=== 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: 4047e199d09ff4d3c3dd9bcab791603060e8611a ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, Cray-FFTW 3.3.8.10, # COSMA 2.6.6, ELPA 2022.11.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.2.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.2, # SIRIUS 7.4.3, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Author: Matthias Krack (31.05.2023) # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.2 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4047e199d09ff4d3c3dd9bcab791603060e8611a_performance_tests/01 job id: 47165416 --- 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/4047e199d09ff4d3c3dd9bcab791603060e8611a_performance_tests/02 job id: 47165417 --- 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/4047e199d09ff4d3c3dd9bcab791603060e8611a_performance_tests/03 job id: 47165418 --- 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/4047e199d09ff4d3c3dd9bcab791603060e8611a_performance_tests/04 job id: 47165419 --- 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/4047e199d09ff4d3c3dd9bcab791603060e8611a_performance_tests/05 job id: 47165421 --- 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/4047e199d09ff4d3c3dd9bcab791603060e8611a_performance_tests/06 job id: 47165422 --- 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/4047e199d09ff4d3c3dd9bcab791603060e8611a_performance_tests/07 job id: 47165423 --- 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/4047e199d09ff4d3c3dd9bcab791603060e8611a_performance_tests/08 job id: 47165424 --- 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/4047e199d09ff4d3c3dd9bcab791603060e8611a_performance_tests/09 job id: 47165425 --- 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/4047e199d09ff4d3c3dd9bcab791603060e8611a_performance_tests/10 job id: 47165426 --- 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/4047e199d09ff4d3c3dd9bcab791603060e8611a_performance_tests/11 job id: 47165428 --- 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/4047e199d09ff4d3c3dd9bcab791603060e8611a_performance_tests/12 job id: 47165429 --- 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/4047e199d09ff4d3c3dd9bcab791603060e8611a_performance_tests/13 job id: 47165430 --- 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/4047e199d09ff4d3c3dd9bcab791603060e8611a_performance_tests/14 job id: 47165431 --- 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/4047e199d09ff4d3c3dd9bcab791603060e8611a_performance_tests/15 job id: 47165432 --- 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/4047e199d09ff4d3c3dd9bcab791603060e8611a_performance_tests/16 job id: 47165433 --- 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/4047e199d09ff4d3c3dd9bcab791603060e8611a_performance_tests/17 job id: 47165434 --- 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/4047e199d09ff4d3c3dd9bcab791603060e8611a_performance_tests/18 job id: 47165436 --- 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/4047e199d09ff4d3c3dd9bcab791603060e8611a_performance_tests/19 job id: 47165437 --- 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/4047e199d09ff4d3c3dd9bcab791603060e8611a_performance_tests/20 job id: 47165438 --- 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/4047e199d09ff4d3c3dd9bcab791603060e8611a_performance_tests/21 job id: 47165439 --- 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/4047e199d09ff4d3c3dd9bcab791603060e8611a_performance_tests/22 job id: 47165440 --- 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/4047e199d09ff4d3c3dd9bcab791603060e8611a_performance_tests/23 job id: 47165441 --- 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/4047e199d09ff4d3c3dd9bcab791603060e8611a_performance_tests/24 job id: 47165443 --- 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/4047e199d09ff4d3c3dd9bcab791603060e8611a_performance_tests/25 job id: 47165444 --- 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/4047e199d09ff4d3c3dd9bcab791603060e8611a_performance_tests/26 job id: 47165445 --- Point --- name: 510 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/1r/12t) --- Point --- name: 511 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: 512 H2O (4 NVE MD steps on 64 nodes) input file: benchmarks/QS/00512_H2O/H2O-512_md.inp required files: [] output file: result.log # nodes = 64 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4047e199d09ff4d3c3dd9bcab791603060e8611a_performance_tests/27 job id: 47165447 --- Point --- name: 601 plot: h2o_512_md regex: CP2K label: (64n/12r/1t) --- Point --- name: 602 plot: h2o_512_md_mem regex: Estimated peak process memory label: (64n/12r/1t) ~~~~~~~ END TEST ~~~~~~~ === END TESTS (description) === ===== PLOTS (description) ===== ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2_mem", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md_mem", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md_mem", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md_mem", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls_mem", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_512_md", title="512 H2O (4 NVE MD steps on 64 nodes)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_512_md_mem", title="512 H2O (4 NVE MD steps on 64 nodes)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" === END PLOTS (description) === ============ RESULTS ============ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4047e199d09ff4d3c3dd9bcab791603060e8611a_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 424 8. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.025 0.040 137.584 137.585 farming_run 1 2.0 136.935 136.936 137.539 137.542 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.468760E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 230 1103589. MP_Allreduce 571 1938539. MP_Sync 25 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 10 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.008 0.026 116.177 116.177 qs_energies 1 2.0 0.000 0.000 115.858 115.860 mp2_main 1 3.0 0.000 0.000 113.096 113.098 mp2_gpw_main 1 4.0 0.019 0.025 111.756 111.758 mp2_ri_gpw_compute_in 1 5.0 0.171 0.172 92.382 92.792 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 54.886 55.296 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.168 41.318 46.520 get_2c_integrals 1 6.0 0.008 0.009 36.726 37.323 integrate_v_rspace 273 8.0 0.434 0.452 24.732 29.822 pw_transfer 6555 10.6 0.373 0.385 27.368 27.895 fft_wrap_pw1pw2 5465 11.4 0.046 0.048 26.043 26.575 grid_integrate_task_list 273 9.0 20.540 26.121 20.540 26.121 fft_wrap_pw1pw2_100 2178 12.4 1.166 1.220 23.568 24.096 rpa_ri_compute_en 1 5.0 0.019 0.021 19.265 19.469 compute_2c_integrals 1 7.0 0.002 0.002 19.369 19.370 compute_2c_integrals_loop_lm 1 8.0 0.003 0.003 18.833 18.920 mp2_eri_2c_integrate_gpw 1 9.0 2.390 2.436 18.830 18.916 cp_fm_cholesky_decompose 12 8.2 17.647 18.185 17.647 18.185 cholesky_decomp 1 7.0 0.000 0.000 16.195 16.767 fft3d_s 5443 13.4 16.158 16.572 16.180 16.593 ao_to_mo_and_store_B_mult_1 272 7.0 10.734 15.312 10.734 15.312 calculate_wavefunction 272 8.0 5.394 5.537 12.510 13.183 rpa_num_int 1 6.0 0.000 0.000 10.948 10.948 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.912 10.943 calc_mat_Q 8 8.0 0.000 0.000 9.443 9.535 contract_S_to_Q 8 9.0 0.000 0.000 8.865 8.951 calc_potential_gpw 544 9.5 0.005 0.005 8.227 8.593 parallel_gemm_fm 14 9.1 0.000 0.000 8.452 8.540 parallel_gemm_fm_cosma 14 10.1 8.452 8.540 8.452 8.540 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.181 8.411 potential_pw2rs 545 10.0 0.107 0.109 7.689 8.391 create_integ_mat 1 6.0 0.023 0.027 7.810 7.810 collocate_single_gaussian 272 10.0 0.039 0.042 7.449 7.690 array2fm 1 7.0 0.000 0.000 6.667 7.103 pw_scatter_s 2720 13.7 4.445 4.578 4.445 4.578 pw_gather_s 2722 13.2 3.869 4.233 3.869 4.233 array2fm_buffer_send 1 8.0 2.913 3.161 2.913 3.161 scf_env_do_scf 1 3.0 0.000 0.000 2.365 2.365 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.756958, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2813.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4047e199d09ff4d3c3dd9bcab791603060e8611a_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 205321. MP_Allreduce 424 9. