2 0.482 0.489 6.749 6.829 mp_irecv_dv 76098 16.2 2.591 6.707 2.591 6.707 density_rs2pw 128 9.7 0.006 0.006 6.278 6.538 calculate_dm_sparse 128 9.5 0.001 0.001 6.194 6.277 fft3d_ps 1291 14.7 2.800 2.907 5.880 5.952 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.514 5.557 grid_collocate_task_list 128 9.7 5.203 5.522 5.203 5.522 mp_alltoall_d11v 2415 14.1 4.860 5.504 4.860 5.504 rs_pw_transfer 1046 11.9 0.013 0.014 4.649 4.946 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.384 4.477 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 4.420 4.474 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 4.415 4.416 potential_pw2rs 128 12.3 0.023 0.023 4.296 4.315 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.212 4.281 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=187.564000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1073.454545, yerr=20.173627 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/007813510527d819d1aa75b16315fe3ed2c20304_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456039424 0.0% 0.0% 100.0% flops 32 x 32 x 32 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 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 94.243758E+12 0.0% 0.0% 100.0% flops max/rank 5.910792E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806547744 0.0% 0.0% 100.0% number of processed stacks 1978768 0.0% 0.0% 100.0% average stack size 0.0 0.0 3439.8 marketing flops 145.661668E+12 ------------------------------------------------------------------------------- # multiplications 2534 max memory usage/rank 1.506460E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 243264 MPI messages size (bytes): total size 1.342058E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.516877E+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 115488 60548972544 4194304 < size <= 16777216 105840 554906419200 16777216 < size 20352 726592466352 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 9025 51. MP_Alltoall 9734 793691. MP_ISend 40500 2096701. MP_IRecv 40500 2095807. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 57630. MP_Allreduce 11104 1171. MP_Sync 88 MP_Alltoall 1724 18848034. 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 0.060 0.080 172.402 172.403 qs_mol_dyn_low 1 2.0 0.003 0.003 171.970 171.996 qs_forces 11 3.9 0.003 0.003 171.878 171.881 qs_energies 11 4.9 0.001 0.002 164.576 164.583 scf_env_do_scf 11 5.9 0.001 0.001 147.628 147.640 velocity_verlet 10 3.0 0.002 0.002 113.305 113.309 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 111.409 111.411 qs_scf_new_mos 118 7.6 0.001 0.001 75.617 75.752 qs_scf_loop_do_ot 118 8.6 0.001 0.001 75.617 75.751 dbcsr_multiply_generic 2534 12.6 0.183 0.190 74.479 75.062 ot_scf_mini 118 9.6 0.004 0.004 71.148 71.272 multiply_cannon 2534 13.6 0.584 0.621 54.707 59.481 multiply_cannon_loop 2534 14.6 0.452 0.463 50.074 51.174 ot_mini 118 10.6 0.001 0.001 39.774 39.886 init_scf_loop 11 6.9 0.000 0.000 36.068 36.071 mp_waitall_1 130986 16.6 26.266 33.241 26.266 33.241 prepare_preconditioner 11 7.9 0.000 0.000 32.187 32.221 make_preconditioner 11 8.9 0.000 0.000 32.187 32.221 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.984 30.276 rebuild_ks_matrix 129 8.3 0.001 0.001 28.721 28.878 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.018 28.720 28.877 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.264 