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.029 0.039 396.317 396.318 farming_run 1 2.0 395.639 395.643 396.275 396.278 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827128576 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788821 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.230987E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 743 386399. MP_Allreduce 2021 21391. MP_Sync 37 MP_Alltoall 77 MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.030 209.253 209.253 qs_energies 1 2.0 0.000 0.000 208.971 209.042 scf_env_do_scf 1 3.0 0.000 0.000 106.373 106.373 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.445 105.454 rebuild_ks_matrix 4 6.0 0.000 0.000 105.370 105.378 qs_ks_build_kohn_sham_matrix 4 7.0 0.056 0.064 105.370 105.378 hfx_ks_matrix 4 8.0 0.001 0.001 104.984 104.989 integrate_four_center 4 9.0 0.144 0.457 104.984 104.988 mp2_main 1 3.0 0.000 0.000 102.306 102.378 mp2_gpw_main 1 4.0 0.031 0.046 101.371 101.441 integrate_four_center_main 4 10.0 0.086 0.562 96.629 98.794 integrate_four_center_bin 269 11.0 96.543 98.791 96.543 98.791 init_scf_loop 1 4.0 0.000 0.000 92.162 92.163 mp2_ri_gpw_compute_in 1 5.0 0.064 0.064 74.504 75.561 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.002 55.058 mp2_eri_3c_integrate_gpw 91 7.0 0.145 0.163 41.615 46.821 integrate_v_rspace 95 8.0 0.398 0.562 28.002 33.019 pw_transfer 2240 10.6 0.145 0.184 29.944 30.351 fft_wrap_pw1pw2 1868 11.4 0.018 0.023 28.947 29.326 grid_integrate_task_list 95 9.0 23.289 28.516 23.289 28.516 mp2_ri_gpw_compute_en 1 5.0 0.064 0.130 26.718 28.431 ao_to_mo_and_store_B_mult_1 91 7.0 10.705 28.270 10.705 28.270 fft_wrap_pw1pw2_100 730 12.4 1.285 1.502 26.653 27.069 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.836 1.900 25.008 25.018 get_2c_integrals 1 6.0 0.000 0.000 20.409 20.436 compute_2c_integrals 1 7.0 0.002 0.003 19.386 19.390 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 18.895 19.257 mp2_eri_2c_integrate_gpw 1 9.0 1.736 1.896 18.894 19.256 fft3d_s 1823 13.4 18.435 18.725 18.449 18.738 scf_env_do_scf_inner_loop 4 4.0 0.002 0.074 14.208 14.209 calculate_wavefunction 91 8.0 2.010 2.044 9.730 9.942 mp2_ri_gpw_compute_en_expansio 172 7.0 0.559 0.596 8.714 9.405 potential_pw2rs 186 10.0 0.033 0.035 8.649 9.283 local_gemm 172 8.0 8.155 8.828 8.155 8.828 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.263 8.641 mp2_ri_gpw_compute_en_comm 22 7.0 0.499 0.520 8.059 8.503 calc_potential_gpw 182 9.5 0.002 0.003 7.923 8.158 collocate_single_gaussian 91 10.0 0.017 0.020 7.900 8.147 mp_sync 37 10.5 3.011 6.979 3.011 6.979 mp_sendrecv_dm3 2068 8.0 6.094 6.526 6.094 6.526 mp2_ri_gpw_compute_en_ener 172 7.0 6.354 6.462 6.354 6.462 pw_gather_s 912 13.2 4.918 5.436 4.918 5.436 pw_scatter_s 910 13.7 3.913 4.199 3.913 4.199 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.357983, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1518.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4047e199d09ff4d3c3dd9bcab791603060e8611a_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 452.018176E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62385. MP_Allreduce 10329 270. MP_Sync 530 MP_Alltoall 2083 444348. MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.027 54.463 54.464 qs_mol_dyn_low 1 2.0 0.003 0.007 53.953 54.240 qs_forces 11 3.9 0.002 0.003 53.657 53.659 qs_energies 11 4.9 0.001 0.001 52.167 52.199 scf_env_do_scf 11 5.9 0.000 0.001 46.116 46.116 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 44.072 44.073 qs_scf_new_mos 108 7.5 0.000 0.001 33.946 34.215 qs_scf_loop_do_ot 108 8.5 0.000 0.001 33.945 34.214 dbcsr_multiply_generic 2286 12.5 0.096 0.100 33.748 34.190 ot_scf_mini 108 9.5 0.002 0.002 32.290 32.448 multiply_cannon 2286 13.5 0.179 0.187 26.060 27.669 multiply_cannon_loop 2286 14.5 1.464 1.581 25.400 27.023 velocity_verlet 10 3.0 0.001 0.002 26.358 26.361 ot_mini 108 10.5 0.001 0.001 19.581 19.824 qs_ot_get_derivative 108 11.5 0.001 0.001 16.619 16.799 mp_waitall_1 245248 16.5 8.428 14.530 8.428 14.530 multiply_cannon_metrocomm3 54864 15.5 0.067 0.075 6.017 12.842 multiply_cannon_multrec 54864 15.5 4.264 6.570 7.761 11.223 qs_ot_get_p 119 10.4 0.002 0.005 8.038 8.330 rebuild_ks_matrix 119 8.3 0.000 0.000 7.893 8.024 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.012 7.893 8.023 mp_sum_l 7287 12.8 5.724 7.319 5.724 7.319 multiply_cannon_sync_h2d 54864 15.5 5.919 7.206 5.919 7.206 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.939 7.057 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.650 6.113 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.426 5.530 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.315 5.362 dbcsr_mm_accdrv_process 76910 16.1 1.133 1.848 3.420 4.906 init_scf_run 11 5.9 0.000 0.001 4.764 4.764 scf_env_initial_rho_setup 11 6.9 0.001 0.002 4.764 4.764 sum_up_and_integrate 119 10.3 0.012 0.014 4.527 4.534 integrate_v_rspace 119 11.3 0.002 0.002 4.515 4.523 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.486 4.486 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.232 4.347 calculate_rho_elec 119 8.7 0.012 0.017 4.231 4.346 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.288 4.288 cp_fm_redistribute_end 50 14.0 2.187 4.262 2.192 4.265 cp_fm_diag_elpa_base 50 14.0 2.067 4.153 2.071 4.161 calculate_dm_sparse 119 9.5 0.000 0.001 2.941 3.084 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.892 3.069 apply_single 119 13.6 0.000 0.000 2.892 3.069 multiply_cannon_metrocomm1 54864 15.5 0.053 0.058 1.751 3.003 jit_kernel_multiply 13 15.8 2.226 2.985 2.226 2.985 rs_pw_transfer 974 11.9 0.011 0.013 2.711 2.812 ot_diis_step 108 11.5 0.006 0.006 2.675 2.676 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.479 2.568 calculate_first_density_matrix 1 7.0 0.000 0.001 2.538 2.542 density_rs2pw 119 9.7 0.004 0.005 2.193 2.311 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.305 2.307 acc_transpose_blocks 54864 15.5 0.225 0.251 1.746 2.237 wfi_extrapolate 11 7.9 0.001 0.001 2.155 2.155 grid_integrate_task_list 119 12.3 2.023 2.148 2.023 2.148 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.022 2.074 init_scf_loop 11 6.9 0.000 0.001 2.026 2.026 mp_sum_d 4135 12.0 1.318 1.900 1.318 1.900 potential_pw2rs 119 12.3 0.004 0.004 1.859 1.875 pw_transfer 1439 11.6 0.052 0.058 1.734 1.805 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.659 1.729 make_m2s 4572 13.5 0.053 0.055 1.643 1.694 make_images 4572 14.5 0.133 0.139 1.560 1.611 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.453 1.483 grid_collocate_task_list 119 9.7 1.360 1.442 1.360 1.442 mp_alltoall_d11v 2130 13.8 1.276 1.426 1.276 1.426 mp_waitany 12084 13.8 1.257 1.374 1.257 1.374 fft3d_ps 1201 14.6 0.375 0.478 1.301 1.368 fft_wrap_pw1pw2_140 487 13.2 0.185 0.201 1.275 1.345 dbcsr_dot_sd 1205 11.9 0.049 0.060 0.725 1.092 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=54.464000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=431.181818, yerr=1.113404 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4047e199d09ff4d3c3dd9bcab791603060e8611a_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 486.363136E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2194560 MPI messages size (bytes): total size 310.646604E+09 min size 0.000000E+00 max size 1.145520E+06 average size 141.553031E+03 MPI breakdown and total messages size (bytes): size <= 128 724648 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 281952 4619501568 32768 < size <= 131072 494448 39143342080 131072 < size <= 4194304 440000 264807943488 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62664. MP_Allreduce 10306 303. MP_Sync 54 MP_Alltoall 2060 1847534. MP_SendRecv 16779 37093. MP_ISendRecv 16779 37093. MP_Wait 23539 MP_comm_split 50 MP_ISend 5720 128509. MP_IRecv 5720 128509. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.029 39.886 39.887 qs_mol_dyn_low 1 2.0 0.003 0.003 39.701 39.709 qs_forces 11 3.9 0.002 0.003 39.025 39.026 qs_energies 11 4.9 0.001 0.001 37.315 37.319 scf_env_do_scf 11 5.9 0.000 0.001 32.012 32.013 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 29.536 29.536 dbcsr_multiply_generic 2286 12.5 0.102 0.105 21.651 22.013 qs_scf_new_mos 108 7.5 0.001 0.001 20.134 20.379 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.134 20.378 ot_scf_mini 108 9.5 0.002 0.003 19.234 19.408 velocity_verlet 10 3.0 0.001 0.002 19.066 19.080 multiply_cannon 2286 13.5 0.206 0.216 16.630 18.301 multiply_cannon_loop 2286 14.5 0.905 0.975 15.508 16.861 ot_mini 108 10.5 0.001 0.001 11.864 12.103 mp_waitall_1 200699 16.5 5.788 10.992 5.788 10.992 multiply_cannon_metrocomm3 27432 15.5 0.066 0.068 4.192 9.653 qs_ot_get_derivative 108 11.5 0.001 0.001 9.408 9.587 multiply_cannon_multrec 27432 15.5 1.998 4.208 6.070 8.918 rebuild_ks_matrix 119 8.3 0.000 0.000 7.314 7.448 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 7.314 7.447 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.450 6.571 dbcsr_mm_accdrv_process 47894 16.0 3.090 5.308 4.002 5.908 qs_ot_get_p 119 10.4 0.001 0.001 4.456 4.693 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.678 4.550 sum_up_and_integrate 119 10.3 0.025 0.027 4.300 4.307 integrate_v_rspace 119 11.3 0.002 0.002 4.275 4.284 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.070 4.146 apply_single 119 13.6 0.000 0.000 3.070 4.146 init_scf_run 11 5.9 0.000 0.001 4.099 4.100 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.099 4.099 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.951 3.984 calculate_rho_elec 119 8.7 0.021 0.024 3.950 3.984 mp_sum_l 7287 12.8 2.029 3.972 2.029 3.972 rs_pw_transfer 974 11.9 0.010 0.011 2.765 3.258 qs_ot_p2m_diag 50 11.0 0.008 0.012 3.010 3.028 multiply_cannon_sync_h2d 27432 15.5 2.189 2.820 2.189 2.820 make_m2s 4572 13.5 0.052 0.054 2.535 2.763 density_rs2pw 119 9.7 0.004 0.004 2.229 2.744 make_images 4572 14.5 0.201 0.238 2.445 2.671 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.571 2.571 calculate_first_density_matrix 1 7.0 0.000 0.002 2.531 2.532 init_scf_loop 11 6.9 0.000 0.001 2.454 2.455 ot_diis_step 108 11.5 0.011 0.011 2.407 2.408 jit_kernel_multiply 11 16.1 0.858 2.289 0.858 2.289 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.238 2.238 calculate_dm_sparse 119 9.5 0.000 0.001 2.155 2.234 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.141 2.233 cp_fm_redistribute_end 50 14.0 1.137 2.213 1.140 2.215 cp_fm_diag_elpa_base 50 14.0 1.043 2.115 1.071 2.154 potential_pw2rs 119 12.3 0.006 0.006 2.013 2.023 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.009 2.011 pw_transfer 1439 11.6 0.066 0.070 1.959 1.991 grid_integrate_task_list 119 12.3 1.832 1.953 1.832 1.953 fft_wrap_pw1pw2 1201 12.6 0.008 0.009 1.867 1.902 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.797 1.837 make_images_data 4572 15.5 0.044 0.050 1.160 1.599 prepare_preconditioner 11 7.9 0.000 0.000 1.547 1.574 make_preconditioner 11 8.9 0.000 0.000 1.547 1.574 acc_transpose_blocks 27432 15.5 0.110 0.116 1.253 1.573 hybrid_alltoall_any 4725 16.4 0.051 0.112 1.006 1.522 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.447 1.506 wfi_extrapolate 11 7.9 0.001 0.001 1.503 1.503 fft_wrap_pw1pw2_140 487 13.2 0.201 0.212 1.463 1.500 fft3d_ps 1201 14.6 0.524 0.580 1.419 1.448 grid_collocate_task_list 119 9.7 1.271 1.423 1.271 1.423 mp_alltoall_d11v 2130 13.8 1.210 1.375 1.210 1.375 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.363 1.372 mp_allgather_i34 2286 14.5 0.565 1.331 0.565 1.331 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.182 1.229 rs_pw_transfer_RS2PW_140 130 11.5 0.138 0.146 0.542 1.040 mp_waitany 5720 13.7 0.523 1.028 0.523 1.028 mp_sum_d 4135 12.0 0.561 1.000 0.561 1.000 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.921 0.936 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.935 0.935 acc_transpose_blocks_kernels 27432 16.5 0.183 0.272 0.702 0.928 rs_pw_transfer_PW2RS_50 119 14.3 0.588 0.607 0.780 0.846 make_images_sizes 4572 15.5 0.005 0.005 0.554 0.809 mp_alltoall_i44 4572 16.5 0.550 0.805 0.550 0.805 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=39.887000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=462.636364, yerr=1.666391 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4047e199d09ff4d3c3dd9bcab791603060e8611a_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 521.961472E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63497. MP_Allreduce 10155 305. MP_Sync 54 MP_Alltoall 1821 1607886. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.020 0.034 36.739 36.740 qs_mol_dyn_low 1 2.0 0.003 0.005 36.475 36.482 qs_forces 11 3.9 0.004 0.014 36.416 36.417 qs_energies 11 4.9 0.002 0.005 34.710 34.713 scf_env_do_scf 11 5.9 0.001 0.003 29.509 29.509 scf_env_do_scf_inner_loop 108 6.5 0.002 0.008 26.451 26.452 velocity_verlet 10 3.0 0.001 0.002 17.561 17.562 dbcsr_multiply_generic 2286 12.5 0.097 0.099 17.435 17.513 qs_scf_new_mos 108 7.5 0.001 0.001 16.594 16.606 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.593 16.605 ot_scf_mini 108 9.5 0.002 0.003 15.837 15.847 multiply_cannon 2286 13.5 0.195 0.198 13.405 14.111 multiply_cannon_loop 2286 14.5 0.637 0.667 12.486 13.232 ot_mini 108 10.5 0.001 0.001 9.468 9.482 qs_ot_get_derivative 108 11.5 0.001 0.002 7.913 7.924 rebuild_ks_matrix 119 8.3 0.000 0.000 7.342 7.361 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.015 7.341 7.360 multiply_cannon_multrec 18288 15.5 1.960 2.918 6.867 7.197 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.491 6.507 dbcsr_mm_accdrv_process 38222 16.0 4.048 5.443 4.822 5.514 sum_up_and_integrate 119 10.3 0.030 0.032 4.832 4.835 integrate_v_rspace 119 11.3 0.003 0.004 4.801 4.807 mp_waitall_1 158411 16.6 3.613 4.549 3.613 4.549 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.439 4.443 calculate_rho_elec 119 8.7 0.031 0.032 4.438 4.443 qs_ot_get_p 119 10.4 0.001 0.001 4.078 4.093 init_scf_run 11 5.9 0.000 0.001 3.911 3.911 scf_env_initial_rho_setup 11 6.9 0.001 0.002 3.911 3.911 rs_pw_transfer 974 11.9 0.009 0.010 3.630 3.901 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.954 3.586 density_rs2pw 119 9.7 0.004 0.004 2.895 3.159 init_scf_loop 11 6.9 0.000 0.002 3.035 3.035 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.756 2.762 make_m2s 4572 13.5 0.044 0.045 2.415 2.582 potential_pw2rs 119 12.3 0.007 0.008 2.524 2.543 make_images 4572 14.5 0.192 0.206 2.329 2.493 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.455 2.456 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.091 2.418 apply_single 119 13.6 0.000 0.000 2.091 2.418 pw_transfer 1439 11.6 0.065 0.069 2.347 2.376 calculate_first_density_matrix 1 7.0 0.000 0.002 2.365 2.366 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.254 2.281 prepare_preconditioner 11 7.9 0.000 0.000 2.154 2.155 make_preconditioner 11 8.9 0.001 0.005 2.154 2.155 jit_kernel_multiply 11 15.8 0.722 2.135 0.722 2.135 multiply_cannon_metrocomm3 18288 15.5 0.043 0.044 1.418 2.129 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.113 2.123 cp_fm_diag_elpa_base 50 14.0 2.086 2.099 2.112 2.122 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.986 2.081 calculate_dm_sparse 119 9.5 0.000 0.001 1.962 1.971 grid_integrate_task_list 119 12.3 1.782 1.877 1.782 1.877 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.809 1.816 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.802 1.810 fft3d_ps 1201 14.6 0.535 0.554 1.743 1.776 fft_wrap_pw1pw2_140 487 13.2 0.251 0.258 1.610 1.635 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.566 1.571 multiply_cannon_sync_h2d 18288 15.5 1.369 1.552 1.369 1.552 mp_sum_l 7287 12.8 1.225 1.538 1.225 1.538 ot_diis_step 108 11.5 0.012 0.014 1.532 1.532 wfi_extrapolate 11 7.9 0.008 0.056 1.471 1.472 grid_collocate_task_list 119 9.7 1.241 1.374 1.241 1.374 make_images_data 4572 15.5 0.044 0.048 1.075 1.338 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.313 1.316 acc_transpose_blocks 18288 15.5 0.077 0.079 1.256 1.278 hybrid_alltoall_any 4725 16.4 0.055 0.114 0.897 1.154 mp_alltoall_d11v 2130 13.8 0.917 1.086 0.917 1.086 mp_alltoall_z22v 1201 16.6 1.016 1.085 1.016 1.085 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 0.996 0.998 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.949 0.970 mp_sendrecv_dv 11067 12.7 0.936 0.964 0.936 0.964 cp_fm_cholesky_invert 11 10.9 0.958 0.963 0.958 0.963 make_images_sizes 4572 15.5 0.005 0.005 0.652 0.940 mp_alltoall_i44 4572 16.5 0.648 0.935 0.648 0.935 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.924 0.927 rs_pw_transfer_PW2RS_50 119 14.3 0.401 0.411 0.786 0.877 mp_waitany 9880 13.7 0.597 0.874 0.597 0.874 rs_pw_transfer_RS2PW_140 130 11.5 0.118 0.121 0.581 0.847 acc_transpose_blocks_kernels 18288 16.5 0.212 0.222 0.800 0.816 rs_pw_transfer_RS2PW_50 119 11.7 0.242 0.251 0.734 0.811 mp_allgather_i34 2286 14.5 0.379 0.772 0.379 0.772 dbcsr_complete_redistribute 329 12.2 0.092 0.098 0.643 0.760 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=36.740000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.181818, yerr=1.945540 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4047e199d09ff4d3c3dd9bcab791603060e8611a_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 544.485376E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63496. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.020 0.032 42.181 42.182 qs_mol_dyn_low 1 2.0 0.003 0.004 41.933 41.942 qs_forces 11 3.9 0.002 0.002 41.869 41.869 qs_energies 11 4.9 0.001 0.001 39.972 39.979 scf_env_do_scf 11 5.9 0.000 0.001 34.286 34.287 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 29.725 29.726 velocity_verlet 10 3.0 0.001 0.001 21.859 21.861 dbcsr_multiply_generic 2286 12.5 0.101 0.104 20.164 20.296 qs_scf_new_mos 108 7.5 0.001 0.001 19.341 19.396 qs_scf_loop_do_ot 108 8.5 0.001 0.001 19.340 19.395 ot_scf_mini 108 9.5 0.002 0.002 18.320 18.379 multiply_cannon 2286 13.5 0.224 0.256 15.179 15.757 multiply_cannon_loop 2286 14.5 0.948 0.979 13.932 14.393 ot_mini 108 10.5 0.001 0.001 10.673 10.731 multiply_cannon_multrec 27432 15.5 2.389 3.083 8.847 9.206 qs_ot_get_derivative 108 11.5 0.001 0.001 8.743 8.802 rebuild_ks_matrix 119 8.3 0.000 0.000 7.589 7.657 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.012 7.589 7.657 dbcsr_mm_accdrv_process 47916 15.9 5.549 7.115 6.370 7.518 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.728 6.786 qs_ot_get_p 119 10.4 0.001 0.001 4.742 4.806 sum_up_and_integrate 119 10.3 0.035 0.037 4.696 4.705 integrate_v_rspace 119 11.3 0.003 0.003 4.661 4.671 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.541 4.569 calculate_rho_elec 119 8.7 0.040 0.046 4.541 4.569 init_scf_loop 11 6.9 0.000 0.000 4.526 4.526 init_scf_run 11 5.9 0.000 0.001 4.243 4.243 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.243 4.243 mp_waitall_1 137007 16.6 3.092 3.880 3.092 3.880 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.224 3.684 rs_pw_transfer 974 11.9 0.009 0.009 3.399 3.587 prepare_preconditioner 11 7.9 0.000 0.000 3.549 3.556 make_preconditioner 11 8.9 0.000 0.000 3.549 3.556 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.116 3.466 make_m2s 4572 13.5 0.054 0.055 3.098 3.248 make_images 4572 14.5 0.271 0.332 2.990 3.137 density_rs2pw 119 9.7 0.004 0.004 2.893 3.076 qs_ot_p2m_diag 50 11.0 0.015 0.022 3.034 3.048 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.656 2.657 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.177 2.648 apply_single 119 13.6 0.000 0.000 2.177 2.647 pw_transfer 1439 11.6 0.066 0.070 2.428 2.468 calculate_first_density_matrix 1 7.0 0.000 0.000 2.456 2.459 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.376 2.409 potential_pw2rs 119 12.3 0.009 0.009 2.392 2.404 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.335 2.378 calculate_dm_sparse 119 9.5 0.000 0.001 2.272 