26.413 multiply_cannon_multrec 10136 15.6 10.503 14.853 17.814 20.712 qs_ot_get_derivative 118 11.6 0.002 0.002 19.836 19.949 ot_diis_step 118 11.6 0.020 0.021 19.870 19.870 multiply_cannon_metrocomm3 10136 15.6 0.023 0.024 12.283 19.846 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.379 19.788 apply_single 129 13.6 0.001 0.001 19.379 19.787 qs_ot_get_p 129 10.4 0.001 0.001 18.377 18.490 cp_fm_cholesky_invert 11 10.9 18.430 18.436 18.430 18.436 make_m2s 5068 13.6 0.066 0.073 16.111 18.253 make_images 5068 14.6 2.327 2.605 15.804 17.950 qs_ot_p2m_diag 84 11.4 0.502 0.507 14.614 14.631 sum_up_and_integrate 129 10.3 0.182 0.192 14.107 14.153 integrate_v_rspace 129 11.3 0.004 0.004 13.925 13.977 cp_dbcsr_syevd 84 12.4 0.005 0.005 13.441 13.443 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.168 13.204 calculate_rho_elec 129 8.7 0.260 0.271 13.167 13.203 multiply_cannon_sync_h2d 10136 15.6 11.674 12.294 11.674 12.294 make_images_data 5068 15.6 0.053 0.061 9.796 12.200 hybrid_alltoall_any 5255 16.5 0.790 3.662 9.570 12.126 cp_fm_diag_elpa 84 13.4 0.000 0.000 10.389 10.390 cp_fm_diag_elpa_base 84 14.4 10.136 10.219 10.383 10.383 init_scf_run 11 5.9 0.000 0.001 10.180 10.180 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.180 10.180 cp_fm_cholesky_decompose 22 10.9 8.068 8.180 8.068 8.180 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 8.059 8.146 grid_integrate_task_list 129 12.3 7.836 8.064 7.836 8.064 multiply_cannon_metrocomm1 10136 15.6 0.030 0.030 4.651 7.960 pw_transfer 1559 11.6 0.083 0.091 7.791 7.809 mp_allgather_i34 2534 14.6 2.824 7.717 2.824 7.717 fft_wrap_pw1pw2 1301 12.7 0.010 0.010 7.570 7.595 dbcsr_mm_accdrv_process 20954 16.1 2.647 3.457 6.937 7.553 wfi_extrapolate 11 7.9 0.001 0.001 7.300 7.300 fft_wrap_pw1pw2_140 527 13.2 0.508 0.528 6.640 6.668 density_rs2pw 129 9.7 0.005 0.006 5.982 6.347 calculate_dm_sparse 129 9.5 0.001 0.001 6.117 6.175 grid_collocate_task_list 129 9.7 5.525 5.831 5.525 5.831 fft3d_ps 1301 14.7 2.741 2.804 5.715 5.747 dbcsr_complete_redistribute 397 12.7 2.117 2.228 5.275 5.684 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.219 5.220 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.138 5.154 mp_alltoall_d11v 2429 14.1 4.675 5.100 4.675 5.100 rs_pw_transfer 1054 12.0 0.013 0.013 4.160 4.412 potential_pw2rs 129 12.3 0.027 0.028 4.137 4.153 multiply_cannon_metrocomm4 7602 15.6 0.024 0.027 1.883 4.096 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.052 4.088 mp_irecv_dv 29142 15.9 1.846 4.020 1.846 4.020 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.559 3.866 copy_fm_to_dbcsr 210 11.7 0.002 0.002 3.486 3.814 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.691 3.725 copy_dbcsr_to_fm 187 11.8 0.004 0.004 3.509 3.607 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.438 3.460 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=172.403000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1409.727273, yerr=43.830440 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/007813510527d819d1aa75b16315fe3ed2c20304_performance_tests/20/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 11.606413E+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 1947808 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.6 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 