2.328 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.246 2.257 cp_fm_diag_elpa_base 50 14.0 2.201 2.222 2.243 2.254 mp_sum_l 7287 12.8 1.461 2.132 1.461 2.132 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.987 1.989 grid_integrate_task_list 119 12.3 1.808 1.889 1.808 1.889 ot_diis_step 108 11.5 0.012 0.012 1.877 1.877 fft3d_ps 1201 14.6 0.557 0.609 1.776 1.810 fft_wrap_pw1pw2_140 487 13.2 0.286 0.298 1.754 1.801 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.718 1.740 wfi_extrapolate 11 7.9 0.001 0.001 1.714 1.714 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.887 1.598 make_images_data 4572 15.5 0.044 0.047 1.322 1.546 acc_transpose_blocks 27432 15.5 0.112 0.114 1.496 1.526 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.394 1.405 grid_collocate_task_list 119 9.7 1.247 1.372 1.247 1.372 dbcsr_complete_redistribute 329 12.2 0.123 0.153 1.052 1.349 cp_fm_cholesky_invert 11 10.9 1.337 1.342 1.337 1.342 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.311 1.327 hybrid_alltoall_any 4725 16.4 0.062 0.151 1.111 1.321 cp_fm_upper_to_full 72 14.2 0.868 1.245 0.868 1.245 jit_kernel_multiply 6 15.9 0.758 1.226 0.758 1.226 mp_alltoall_d11v 2130 13.8 1.069 1.206 1.069 1.206 mp_alltoall_z22v 1201 16.6 1.103 1.142 1.103 1.142 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.098 1.098 multiply_cannon_sync_h2d 27432 15.5 0.954 1.061 0.954 1.061 make_images_sizes 4572 15.5 0.005 0.005 0.751 1.025 mp_alltoall_i44 4572 16.5 0.746 1.020 0.746 1.020 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.710 1.002 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.955 0.965 mp_sendrecv_dv 8211 12.7 0.935 0.959 0.935 0.959 acc_transpose_blocks_kernels 27432 16.5 0.271 0.294 0.848 0.873 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.794 0.871 mp_alltoall_i22 627 13.8 0.517 0.844 0.517 0.844 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=42.182000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=518.545455, yerr=1.924183 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4047e199d09ff4d3c3dd9bcab791603060e8611a_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 599.416832E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63495. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.029 29.543 29.544 qs_mol_dyn_low 1 2.0 0.003 0.004 29.088 29.096 qs_forces 11 3.9 0.002 0.003 28.822 28.823 qs_energies 11 4.9 0.001 0.001 27.099 27.101 scf_env_do_scf 11 5.9 0.000 0.001 21.821 21.821 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 19.295 19.296 velocity_verlet 10 3.0 0.001 0.001 14.660 14.676 dbcsr_multiply_generic 2286 12.5 0.094 0.099 12.877 13.044 qs_scf_new_mos 108 7.5 0.001 0.001 11.235 11.258 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.235 11.257 multiply_cannon 2286 13.5 0.230 0.241 10.349 10.915 ot_scf_mini 108 9.5 0.002 0.002 10.560 10.585 multiply_cannon_loop 2286 14.5 0.332 0.342 9.451 9.713 multiply_cannon_multrec 9144 15.5 1.714 2.017 6.183 6.423 rebuild_ks_matrix 119 8.3 0.000 0.000 6.014 6.034 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.013 6.034 ot_mini 108 10.5 0.001 0.001 5.977 6.006 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.348 5.366 qs_ot_get_derivative 108 11.5 0.001 0.001 4.686 4.710 dbcsr_mm_accdrv_process 12550 15.8 3.062 4.375 4.369 4.471 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.627 3.864 calculate_rho_elec 119 8.7 0.060 0.061 3.626 3.863 init_scf_run 11 5.9 0.000 0.001 3.852 3.852 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.852 3.852 sum_up_and_integrate 119 10.3 0.037 0.041 3.696 3.700 integrate_v_rspace 119 11.3 0.003 0.003 3.658 3.663 qs_ot_get_p 119 10.4 0.001 0.001 2.806 2.844 calculate_first_density_matrix 1 7.0 0.000 0.000 2.659 2.693 init_scf_loop 11 6.9 0.000 0.000 2.505 2.506 mp_waitall_1 115863 16.7 1.762 2.374 1.762 2.374 calculate_dm_sparse 119 9.5 0.000 0.000 2.122 2.175 jit_kernel_multiply 10 15.6 1.269 2.100 1.269 2.100 pw_transfer 1439 11.6 0.066 0.069 2.076 2.086 density_rs2pw 119 9.7 0.004 0.004 1.877 2.020 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.982 1.992 grid_integrate_task_list 119 12.3 1.851 1.929 1.851 1.929 make_m2s 4572 13.5 0.034 0.035 1.750 1.904 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.868 1.870 rs_pw_transfer 974 11.9 0.008 0.008 1.710 1.853 make_images 4572 14.5 0.268 0.304 1.659 1.812 prepare_preconditioner 11 7.9 0.000 0.000 1.763 1.767 make_preconditioner 11 8.9 0.000 0.000 1.763 1.767 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.685 1.687 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.651 1.679 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.670 1.670 fft_wrap_pw1pw2_140 487 13.2 0.364 0.372 1.617 1.629 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.540 1.580 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.556 1.569 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.519 1.526 potential_pw2rs 119 12.3 0.010 0.011 1.417 1.421 grid_collocate_task_list 119 9.7 1.297 1.412 1.297 1.412 acc_transpose_blocks 9144 15.5 0.039 0.040 0.983 1.399 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.377 1.385 cp_fm_diag_elpa_base 50 14.0 1.349 1.367 1.375 1.383 fft3d_ps 1201 14.6 0.561 0.572 1.316 1.324 ot_diis_step 108 11.5 0.013 0.013 1.277 1.278 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.216 1.216 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.148 1.168 apply_single 119 13.6 0.000 0.000 1.148 1.167 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.154 1.166 acc_transpose_blocks_kernels 9144 16.5 0.117 0.120 0.734 1.152 mp_alltoall_d11v 2130 13.8 0.821 1.126 0.821 1.126 wfi_extrapolate 11 7.9 0.001 0.001 1.105 1.105 jit_kernel_transpose 5 15.6 0.617 1.038 0.617 1.038 hybrid_alltoall_any 4725 16.4 0.063 0.177 0.789 1.018 make_images_data 4572 15.5 0.038 0.041 0.811 0.991 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.866 0.915 cp_fm_cholesky_invert 11 10.9 0.893 0.896 0.893 0.896 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.543 0.863 multiply_cannon_metrocomm1 9144 15.5 0.022 0.022 0.406 0.851 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.788 0.790 multiply_cannon_sync_h2d 9144 15.5 0.704 0.787 0.704 0.787 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.719 0.729 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.671 0.723 mp_alltoall_z22v 1201 16.6 0.629 0.657 0.629 0.657 rs_scatter_matrices 130 9.7 0.140 0.151 0.408 0.649 mp_allgather_i34 2286 14.5 0.231 0.633 0.231 0.633 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=29.544000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=569.090909, yerr=3.232071 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4047e199d09ff4d3c3dd9bcab791603060e8611a_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 754.819072E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63729. MP_Allreduce 10154 429. MP_Sync 54 MP_Alltoall 1582 7383731. MP_SendRecv 2499 189067. MP_ISendRecv 2499 189067. MP_Wait 6399 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.033 42.454 42.455 qs_mol_dyn_low 1 2.0 0.003 0.004 42.134 42.141 qs_forces 11 3.9 0.002 0.003 41.865 41.866 qs_energies 11 4.9 0.001 0.001 39.844 39.847 scf_env_do_scf 11 5.9 0.001 0.001 33.528 33.529 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 25.716 25.717 velocity_verlet 10 3.0 0.001 0.001 23.588 23.619 dbcsr_multiply_generic 2286 12.5 0.101 0.104 17.981 18.102 qs_scf_new_mos 108 7.5 0.001 0.001 15.910 16.007 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.910 16.006 multiply_cannon 2286 13.5 0.300 0.311 14.140 14.960 ot_scf_mini 108 9.5 0.002 0.002 14.836 14.932 multiply_cannon_loop 2286 14.5 0.345 0.351 12.909 13.770 multiply_cannon_multrec 9144 15.5 3.566 5.210 8.972 9.157 ot_mini 108 10.5 0.001 0.001 8.881 8.995 init_scf_loop 11 6.9 0.000 0.000 7.784 7.785 rebuild_ks_matrix 119 8.3 0.000 0.000 7.292 7.432 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.292 7.431 qs_ot_get_derivative 108 11.5 0.001 0.001 6.885 6.983 prepare_preconditioner 11 7.9 0.000 0.000 6.832 6.845 make_preconditioner 11 8.9 0.000 0.000 6.832 6.845 dbcsr_mm_accdrv_process 12550 15.8 4.281 5.731 5.287 6.784 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.399 6.720 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.570 6.696 cp_fm_upper_to_full 72 14.2 3.205 4.585 3.205 4.585 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.425 4.539 calculate_rho_elec 119 8.7 0.118 0.121 4.424 4.539 init_scf_run 11 5.9 0.000 0.001 4.276 4.276 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.276 4.276 sum_up_and_integrate 119 10.3 0.064 0.065 4.037 4.044 integrate_v_rspace 119 11.3 0.003 0.003 3.973 3.979 qs_ot_get_p 119 10.4 0.001 0.001 3.241 3.376 mp_waitall_1 94719 16.7 2.484 3.319 2.484 3.319 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.448 2.878 dbcsr_complete_redistribute 329 12.2 0.287 0.293 1.963 2.799 calculate_first_density_matrix 1 7.0 0.000 0.000 2.677 2.795 pw_transfer 1439 11.6 0.069 0.069 2.703 2.710 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.605 2.612 calculate_dm_sparse 119 9.5 0.000 0.000 2.412 2.526 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.653 2.473 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.172 2.427 apply_single 119 13.6 0.000 0.000 2.172 2.427 make_m2s 4572 13.5 0.037 0.037 2.256 2.407 make_images 4572 14.5 0.352 0.382 2.134 2.284 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.428 2.239 mp_alltoall_i22 627 13.8 1.406 2.229 1.406 2.229 density_rs2pw 119 9.7 0.004 0.004 2.212 2.227 fft_wrap_pw1pw2_140 487 13.2 0.615 0.619 2.177 2.186 multiply_cannon_metrocomm3 9144 15.5 0.019 0.020 1.354 2.166 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.100 2.100 grid_integrate_task_list 119 12.3 2.062 2.071 2.062 2.071 ot_diis_step 108 11.5 0.014 0.015 1.974 1.974 qs_ot_p2m_diag 50 11.0 0.042 0.043 1.962 1.963 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.843 1.888 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.761 1.763 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.680 1.729 mp_sum_l 7287 12.8 1.037 1.714 1.037 1.714 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.687 1.687 fft3d_ps 1201 14.6 0.595 0.605 1.623 1.627 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.486 1.599 rs_pw_transfer 974 11.9 0.009 0.009 1.517 1.538 grid_collocate_task_list 119 9.7 1.510 1.519 1.510 1.519 potential_pw2rs 119 12.3 0.014 0.014 1.433 1.436 cp_fm_cholesky_invert 11 10.9 1.425 1.428 1.425 1.428 wfi_extrapolate 11 7.9 0.001 0.001 1.417 1.417 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.396 1.397 cp_fm_diag_elpa_base 50 14.0 1.245 1.305 1.395 1.395 hybrid_alltoall_any 4725 16.4 0.087 0.147 1.109 1.337 jit_kernel_multiply 6 15.6 0.979 1.302 0.979 1.302 mp_alltoall_d11v 2130 13.8 1.179 1.297 1.179 1.297 make_images_data 4572 15.5 0.042 0.044 1.046 1.254 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.131 1.151 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.094 1.116 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.988 1.045 multiply_cannon_sync_h2d 9144 15.5 1.037 1.045 1.037 1.045 acc_transpose_blocks 9144 15.5 0.039 0.040 0.946 0.950 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.933 0.947 qs_create_task_list 11 7.9 0.001 0.001 0.932 0.945 generate_qs_task_list 11 8.9 0.366 0.386 0.931 0.944 mp_alltoall_z22v 1201 16.6 0.893 0.915 0.893 0.915 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=42.455000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=708.000000, yerr=12.749332 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4047e199d09ff4d3c3dd9bcab791603060e8611a_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 501.559296E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66218. MP_Allreduce 9776 488. MP_Sync 52 MP_Alltoall 1938 926028. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.027 82.293 82.294 qs_mol_dyn_low 1 2.0 0.003 0.003 82.041 82.051 qs_forces 11 3.9 0.003 0.003 81.675 81.676 qs_energies 11 4.9 0.001 0.001 78.773 78.792 scf_env_do_scf 11 5.9 0.000 0.001 69.873 69.876 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 64.284 64.284 dbcsr_multiply_generic 2055 12.4 0.110 0.114 50.563 50.739 qs_scf_new_mos 99 7.5 0.000 0.001 46.739 46.860 qs_scf_loop_do_ot 99 8.5 0.001 0.001 46.738 46.860 ot_scf_mini 99 9.5 0.002 0.002 44.378 44.482 multiply_cannon 2055 13.4 0.177 0.183 42.146 43.029 velocity_verlet 10 3.0 0.001 0.002 42.799 42.803 multiply_cannon_loop 2055 14.4 1.514 1.559 41.184 41.988 ot_mini 99 10.5 0.001 0.001 26.141 26.228 qs_ot_get_derivative 99 11.5 0.001 0.001 19.427 19.510 multiply_cannon_multrec 49320 15.4 12.479 13.149 17.421 18.176 rebuild_ks_matrix 110 8.3 0.000 0.000 14.709 14.807 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.708 14.806 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.869 12.954 mp_waitall_1 220248 16.4 10.524 11.349 10.524 11.349 multiply_cannon_sync_h2d 49320 15.4 10.366 10.948 10.366 10.948 qs_ot_get_p 110 10.4 0.001 0.001 9.757 9.885 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.224 7.751 apply_single 110 13.6 0.000 0.001 7.224 7.751 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.184 7.737 multiply_cannon_metrocomm3 49320 15.4 0.078 0.080 6.372 7.497 sum_up_and_integrate 110 10.3 0.036 0.042 7.179 7.196 integrate_v_rspace 110 11.3 0.003 0.003 7.144 7.169 init_scf_run 11 5.9 0.000 0.001 6.860 6.861 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.860 6.860 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.731 6.854 calculate_rho_elec 110 8.6 0.021 0.026 6.730 6.853 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.740 6.775 ot_diis_step 99 11.5 0.006 0.006 6.536 6.537 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.835 5.836 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.620 5.668 init_scf_loop 11 6.9 0.000 0.001 5.561 5.562 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.262 5.286 cp_fm_diag_elpa_base 48 14.0 5.248 5.275 5.260 5.284 dbcsr_mm_accdrv_process 87628 16.1 1.934 2.020 4.818 5.112 mp_sum_l 6594 12.7 4.061 4.869 4.061 4.869 rs_pw_transfer 902 11.9 0.011 0.013 3.759 4.263 wfi_extrapolate 11 7.9 0.001 0.001 4.079 4.079 make_m2s 4110 13.4 0.061 0.065 3.887 4.001 density_rs2pw 110 9.6 0.004 0.005 3.483 3.979 make_images 4110 14.4 0.178 0.191 3.791 3.908 calculate_dm_sparse 110 9.5 0.001 0.001 3.759 3.850 multiply_cannon_metrocomm1 49320 15.4 0.062 0.065 2.290 3.653 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.558 3.562 grid_integrate_task_list 110 12.3 3.259 3.413 3.259 3.413 prepare_preconditioner 11 7.9 0.000 0.000 3.388 3.412 make_preconditioner 11 8.9 0.000 0.000 3.388 3.412 pw_transfer 1331 11.6 0.055 0.068 3.233 3.317 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.223 3.255 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.144 3.233 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.176 3.229 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.173 3.228 fft_wrap_pw1pw2_140 451 13.1 0.452 0.496 2.674 2.767 potential_pw2rs 110 12.3 0.006 0.007 2.664 2.689 calculate_first_density_matrix 1 7.0 0.000 0.001 2.674 2.681 jit_kernel_multiply 13 15.9 2.607 2.636 2.607 2.636 mp_alltoall_d11v 2046 13.8 2.065 2.522 2.065 2.522 mp_waitany 14300 13.8 1.845 2.446 1.845 2.446 fft3d_ps 1111 14.6 0.796 0.879 2.339 2.402 grid_collocate_task_list 110 9.6 2.162 2.381 2.162 2.381 acc_transpose_blocks 49320 15.4 0.210 0.227 2.087 2.173 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.970 1.983 make_images_data 4110 15.4 0.042 0.046 1.766 1.879 cp_fm_cholesky_invert 11 10.9 1.828 1.832 1.828 1.832 hybrid_alltoall_any 4261 16.3 0.082 0.482 1.533 1.809 mp_sum_d 3889 11.9 1.248 1.739 1.248 1.739 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.675 1.704 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=82.294000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=475.454545, yerr=2.965365 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4047e199d09ff4d3c3dd9bcab791603060e8611a_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 589.438976E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1972800 MPI messages size (bytes): total size 1.077520E+12 min size 0.000000E+00 max size 4.537280E+06 average size 546.188250E+03 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 222984 1826684928 8192 < size <= 32768 520356 13399818240 32768 < size <= 131072 372336 35386294272 131072 < size <= 4194304 787758 788321309808 4194304 < size <= 16777216 54450 238588003280 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66437. MP_Allreduce 9775 566. MP_Sync 52 MP_Alltoall 1717 2882955. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.022 0.036 74.490 74.491 qs_mol_dyn_low 1 2.0 0.003 0.004 74.177 74.186 qs_forces 11 3.9 0.004 0.009 74.021 74.023 qs_energies 11 4.9 0.002 0.007 70.584 70.590 scf_env_do_scf 11 5.9 0.001 0.002 61.507 61.510 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 52.773 52.775 dbcsr_multiply_generic 2055 12.4 0.118 0.122 38.630 38.854 velocity_verlet 10 3.0 0.001 0.001 38.530 38.533 qs_scf_new_mos 99 7.5 0.001 0.001 35.193 35.350 qs_scf_loop_do_ot 99 8.5 0.001 0.001 35.192 35.350 ot_scf_mini 99 9.5 0.003 0.004 33.508 33.646 multiply_cannon 2055 13.4 0.221 0.245 31.304 32.604 multiply_cannon_loop 2055 14.4 0.924 0.947 29.878 30.863 ot_mini 99 10.5 0.001 0.001 18.919 19.065 multiply_cannon_multrec 24660 15.4 7.611 9.344 14.016 15.827 rebuild_ks_matrix 110 8.3 0.000 0.000 14.514 14.656 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.016 14.513 14.656 qs_ot_get_derivative 99 11.5 0.001 0.001 13.080 13.221 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.775 12.893 mp_waitall_1 176588 16.5 8.211 10.620 8.211 10.620 init_scf_loop 11 6.9 0.001 0.005 8.683 8.687 multiply_cannon_sync_h2d 24660 15.4 7.004 8.061 7.004 8.061 qs_ot_get_p 110 10.4 0.001 0.002 7.868 8.028 multiply_cannon_metrocomm3 24660 15.4 0.067 0.072 5.129 7.900 sum_up_and_integrate 110 10.3 0.053 0.059 7.294 7.305 integrate_v_rspace 110 11.3 0.003 0.003 7.241 7.254 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.538 7.226 apply_single 110 13.6 0.000 0.001 6.538 7.226 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.912 6.921 calculate_rho_elec 110 8.6 0.040 0.048 6.911 6.920 prepare_preconditioner 11 7.9 0.000 0.001 6.573 6.590 make_preconditioner 11 8.9 0.000 0.001 6.572 6.590 dbcsr_mm_accdrv_process 52282 16.1 4.796 5.674 6.244 6.507 init_scf_run 11 5.9 0.000 0.001 6.494 6.495 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.494 6.494 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.140 6.308 ot_diis_step 99 11.5 0.010 0.011 5.788 5.789 qs_ot_p2m_diag 48 11.0 0.028 0.043 5.705 5.725 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.840 