2.933182E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 99400 MPI messages size (bytes): total size 1.127422E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.342275E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 44768 34745614336 4194304 < size <= 16777216 43984 376564613120 16777216 < size 10032 716108638608 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4003 59127. MP_Allreduce 11005 1515. MP_Sync 86 MP_Alltoall 1700 36954383. MP_ISendRecv 3556 218624. MP_Wait 11506 MP_ISend 6360 1080477. MP_IRecv 6360 1080477. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.023 0.043 290.217 290.217 qs_mol_dyn_low 1 2.0 0.003 0.003 289.701 289.713 qs_forces 11 3.9 0.003 0.003 289.612 289.614 qs_energies 11 4.9 0.002 0.002 281.049 281.064 scf_env_do_scf 11 5.9 0.001 0.001 259.521 259.531 velocity_verlet 10 3.0 0.002 0.002 209.125 209.133 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 133.151 133.152 init_scf_loop 11 6.9 0.000 0.000 126.114 126.115 prepare_preconditioner 11 7.9 0.000 0.000 121.447 121.474 make_preconditioner 11 8.9 0.000 0.000 121.447 121.474 make_full_inverse_cholesky 11 9.9 0.000 0.000 96.875 118.587 qs_scf_new_mos 116 7.6 0.001 0.001 91.674 91.786 qs_scf_loop_do_ot 116 8.6 0.001 0.001 91.673 91.785 ot_scf_mini 116 9.6 0.004 0.004 86.938 86.963 dbcsr_multiply_generic 2485 12.5 0.216 0.223 83.627 84.264 cp_fm_upper_to_full 104 14.8 53.994 77.561 53.994 77.561 multiply_cannon 2485 13.5 0.679 0.722 58.597 58.976 multiply_cannon_loop 2485 14.5 0.465 0.476 55.056 56.333 ot_mini 116 10.6 0.001 0.001 45.322 45.349 dbcsr_complete_redistribute 393 12.7 3.986 4.024 30.197 43.066 copy_fm_to_dbcsr 208 11.6 0.001 0.002 26.838 39.717 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 24.526 37.343 mp_alltoall_i22 712 14.1 22.294 35.108 22.294 35.108 cp_fm_cholesky_invert 11 10.9 33.727 33.733 33.727 33.733 rebuild_ks_matrix 127 8.3 0.001 0.001 33.149 33.201 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.018 33.148 33.200 mp_waitall_1 104546 16.7 28.169 31.892 28.169 31.892 qs_ks_update_qs_env 127 7.6 0.001 0.001 30.916 30.968 qs_ot_get_p 127 10.4 0.001 0.001 26.215 26.248 qs_ot_get_derivative 116 11.6 0.001 0.002 25.314 25.341 qs_ot_p2m_diag 82 11.4 0.867 0.872 22.234 22.263 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 20.179 20.598 apply_single 127 13.6 0.001 0.001 20.179 20.598 cp_dbcsr_syevd 82 12.4 0.005 0.006 20.521 20.523 multiply_cannon_metrocomm3 9940 15.5 0.023 0.023 19.045 20.299 make_m2s 4970 13.5 0.074 0.076 19.385 20.261 ot_diis_step 116 11.6 0.022 0.022 19.979 19.979 make_images 4970 14.5 3.731 3.835 18.913 19.790 multiply_cannon_multrec 9940 15.5 10.463 12.166 17.805 17.887 cp_fm_diag_elpa 82 13.4 0.000 0.000 17.343 17.344 cp_fm_diag_elpa_base 82 14.4 12.884 14.490 17.339 17.339 sum_up_and_integrate 127 10.3 0.318 0.321 15.491 15.588 multiply_cannon_sync_h2d 9940 15.5 15.530 15.537 15.530 15.537 integrate_v_rspace 127 11.3 0.004 0.004 15.172 15.271 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.820 14.837 calculate_rho_elec 127 8.7 0.479 0.480 14.819 14.836 make_images_data 4970 15.5 0.059 0.064 10.216 11.872 hybrid_alltoall_any 5155 16.4 1.291 3.028 10.332 11.840 init_scf_run 11 5.9 0.000 0.001 11.760 11.761 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.760 11.760 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.335 