5.531 make_m2s 4110 13.4 0.057 0.059 4.782 5.270 cp_dbcsr_syevd 48 12.0 0.003 0.003 5.222 5.223 make_images 4110 14.4 0.401 0.445 4.669 5.153 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.495 4.505 cp_fm_diag_elpa_base 48 14.0 4.435 4.467 4.492 4.503 density_rs2pw 110 9.6 0.004 0.005 3.958 4.460 rs_pw_transfer 902 11.9 0.012 0.013 3.788 4.347 pw_transfer 1331 11.6 0.067 0.076 4.135 4.296 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 4.027 4.191 wfi_extrapolate 11 7.9 0.001 0.001 3.685 3.685 qs_ot_get_derivative_diag 47 12.0 0.003 0.014 3.493 3.567 grid_integrate_task_list 110 12.3 3.136 3.366 3.136 3.366 fft_wrap_pw1pw2_140 451 13.1 0.515 0.529 3.185 3.353 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.273 3.275 calculate_dm_sparse 110 9.5 0.001 0.001 3.181 3.210 cp_fm_cholesky_invert 11 10.9 3.172 3.181 3.172 3.181 fft3d_ps 1111 14.6 1.113 1.336 3.008 3.176 make_images_data 4110 15.4 0.046 0.050 2.596 3.081 potential_pw2rs 110 12.3 0.008 0.009 3.019 3.048 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.893 2.958 hybrid_alltoall_any 4261 16.3 0.102 0.439 2.258 2.927 calculate_first_density_matrix 1 7.0 0.001 0.003 2.694 2.696 grid_collocate_task_list 110 9.6 2.102 2.539 2.102 2.539 mp_alltoall_d11v 2046 13.8 1.991 2.526 1.991 2.526 mp_sum_l 6594 12.7 1.921 2.497 1.921 2.497 jit_kernel_multiply 11 16.2 1.089 2.262 1.089 2.262 qs_ot_get_orbitals 99 10.5 0.001 0.001 2.026 2.050 qs_energies_init_hamiltonians 11 5.9 0.016 0.066 1.974 2.022 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.897 1.918 mp_waitany 10164 13.8 1.326 1.861 1.326 1.861 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.741 1.752 mp_allgather_i34 2055 14.4 0.754 1.746 0.754 1.746 cp_fm_cholesky_decompose 22 10.9 1.714 1.720 1.714 1.720 mp_alltoall_z22v 1111 16.6 1.605 1.702 1.605 1.702 multiply_cannon_metrocomm4 22605 15.4 0.077 0.080 0.785 1.691 mp_irecv_dv 57340 16.2 0.658 1.571 0.658 1.571 dbcsr_complete_redistribute 325 12.2 0.232 0.291 1.279 1.565 acc_transpose_blocks 24660 15.4 0.113 0.116 1.539 1.563 rs_pw_transfer_RS2PW_140 121 11.5 0.205 0.214 1.014 1.531 rs_gather_matrices 110 12.3 0.193 0.214 1.006 1.500 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=74.491000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=557.363636, yerr=7.401139 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4047e199d09ff4d3c3dd9bcab791603060e8611a_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 661.168128E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66428. MP_Allreduce 9774 562. MP_Sync 52 MP_Alltoall 1496 4511006. MP_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.036 59.925 59.926 qs_mol_dyn_low 1 2.0 0.003 0.004 59.554 59.563 qs_forces 11 3.9 0.003 0.004 59.481 59.482 qs_energies 11 4.9 0.001 0.002 56.252 56.256 scf_env_do_scf 11 5.9 0.000 0.001 48.319 48.320 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 39.837 39.837 velocity_verlet 10 3.0 0.001 0.001 32.560 32.562 dbcsr_multiply_generic 2055 12.4 0.108 0.114 28.264 28.495 qs_scf_new_mos 99 7.5 0.001 0.001 24.655 24.739 qs_scf_loop_do_ot 99 8.5 0.001 0.001 24.654 24.739 ot_scf_mini 99 9.5 0.002 0.003 23.433 23.537 multiply_cannon 2055 13.4 0.212 0.224 22.120 23.345 multiply_cannon_loop 2055 14.4 0.617 0.633 20.938 21.913 ot_mini 99 10.5 0.001 0.001 13.371 13.472 rebuild_ks_matrix 110 8.3 0.000 0.000 12.448 12.569 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.448 12.568 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.937 11.047 multiply_cannon_multrec 16440 15.4 3.939 4.919 9.797 10.715 mp_waitall_1 139946 16.5 6.802 9.957 6.802 9.957 qs_ot_get_derivative 99 11.5 0.001 0.001 8.976 9.083 init_scf_loop 11 6.9 0.000 0.000 8.449 8.449 multiply_cannon_metrocomm3 16440 15.4 0.042 0.045 4.213 7.275 prepare_preconditioner 11 7.9 0.000 0.000 6.678 6.695 make_preconditioner 11 8.9 0.000 0.000 6.678 6.695 sum_up_and_integrate 110 10.3 0.060 0.061 6.601 6.614 integrate_v_rspace 110 11.3 0.003 0.003 6.540 6.554 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.989 6.330 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.104 6.112 calculate_rho_elec 110 8.6 0.059 0.059 6.104 6.112 dbcsr_mm_accdrv_process 34862 16.1 4.732 5.389 5.711 5.860 init_scf_run 11 5.9 0.000 0.001 5.551 5.552 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.551 5.551 qs_ot_get_p 110 10.4 0.001 0.001 5.284 5.412 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.929 5.349 apply_single 110 13.6 0.000 0.000 4.929 5.349 make_m2s 4110 13.4 0.049 0.051 4.094 4.477 density_rs2pw 110 9.6 0.004 0.005 3.141 4.374 ot_diis_step 99 11.5 0.011 0.011 4.367 4.367 make_images 4110 14.4 0.393 0.512 3.977 4.358 multiply_cannon_sync_h2d 16440 15.4 3.701 4.328 3.701 4.328 rs_pw_transfer 902 11.9 0.010 0.011 2.692 3.922 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.106 3.748 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.706 3.709 pw_transfer 1331 11.6 0.066 0.074 3.689 3.698 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.582 3.592 grid_integrate_task_list 110 12.3 3.192 3.397 3.192 3.397 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.367 3.367 fft_wrap_pw1pw2_140 451 13.1 0.635 0.641 3.088 3.099 wfi_extrapolate 11 7.9 0.001 0.001 2.978 2.978 make_images_data 4110 15.4 0.042 0.046 2.362 2.857 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.836 2.845 cp_fm_diag_elpa_base 48 14.0 2.770 2.801 2.834 2.844 hybrid_alltoall_any 4261 16.3 0.106 0.377 2.057 2.758 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.733 2.734 calculate_dm_sparse 110 9.5 0.001 0.001 2.606 2.638 cp_fm_cholesky_invert 11 10.9 2.631 2.637 2.631 2.637 grid_collocate_task_list 110 9.6 2.124 2.515 2.124 2.515 calculate_first_density_matrix 1 7.0 0.000 0.000 2.480 2.480 mp_waitany 17072 13.8 1.188 2.468 1.188 2.468 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.417 2.461 fft3d_ps 1111 14.6 1.094 1.105 2.406 2.415 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.291 2.349 potential_pw2rs 110 12.3 0.011 0.012 2.303 2.312 multiply_cannon_metrocomm4 14385 15.4 0.044 0.048 0.880 2.287 mp_alltoall_d11v 2046 13.8 1.782 2.241 1.782 2.241 mp_irecv_dv 48980 15.7 0.810 2.164 0.810 2.164 rs_pw_transfer_RS2PW_140 121 11.5 0.174 0.178 0.902 2.135 jit_kernel_multiply 10 16.5 0.584 2.012 0.584 2.012 mp_sum_l 6594 12.7 1.433 1.999 1.433 1.999 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.942 1.942 dbcsr_complete_redistribute 325 12.2 0.325 0.343 1.393 1.850 cp_fm_upper_to_full 70 14.2 1.372 1.742 1.372 1.742 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.708 1.721 cp_fm_cholesky_decompose 22 10.9 1.590 1.610 1.590 1.610 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.567 1.579 mp_allgather_i34 2055 14.4 0.504 1.520 0.504 1.520 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.356 1.464 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.970 1.415 rs_gather_matrices 110 12.3 0.234 0.264 0.967 1.374 acc_transpose_blocks 16440 15.4 0.074 0.077 1.227 1.243 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.235 1.241 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=59.926000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=625.090909, yerr=8.670268 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4047e199d09ff4d3c3dd9bcab791603060e8611a_performance_tests/12/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 601.317074E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 4916280 0.0% 0.0% 100.0% average stack size 0.0 0.0 200.2 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 729.583616E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66426. MP_Allreduce 9774 603. MP_Sync 52 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.033 72.350 72.350 qs_mol_dyn_low 1 2.0 0.003 0.004 72.001 72.010 qs_forces 11 3.9 0.003 0.003 71.927 71.928 qs_energies 11 4.9 0.001 0.001 68.370 68.374 scf_env_do_scf 11 5.9 0.000 0.001 59.493 59.495 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 45.839 45.839 velocity_verlet 10 3.0 0.001 0.001 40.198 40.200 dbcsr_multiply_generic 2055 12.4 0.117 0.122 31.027 31.241 qs_scf_new_mos 99 7.5 0.001 0.001 29.741 29.859 qs_scf_loop_do_ot 99 8.5 0.001 0.001 29.741 29.858 ot_scf_mini 99 9.5 0.003 0.003 28.065 28.177 multiply_cannon 2055 13.4 0.240 0.263 23.125 24.434 multiply_cannon_loop 2055 14.4 0.889 0.911 21.545 22.081 ot_mini 99 10.5 0.001 0.001 15.194 15.315 multiply_cannon_multrec 24660 15.4 4.259 6.862 12.971 14.278 init_scf_loop 11 6.9 0.000 0.000 13.602 13.603 rebuild_ks_matrix 110 8.3 0.000 0.000 12.818 12.978 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.817 12.978 prepare_preconditioner 11 7.9 0.000 0.000 11.758 11.773 make_preconditioner 11 8.9 0.000 0.000 11.758 11.773 make_full_inverse_cholesky 11 9.9 0.000 0.000 9.936 11.454 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.299 11.444 qs_ot_get_derivative 99 11.5 0.001 0.001 10.966 11.089 dbcsr_mm_accdrv_process 52304 16.0 6.958 8.379 8.567 9.505 qs_ot_get_p 110 10.4 0.001 0.001 7.686 7.820 sum_up_and_integrate 110 10.3 0.067 0.071 7.037 7.052 mp_waitall_1 121746 16.5 5.201 7.029 5.201 7.029 integrate_v_rspace 110 11.3 0.003 0.003 6.970 6.986 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.780 6.792 calculate_rho_elec 110 8.6 0.078 0.081 6.780 6.792 make_m2s 4110 13.4 0.059 0.060 6.113 6.388 make_images 4110 14.4 0.575 0.693 5.970 6.241 init_scf_run 11 5.9 0.000 0.001 6.084 6.085 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.084 6.085 qs_ot_p2m_diag 48 11.0 0.054 0.062 5.584 5.598 cp_dbcsr_syevd 48 12.0 0.003 0.003 5.072 5.073 cp_fm_upper_to_full 70 14.2 3.475 5.046 3.475 5.046 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.338 4.346 cp_fm_diag_elpa_base 48 14.0 4.166 4.232 4.335 4.343 ot_diis_step 99 11.5 0.011 0.012 4.170 4.170 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.061 4.166 apply_single 110 13.6 0.000 0.000 4.061 4.166 density_rs2pw 110 9.6 0.004 0.004 3.624 4.144 pw_transfer 1331 11.6 0.066 0.077 4.039 4.084 dbcsr_complete_redistribute 325 12.2 0.410 0.453 2.890 4.061 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.932 3.981 cp_fm_cholesky_invert 11 10.9 3.852 3.862 3.852 3.862 rs_pw_transfer 902 11.9 0.010 0.010 3.195 3.721 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.528 3.581 multiply_cannon_metrocomm3 24660 15.4 0.036 0.037 1.583 3.533 make_images_data 4110 15.4 0.045 0.048 3.130 3.488 multiply_cannon_sync_h2d 24660 15.4 3.160 3.482 3.160 3.482 grid_integrate_task_list 110 12.3 3.246 3.481 3.246 3.481 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.370 3.433 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.245 3.415 hybrid_alltoall_any 4261 16.3 0.120 0.456 2.584 3.317 fft_wrap_pw1pw2_140 451 13.1 0.664 0.685 3.205 3.259 wfi_extrapolate 11 7.9 0.001 0.001 3.256 3.256 calculate_dm_sparse 110 9.5 0.001 0.001 3.131 3.163 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.998 2.999 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.813 2.964 mp_alltoall_i22 605 13.7 1.718 2.914 1.718 2.914 fft3d_ps 1111 14.6 1.086 1.116 2.700 2.722 calculate_first_density_matrix 1 7.0 0.000 0.000 2.715 2.718 potential_pw2rs 110 12.3 0.012 0.013 2.646 2.668 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.584 2.617 mp_alltoall_d11v 2046 13.8 1.984 2.546 1.984 2.546 jit_kernel_multiply 12 15.8 1.272 2.542 1.272 2.542 grid_collocate_task_list 110 9.6 2.222 2.500 2.222 2.500 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.252 2.253 cp_fm_cholesky_decompose 22 10.9 2.191 2.245 2.191 2.245 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.822 1.853 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.764 1.778 mp_allgather_i34 2055 14.4 0.683 1.725 0.683 1.725 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.704 1.718 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.601 1.705 mp_waitany 13376 13.8 1.182 1.695 1.182 1.695 multiply_cannon_metrocomm4 20550 15.4 0.059 0.063 0.843 1.664 mp_irecv_dv 62702 16.1 0.743 1.584 0.743 1.584 acc_transpose_blocks 24660 15.4 0.105 0.107 1.548 1.569 mp_sum_l 6594 12.7 1.105 1.564 1.105 1.564 rs_gather_matrices 110 12.3 0.267 0.297 0.991 1.520 mp_alltoall_z22v 1111 16.6 1.400 1.469 1.400 1.469 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=72.350000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=691.636364, yerr=8.293182 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4047e199d09ff4d3c3dd9bcab791603060e8611a_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 836.362240E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66424. MP_Allreduce 9774 644. MP_Sync 52 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.029 55.988 55.989 qs_mol_dyn_low 1 2.0 0.003 0.003 55.677 55.708 qs_forces 11 3.9 0.003 0.003 55.611 55.612 qs_energies 11 4.9 0.001 0.001 51.875 51.882 scf_env_do_scf 11 5.9 0.000 0.001 43.227 43.228 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 35.436 35.437 velocity_verlet 10 3.0 0.001 0.001 31.307 31.310 dbcsr_multiply_generic 2055 12.4 0.105 0.109 23.103 23.245 qs_scf_new_mos 99 7.5 0.001 0.001 20.286 20.335 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.286 20.335 ot_scf_mini 99 9.5 0.002 0.002 19.045 19.068 multiply_cannon 2055 13.4 0.249 0.262 17.647 18.949 multiply_cannon_loop 2055 14.4 0.322 0.335 16.305 16.546 rebuild_ks_matrix 110 8.3 0.000 0.000 12.011 12.038 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 12.010 12.037 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.629 10.651 ot_mini 99 10.5 0.001 0.001 10.298 10.314 multiply_cannon_multrec 8220 15.4 3.252 4.528 7.838 8.714 init_scf_loop 11 6.9 0.000 0.000 7.744 7.746 mp_waitall_1 103326 16.6 5.972 7.648 5.972 7.648 sum_up_and_integrate 110 10.3 0.080 0.081 6.579 6.589 qs_ot_get_derivative 99 11.5 0.001 0.001 6.527 6.550 integrate_v_rspace 110 11.3 0.003 0.003 6.499 6.509 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.353 6.367 calculate_rho_elec 110 8.6 0.114 0.114 6.352 6.367 prepare_preconditioner 11 7.9 0.000 0.000 6.074 6.077 make_preconditioner 11 8.9 0.000 0.000 6.074 6.077 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.668 5.742 init_scf_run 11 5.9 0.000 0.001 5.526 5.526 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.526 5.526 dbcsr_mm_accdrv_process 17442 15.9 2.947 3.932 4.457 5.443 qs_ot_get_p 110 10.4 0.001 0.001 4.821 4.836 make_m2s 4110 13.4 0.038 0.040 4.224 4.499 make_images 4110 14.4 0.637 0.695 4.092 4.367 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 2.935 4.250 pw_transfer 1331 11.6 0.066 0.071 3.984 3.995 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.876 3.889 ot_diis_step 99 11.5 0.012 0.012 3.747 3.748 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.660 3.680 apply_single 110 13.6 0.000 0.000 3.659 3.680 grid_integrate_task_list 110 12.3 3.356 3.578 3.356 3.578 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.467 3.471 density_rs2pw 110 9.6 0.004 0.004 3.063 3.434 fft_wrap_pw1pw2_140 451 13.1 0.832 0.848 3.334 3.362 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.164 3.164 multiply_cannon_sync_h2d 8220 15.4 2.915 2.998 2.915 2.998 cp_fm_cholesky_invert 11 10.9 2.924 2.927 2.924 2.927 make_images_data 4110 15.4 0.037 0.042 2.427 2.866 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.797 2.799 hybrid_alltoall_any 4261 16.3 0.200 0.865 2.364 2.784 wfi_extrapolate 11 7.9 0.001 0.001 2.753 2.753 calculate_first_density_matrix 1 7.0 0.000 0.000 2.671 2.672 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.667 2.668 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.638 2.646 cp_fm_diag_elpa_base 48 14.0 2.581 2.604 2.636 2.644 calculate_dm_sparse 110 9.5 0.001 0.001 2.553 2.595 rs_pw_transfer 902 11.9 0.010 0.010 2.193 2.592 grid_collocate_task_list 110 9.6 2.315 2.571 2.315 2.571 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.450 2.465 fft3d_ps 1111 14.6 1.137 1.179 2.424 2.439 potential_pw2rs 110 12.3 0.015 0.015 2.094 2.098 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.773 1.992 mp_alltoall_d11v 2046 13.8 1.645 1.955 1.645 1.955 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.932 1.944 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.775 1.787 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.713 1.717 cp_fm_cholesky_decompose 22 10.9 1.697 1.715 1.697 1.715 mp_allgather_i34 2055 14.4 0.526 1.642 0.526 1.642 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.502 1.632 dbcsr_complete_redistribute 325 12.2 0.560 0.583 1.425 1.527 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.514 1.527 mp_waitany 9240 13.8 1.071 1.505 1.071 1.505 jit_kernel_multiply 7 15.7 1.196 1.492 1.196 1.492 qs_create_task_list 11 7.9 0.000 0.001 1.218 1.325 generate_qs_task_list 11 8.9 0.374 0.441 1.218 1.325 rs_gather_matrices 110 12.3 0.326 0.366 0.958 1.258 rs_pw_transfer_RS2PW_140 121 11.5 0.161 0.165 0.805 1.211 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.148 1.180 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=55.989000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=785.909091, yerr=14.798369 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4047e199d09ff4d3c3dd9bcab791603060e8611a_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 1.380958E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67104. MP_Allreduce 9752 812. MP_Sync 52 MP_Alltoall 1474 16505187. MP_SendRecv 2310 360267. MP_ISendRecv 2310 360267. MP_Wait 5214 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.035 88.964 88.965 qs_mol_dyn_low 1 2.0 0.003 0.003 88.640 88.649 qs_forces 11 3.9 0.003 0.003 88.287 88.288 qs_energies 11 4.9 0.001 0.001 84.063 84.065 scf_env_do_scf 11 5.9 0.001 0.001 73.723 73.724 velocity_verlet 10 3.0 0.001 0.001 56.876 56.918 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 44.794 44.794 dbcsr_multiply_generic 2055 12.4 0.119 0.125 29.008 29.110 init_scf_loop 11 6.9 0.000 0.000 28.857 28.859 prepare_preconditioner 11 7.9 0.000 0.000 26.822 26.828 make_preconditioner 11 8.9 0.000 0.000 26.822 26.828 qs_scf_new_mos 99 7.5 0.001 0.001 26.487 26.521 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.486 26.520 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.852 26.308 ot_scf_mini 99 9.5 0.002 0.002 24.714 24.739 multiply_cannon 2055 13.4 0.353 0.377 21.793 22.477 multiply_cannon_loop 2055 14.4 0.342 0.344 19.994 20.274 cp_fm_upper_to_full 70 14.2 12.884 18.596 12.884 18.596 rebuild_ks_matrix 110 8.3 0.000 0.001 14.173 14.217 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 14.173 14.217 ot_mini 99 10.5 0.001 0.001 13.725 13.750 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.806 12.845 dbcsr_complete_redistribute 325 12.2 1.017 1.034 7.539 10.763 multiply_cannon_multrec 8220 15.4 4.434 4.625 