9.360 cp_fm_cholesky_decompose 22 10.9 8.995 9.019 8.995 9.019 dbcsr_mm_accdrv_process 20590 16.0 3.814 5.634 7.108 8.858 wfi_extrapolate 11 7.9 0.001 0.001 8.733 8.734 grid_integrate_task_list 127 12.3 8.482 8.647 8.482 8.647 pw_transfer 1535 11.6 0.089 0.089 8.265 8.274 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.034 8.042 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 7.826 7.826 fft_wrap_pw1pw2_140 519 13.2 0.535 0.536 7.103 7.119 mp_alltoall_d11v 2401 14.1 6.857 7.030 6.857 7.030 calculate_dm_sparse 127 9.5 0.001 0.001 6.576 6.688 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.305 6.397 grid_collocate_task_list 127 9.7 6.293 6.325 6.293 6.325 fft3d_ps 1281 14.7 2.733 2.745 6.094 6.103 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.003 6.074 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.665 6.063 density_rs2pw 127 9.7 0.005 0.005 5.951 5.992 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=290.217000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2596.636364, yerr=173.517761 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/007813510527d819d1aa75b16315fe3ed2c20304_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 1.261818E+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 0.023 0.084 84.664 84.665 qs_energies 1 2.0 0.000 0.000 84.173 84.185 ls_scf 1 3.0 0.000 0.000 83.276 83.288 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.528 72.697 multiply_cannon 111 7.7 0.017 0.020 55.880 57.002 multiply_cannon_loop 111 8.7 0.211 0.225 52.470 53.693 ls_scf_main 1 4.0 0.000 0.000 52.165 52.165 density_matrix_trs4 2 5.0 0.002 0.003 46.762 46.833 ls_scf_init_scf 1 4.0 0.000 0.000 28.048 28.049 ls_scf_init_matrix_S 1 5.0 0.000 0.000 26.991 27.038 mp_waitall_1 11316 10.9 22.330 26.282 22.330 26.282 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 24.879 24.897 multiply_cannon_multrec 2664 9.7 8.164 8.936 15.462 17.195 multiply_cannon_sync_h2d 2664 9.7 13.735 15.536 13.735 15.536 make_m2s 222 7.7 0.008 0.012 13.033 13.533 make_images 222 8.7 0.098 0.107 13.011 13.514 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.689 13.401 make_images_data 222 9.7 0.004 0.005 7.554 8.111 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.413 8.100 hybrid_alltoall_any 227 10.6 0.215 1.832 6.458 7.912 dbcsr_mm_accdrv_process 4760 10.4 0.509 0.610 6.916 7.853 dbcsr_mm_accdrv_process_sort 4760 11.4 6.208 7.078 6.208 7.078 calculate_norms 4752 9.8 5.519 6.105 5.519 6.105 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.026 5.175 mp_sum_l 807 5.4 3.135 4.467 3.135 4.467 make_images_sizes 222 9.7 0.000 0.000 0.791 3.727 mp_alltoall_i44 222 10.7 0.790 3.727 0.790 3.727 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.310 3.322 multiply_cannon_metrocomm4 2442 9.7 0.012 0.014 2.068 3.301 mp_irecv_dv 6231 10.9 2.051 3.271 2.051 3.271 arnoldi_extremal 4 6.8 0.000 0.000 3.173 3.192 arnoldi_normal_ev 4 7.8 0.001 0.003 3.173 3.192 build_subspace 16 8.4 0.009 0.012 3.081 3.084 ls_scf_post 1 4.0 0.000 0.000 3.063 3.075 ls_scf_store_result 1 5.0 0.000 0.000 2.844 2.890 dbcsr_special_finalize 555 9.7 0.005 0.006 2.336 2.890 dbcsr_merge_single_wm 555 10.7 0.455 0.595 2.328 2.882 make_images_pack 222 9.7 2.208 2.629 2.209 2.631 dbcsr_sort_data 658 11.4 2.130 2.607 2.130 2.607 dbcsr_matrix_vector_mult 304 9.0 0.003 0.010 2.303 2.540 dbcsr_matrix_vector_mult_local 304 10.0 2.066 2.453 2.068 2.455 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.168 2.248 buffer_matrices_ensure_size 222 8.7 1.760 2.126 1.760 2.126 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.732 1.733 rebuild_ks_matrix 3 7.3 0.000 0.000 1.722 1.723 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.014 1.722 1.723 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=84.665000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1133.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/007813510527d819d1aa75b16315fe3ed2c20304_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 2.104660E+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 0.057 0.085 91.620 91.623 qs_energies 1 2.0 0.000 0.000 91.029 91.051 ls_scf 1 3.0 0.000 0.001 89.721 89.743 dbcsr_multiply_generic 111 6.7 0.015 0.015 75.802 76.192 multiply_cannon 111 7.7 0.028 0.040 53.665 57.329 ls_scf_main 1 4.0 0.000 0.000 54.909 54.921 multiply_cannon_loop 111 8.7 0.116 0.124 50.381 53.901 density_matrix_trs4 2 5.0 0.002 0.003 49.248 49.446 ls_scf_init_scf 1 4.0 0.000 0.001 31.308 31.310 mp_waitall_1 9246 10.9 21.222 30.707 21.222 30.707 ls_scf_init_matrix_S 1 5.0 0.000 0.001 30.146 30.273 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.726 27.742 multiply_cannon_multrec 1332 9.7 13.132 17.032 22.327 27.349 multiply_cannon_metrocomm3 1332 9.7 0.006 0.008 11.693 21.145 make_m2s 222 7.7 0.006 0.007 15.477 16.236 make_images 222 8.7 1.575 1.968 15.447 16.208 dbcsr_mm_accdrv_process 4041 10.4 0.278 0.445 8.791 10.323 make_images_data 222 9.7 0.004 0.004 8.948 9.957 dbcsr_mm_accdrv_process_sort 4041 11.4 8.384 9.878 8.384 9.878 hybrid_alltoall_any 227 10.6 0.523 2.496 8.301 9.207 mp_sum_l 807 5.4 5.378 8.681 5.378 8.681 multiply_cannon_metrocomm4 1221 9.7 0.007 0.008 3.247 7.628 mp_irecv_dv 3311 11.0 3.228 7.567 3.228 7.567 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.149 6.884 calculate_norms 2376 9.8 6.019 6.812 6.019 6.812 multiply_cannon_sync_h2d 1332 9.7 4.790 6.132 4.790 6.132 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.036 5.214 arnoldi_extremal 4 6.8 0.000 0.000 4.680 4.703 arnoldi_normal_ev 4 7.8 0.001 0.004 4.680 4.703 build_subspace 16 8.4 0.014 0.021 4.416 4.423 ls_scf_post 1 4.0 0.003 0.023 3.504 3.526 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.149 3.383 ls_scf_store_result 1 5.0 0.000 0.000 3.198 3.323 dbcsr_matrix_vector_mult_local 304 10.0 2.740 3.226 2.742 3.227 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.207 2.749 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.534 2.638 mp_allgather_i34 111 8.7 0.985 2.476 0.985 2.476 make_images_pack 222 9.7 2.025 2.389 2.027 2.391 dbcsr_sort_data 436 11.2 1.846 2.066 1.846 2.066 dbcsr_data_new 4174 10.1 1.610 1.856 1.610 1.856 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.623000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1705.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/007813510527d819d1aa75b16315fe3ed2c20304_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 2.702422E+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 0.048 0.070 94.350 94.351 qs_energies 1 2.0 0.000 0.000 93.653 93.674 ls_scf 1 3.0 0.000 0.000 92.233 92.256 dbcsr_multiply_generic 111 6.7 0.015 0.016 77.119 77.396 multiply_cannon 111 7.7 0.040 0.087 53.338 57.783 ls_scf_main 1 4.0 0.000 0.000 57.386 57.389 multiply_cannon_loop 111 8.7 0.100 0.106 49.789 52.882 density_matrix_trs4 2 5.0 0.002 0.003 51.490 51.645 mp_waitall_1 7374 11.0 24.055 33.250 24.055 33.250 ls_scf_init_scf 1 4.0 0.000 0.000 31.253 31.256 ls_scf_init_matrix_S 1 5.0 0.000 0.000 30.135 30.200 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.743 27.756 multiply_cannon_multrec 888 9.7 12.583 15.159 21.100 24.320 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.096 22.152 make_m2s 222 7.7 0.006 0.007 17.263 18.439 make_images 222 8.7 1.964 2.305 17.225 18.401 make_images_data 222 9.7 0.003 0.004 9.936 10.975 hybrid_alltoall_any 227 10.6 0.621 2.865 9.532 10.863 dbcsr_mm_accdrv_process 3754 10.4 0.243 0.428 8.037 9.267 dbcsr_mm_accdrv_process_sort 3754 11.4 7.664 8.839 7.664 8.839 mp_sum_l 807 5.4 4.958 8.785 4.958 8.785 multiply_cannon_sync_h2d 888 9.7 6.032 7.669 6.032 7.669 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.827 7.031 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.491 6.778 mp_irecv_dv 2335 11.1 2.475 6.732 2.475 6.732 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.642 6.630 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.941 5.123 arnoldi_extremal 4 6.8 0.000 0.000 5.102 5.119 arnoldi_normal_ev 4 7.8 0.001 0.005 5.102 5.119 build_subspace 16 8.4 0.014 0.020 4.798 4.804 calculate_norms 1584 9.8 4.266 4.600 4.266 4.600 mp_allgather_i34 111 8.7 1.389 3.884 1.389 3.884 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.447 3.788 ls_scf_post 1 4.0 0.000 0.000 3.593 3.613 dbcsr_matrix_vector_mult_local 304 10.0 3.021 3.597 3.023 3.599 ls_scf_store_result 1 5.0 0.000 0.000 3.330 3.416 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.747 2.859 make_images_sizes 222 9.7 0.000 0.000 1.003 2.245 mp_alltoall_i44 222 10.7 1.003 2.245 1.003 2.245 make_images_pack 222 9.7 1.813 2.133 1.815 2.136 dbcsr_sort_data 325 11.1 1.887 2.129 1.887 2.129 dbcsr_data_release 9322 10.9 1.319 2.028 1.319 2.028 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=94.351000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2157.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/007813510527d819d1aa75b16315fe3ed2c20304_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 3.326636E+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 0.032 0.049 96.791 96.792 qs_energies 1 2.0 0.000 0.000 96.195 96.201 ls_scf 1 3.0 0.000 0.000 94.560 94.565 dbcsr_multiply_generic 111 6.7 0.016 0.018 78.388 78.694 ls_scf_main 1 4.0 0.000 0.000 58.670 58.671 multiply_cannon 111 7.7 0.046 0.099 51.462 56.433 density_matrix_trs4 2 5.0 0.002 0.003 52.662 52.794 multiply_cannon_loop 111 8.7 0.115 0.126 46.420 49.919 ls_scf_init_scf 1 4.0 0.000 0.000 32.650 32.652 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.514 31.577 mp_waitall_1 6438 11.0 22.731 29.483 22.731 29.483 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.043 29.055 multiply_cannon_multrec 1332 9.7 14.240 17.894 22.069 25.018 make_m2s 222 7.7 0.007 0.008 21.230 22.622 make_images 222 8.7 3.124 3.589 21.180 22.574 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.100 17.669 make_images_data 222 9.7 0.004 0.004 11.889 13.671 hybrid_alltoall_any 227 10.6 0.796 3.832 11.176 12.981 dbcsr_mm_accdrv_process 3641 10.4 0.227 0.416 7.463 8.975 dbcsr_mm_accdrv_process_sort 3641 11.4 7.090 8.559 7.090 8.559 mp_sum_l 807 5.4 4.133 7.376 4.133 7.376 multiply_cannon_sync_h2d 1332 9.7 5.494 6.215 5.494 6.215 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.078 5.976 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.565 5.970 mp_irecv_dv 3229 10.9 2.055 5.922 2.055 5.922 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.203 5.738 arnoldi_extremal 4 6.8 0.000 0.000 5.277 5.290 arnoldi_normal_ev 4 7.8 0.008 0.046 5.277 5.290 build_subspace 16 8.4 0.015 0.021 4.890 4.897 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.520 4.712 mp_allgather_i34 111 8.7 2.168 4.553 2.168 4.553 calculate_norms 2376 9.8 4.179 4.546 4.179 4.546 dbcsr_matrix_vector_mult 304 9.0 0.006 0.016 3.605 3.879 dbcsr_matrix_vector_mult_local 304 10.0 3.194 3.689 3.195 3.691 dbcsr_sort_data 658 11.4 3.029 3.540 3.029 3.540 ls_scf_post 1 4.0 0.000 0.000 3.240 3.247 dbcsr_special_finalize 555 9.7 0.006 0.007 2.771 3.228 dbcsr_merge_single_wm 555 10.7 0.529 0.666 2.763 3.221 ls_scf_store_result 1 5.0 0.000 0.000 2.982 3.042 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.917 2.997 dbcsr_data_release 10477 10.7 1.579 2.473 1.579 2.473 dbcsr_finalize 304 7.8 0.049 0.061 1.793 2.045 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=96.792000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2742.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/007813510527d819d1aa75b16315fe3ed2c20304_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 4.630589E+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 0.051 0.067 98.203 98.204 qs_energies 1 2.0 0.000 0.000 97.436 97.441 ls_scf 1 3.0 0.000 0.000 95.523 95.526 dbcsr_multiply_generic 111 6.7 0.017 0.018 77.192 77.405 ls_scf_main 1 4.0 0.000 0.000 61.587 61.588 multiply_cannon 111 7.7 0.076 0.194 55.066 60.325 density_matrix_trs4 2 5.0 0.002 0.003 54.625 54.725 multiply_cannon_loop 111 8.7 0.070 0.076 50.454 52.133 mp_waitall_1 5481 11.0 25.881 30.875 25.881 30.875 ls_scf_init_scf 1 4.0 0.000 0.000 30.334 30.339 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.167 29.200 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.011 27.024 multiply_cannon_multrec 444 9.7 14.064 16.630 21.074 23.101 make_m2s 222 7.7 0.004 0.005 17.420 20.038 make_images 222 8.7 3.713 4.409 17.359 19.978 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.819 16.594 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.184 13.931 make_images_data 222 9.7 0.003 0.004 9.677 12.228 hybrid_alltoall_any 227 10.6 0.789 3.764 9.410 11.972 dbcsr_mm_accdrv_process 3003 10.4 0.176 0.341 6.718 7.834 multiply_cannon_sync_h2d 444 9.7 6.576 7.591 6.576 7.591 dbcsr_mm_accdrv_process_sort 3003 11.4 6.403 7.488 6.403 7.488 mp_allgather_i34 111 8.7 2.848 7.079 2.848 7.079 arnoldi_extremal 4 6.8 0.000 0.000 5.917 5.931 arnoldi_normal_ev 4 7.8 0.002 0.005 5.917 5.931 build_subspace 16 8.4 0.015 0.019 5.527 5.537 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.566 4.696 dbcsr_matrix_vector_mult 304 9.0 0.007 0.016 4.222 4.427 mp_sum_l 807 5.4 2.749 4.200 2.749 4.200 dbcsr_matrix_vector_mult_local 304 10.0 3.711 4.188 3.713 4.191 calculate_norms 792 9.8 3.553 3.722 3.553 3.722 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.620 3.650 ls_scf_post 1 4.0 0.000 0.000 3.602 3.606 mp_irecv_dv 1241 11.2 1.601 3.604 1.601 3.604 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.381 3.483 ls_scf_store_result 1 5.0 0.000 0.000 3.372 3.427 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.849 3.305 make_images_sizes 222 9.7 0.000 0.000 1.110 3.195 mp_alltoall_i44 222 10.7 1.109 3.194 1.109 3.194 dbcsr_finalize 304 7.8 0.062 0.078 2.201 2.326 dbcsr_merge_all 275 8.9 0.474 0.534 2.047 2.160 dbcsr_data_release 10123 10.8 1.333 1.994 1.333 1.994 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=98.204000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3585.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/007813510527d819d1aa75b16315fe3ed2c20304_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 8.690651E+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 0.076 0.093 107.754 107.754 qs_energies 1 2.0 0.000 0.000 106.370 106.383 ls_scf 1 3.0 0.000 0.000 103.451 103.464 dbcsr_multiply_generic 111 6.7 0.023 0.028 77.624 77.706 ls_scf_main 1 4.0 0.000 0.000 65.855 65.856 density_matrix_trs4 2 5.0 0.002 0.003 56.983 57.040 multiply_cannon 111 7.7 0.154 0.293 49.963 51.938 multiply_cannon_loop 111 8.7 0.067 0.070 46.476 47.492 ls_scf_init_scf 1 4.0 0.000 0.000 33.932 33.933 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.604 32.625 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.718 29.774 mp_waitall_1 4569 11.1 22.417 26.219 22.417 26.219 make_m2s 222 7.7 0.005 0.005 23.960 24.997 make_images 222 8.7 4.585 4.967 23.854 24.889 multiply_cannon_multrec 444 9.7 17.817 18.644 22.403 23.058 hybrid_alltoall_any 227 10.6 1.660 3.615 12.992 15.712 make_images_data 222 9.7 0.003 0.003 13.195 15.666 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.628 11.175 multiply_cannon_sync_h2d 444 9.7 8.829 8.870 8.829 8.870 arnoldi_extremal 4 6.8 0.000 0.000 7.379 7.391 arnoldi_normal_ev 4 7.8 0.003 0.009 7.379 7.391 build_subspace 16 8.4 0.026 0.037 6.826 6.837 dbcsr_matrix_vector_mult 304 9.0 0.009 0.025 5.485 5.648 dbcsr_matrix_vector_mult_local 304 10.0 5.081 5.400 5.084 5.403 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.028 5.265 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.835 4.926 dbcsr_mm_accdrv_process 1814 10.4 0.253 0.324 4.413 4.547 dbcsr_mm_accdrv_process_sort 1814 11.4 4.114 4.251 4.114 4.251 ls_scf_post 1 4.0 0.000 0.000 3.664 3.676 mp_allgather_i34 111 8.7 1.145 3.617 1.145 3.617 make_images_sizes 222 9.7 0.000 0.000 1.433 3.498 mp_alltoall_i44 222 10.7 1.432 3.498 1.432 3.498 ls_scf_store_result 1 5.0 0.000 0.000 3.412 3.424 calculate_norms 792 9.8 3.263 3.294 3.263 3.294 dbcsr_finalize 304 7.8 0.082 0.089 3.086 3.141 dbcsr_complete_redistribute 5 7.6 1.465 1.512 2.785 2.929 dbcsr_merge_all 275 8.9 0.892 0.920 2.872 2.917 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.889 2.889 dbcsr_data_release 12724 10.6 2.326 2.854 2.326 2.854 matrix_ls_to_qs 2 6.0 0.000 0.000 2.448 2.593 dbcsr_sort_data 325 11.1 2.442 2.494 2.442 2.494 dbcsr_new_transposed 4 7.5 0.247 0.257 2.288 2.304 dbcsr_frobenius_norm 74 6.6 2.057 2.140 2.208 2.257 dbcsr_add_d 103 6.2 0.000 0.000 2.134 2.206 dbcsr_add_anytype 103 7.2 0.859 0.892 2.134 2.206 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.187 2.189 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=107.754000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6882.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 007813510527d819d1aa75b16315fe3ed2c20304 Summary: empty Status: OK