9.735 9.947 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.518 9.738 qs_ot_get_derivative 99 11.5 0.001 0.001 9.131 9.156 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.954 9.132 mp_alltoall_i22 605 13.7 5.603 8.872 5.603 8.872 mp_waitall_1 84994 16.7 7.813 8.754 7.813 8.754 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.770 7.802 calculate_rho_elec 110 8.6 0.225 0.225 7.770 7.801 sum_up_and_integrate 110 10.3 0.151 0.152 7.350 7.364 integrate_v_rspace 110 11.3 0.004 0.004 7.200 7.215 init_scf_run 11 5.9 0.000 0.001 6.057 6.057 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.057 6.057 make_m2s 4110 13.4 0.042 0.043 5.446 5.974 make_images 4110 14.4 0.875 0.921 5.256 5.783 qs_ot_get_p 110 10.4 0.001 0.001 5.735 5.761 dbcsr_mm_accdrv_process 11614 15.7 3.201 3.350 5.159 5.439 cp_fm_cholesky_invert 11 10.9 5.403 5.408 5.403 5.408 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.712 5.174 apply_single 110 13.6 0.000 0.000 4.712 5.174 multiply_cannon_metrocomm3 8220 15.4 0.018 0.019 4.811 5.166 pw_transfer 1331 11.6 0.075 0.076 5.077 5.082 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 4.959 4.965 ot_diis_step 99 11.5 0.015 0.016 4.568 4.569 fft_wrap_pw1pw2_140 451 13.1 1.334 1.342 4.317 4.327 qs_ot_p2m_diag 48 11.0 0.150 0.155 4.166 4.174 multiply_cannon_sync_h2d 8220 15.4 3.942 3.944 3.942 3.944 hybrid_alltoall_any 4261 16.3 0.257 0.557 3.046 3.830 density_rs2pw 110 9.6 0.004 0.004 3.754 3.799 grid_integrate_task_list 110 12.3 3.678 3.744 3.678 3.744 make_images_data 4110 15.4 0.040 0.043 3.059 3.717 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.713 3.713 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.672 3.673 wfi_extrapolate 11 7.9 0.001 0.001 3.368 3.369 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.860 3.320 calculate_dm_sparse 110 9.5 0.001 0.001 3.240 3.258 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.104 3.105 cp_fm_diag_elpa_base 48 14.0 2.557 2.761 3.102 3.103 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.038 3.044 fft3d_ps 1111 14.6 1.304 1.312 2.893 2.900 grid_collocate_task_list 110 9.6 2.672 2.683 2.672 2.683 calculate_first_density_matrix 1 7.0 0.000 0.000 2.566 2.567 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.525 2.555 potential_pw2rs 110 12.3 0.021 0.021 2.441 2.447 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.399 2.414 rs_pw_transfer 902 11.9 0.010 0.011 2.308 2.331 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.185 2.233 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.088 2.188 mp_alltoall_d11v 2046 13.8 1.970 2.085 1.970 2.085 cp_fm_cholesky_decompose 22 10.9 2.051 2.071 2.051 2.071 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.969 1.974 qs_create_task_list 11 7.9 0.001 0.001 1.882 1.929 generate_qs_task_list 11 8.9 0.730 0.784 1.882 1.929 jit_kernel_multiply 10 15.5 1.754 1.905 1.754 1.905 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.781 1.825 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.765 1.781 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=88.965000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1230.000000, yerr=62.545983 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4047e199d09ff4d3c3dd9bcab791603060e8611a_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 637.026304E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175955383376 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4010 57653. MP_Allreduce 11104 795. MP_Sync 87 MP_Alltoall 2226 3098169. MP_SendRecv 24320 18752. MP_ISendRecv 24320 18752. MP_Wait 42476 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.057 206.980 206.981 qs_mol_dyn_low 1 2.0 0.003 0.004 206.479 206.493 qs_forces 11 3.9 0.005 0.005 206.395 206.395 qs_energies 11 4.9 0.001 0.001 200.761 200.782 scf_env_do_scf 11 5.9 0.001 0.001 184.153 184.156 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 163.630 163.631 dbcsr_multiply_generic 2507 12.6 0.182 0.186 124.187 124.561 velocity_verlet 10 3.0 0.001 0.002 123.942 123.943 qs_scf_new_mos 117 7.6 0.001 0.001 123.340 123.555 qs_scf_loop_do_ot 117 8.6 0.001 0.001 123.339 123.554 ot_scf_mini 117 9.6 0.003 0.003 116.750 116.943 multiply_cannon 2507 13.6 0.238 0.249 101.033 102.189 multiply_cannon_loop 2507 14.6 2.083 2.126 98.823 99.893 ot_mini 117 10.6 0.001 0.001 65.400 65.609 multiply_cannon_multrec 60168 15.6 33.360 35.286 41.671 43.785 qs_ot_get_derivative 117 11.6 0.001 0.001 40.659 40.868 rebuild_ks_matrix 128 8.3 0.001 0.001 33.652 33.851 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 33.651 33.850 mp_waitall_1 267128 16.5 28.187 31.202 28.187 31.202 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.175 30.379 qs_ot_get_p 128 10.4 0.001 0.001 29.455 29.702 multiply_cannon_sync_h2d 60168 15.6 27.473 29.187 27.473 29.187 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.379 24.928 apply_single 128 13.6 0.001 0.001 24.378 24.928 ot_diis_step 117 11.6 0.008 0.008 24.495 24.496 qs_ot_p2m_diag 83 11.4 0.079 0.091 22.884 22.941 init_scf_loop 11 6.9 0.000 0.001 20.449 20.450 cp_dbcsr_syevd 83 12.4 0.005 0.005 20.309 20.309 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.407 19.566 multiply_cannon_metrocomm3 60168 15.6 0.111 0.115 15.675 17.585 cp_fm_diag_elpa 83 13.4 0.000 0.000 17.366 17.403 cp_fm_diag_elpa_base 83 14.4 17.298 17.338 17.362 17.399 prepare_preconditioner 11 7.9 0.000 0.000 15.870 15.920 make_preconditioner 11 8.9 0.000 0.000 15.870 15.920 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.114 15.260 sum_up_and_integrate 128 10.3 0.088 0.104 14.180 14.193 integrate_v_rspace 128 11.3 0.004 0.004 14.091 14.108 make_m2s 5014 13.6 0.105 0.114 13.800 14.099 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.858 14.005 calculate_rho_elec 128 8.7 0.046 0.065 13.857 14.005 make_images 5014 14.6 0.399 0.418 13.616 13.924 init_scf_run 11 5.9 0.000 0.001 12.475 12.475 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.475 12.475 density_rs2pw 128 9.7 0.006 0.007 7.104 10.437 mp_sum_l 7950 12.9 8.089 9.394 8.089 9.394 wfi_extrapolate 11 7.9 0.001 0.001 9.142 9.142 cp_fm_cholesky_invert 11 10.9 9.053 9.061 9.053 9.061 rs_pw_transfer 1046 11.9 0.016 0.019 5.710 9.030 calculate_dm_sparse 128 9.5 0.001 0.001 8.484 8.616 dbcsr_mm_accdrv_process 124484 16.2 3.174 3.327 7.873 8.377 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.954 8.068 pw_transfer 1547 11.6 0.074 0.090 7.777 7.968 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.852 7.917 multiply_cannon_metrocomm1 60168 15.6 0.089 0.093 6.011 7.773 fft_wrap_pw1pw2 1291 12.7 0.011 0.013 7.575 7.764 make_images_data 5014 15.6 0.065 0.071 6.809 7.741 grid_integrate_task_list 128 12.3 7.065 7.583 7.065 7.583 hybrid_alltoall_any 5200 16.5 0.290 2.255 5.934 7.441 fft_wrap_pw1pw2_140 523 13.2 1.274 1.321 6.627 6.798 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.668 6.680 grid_collocate_task_list 128 9.7 4.670 6.001 4.670 6.001 mp_waitany 16020 13.9 2.651 6.001 2.651 6.001 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.785 5.839 fft3d_ps 1291 14.7 2.145 2.788 5.340 5.671 mp_alltoall_d11v 2415 14.1 4.257 5.586 4.257 5.586 rs_pw_transfer_RS2PW_140 139 11.5 0.280 0.304 2.147 5.505 cp_fm_cholesky_decompose 22 10.9 4.646 4.659 4.646 4.659 potential_pw2rs 128 12.3 0.009 0.011 4.537 4.557 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=206.981000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=598.818182, yerr=7.456762 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4047e199d09ff4d3c3dd9bcab791603060e8611a_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 2.199914E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 6022464 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.2 marketing flops 145.647559E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 834.465792E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2425920 MPI messages size (bytes): total size 4.132350E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703416E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 71436 2336489472 32768 < size <= 131072 728832 55956209664 131072 < size <= 4194304 1386864 1409906900992 4194304 < size <= 16777216 155760 1473826772352 16777216 < size 68112 1190343475200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4024 57909. MP_Allreduce 11138 958. MP_Sync 87 MP_Alltoall 1983 4663719. MP_SendRecv 12126 47072. MP_ISendRecv 12126 47072. MP_Wait 26114 MP_ISend 11836 212447. MP_IRecv 11836 212447. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.027 0.054 194.333 194.334 qs_mol_dyn_low 1 2.0 0.003 0.005 193.895 193.910 qs_forces 11 3.9 0.004 0.005 193.814 193.815 qs_energies 11 4.9 0.002 0.007 186.963 186.974 scf_env_do_scf 11 5.9 0.001 0.002 170.425 170.435 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 136.720 136.722 velocity_verlet 10 3.0 0.001 0.002 122.413 122.414 dbcsr_multiply_generic 2527 12.6 0.193 0.198 98.544 99.796 qs_scf_new_mos 118 7.6 0.001 0.001 95.967 96.579 qs_scf_loop_do_ot 118 8.6 0.001 0.001 95.966 96.578 ot_scf_mini 118 9.6 0.004 0.005 91.105 91.777 multiply_cannon 2527 13.6 0.483 0.542 78.042 82.386 multiply_cannon_loop 2527 14.6 1.254 1.296 74.922 77.984 ot_mini 118 10.6 0.001 0.001 50.786 51.413 mp_waitall_1 216438 16.6 25.342 38.727 25.342 38.727 multiply_cannon_multrec 30324 15.6 21.940 26.564 31.748 36.555 rebuild_ks_matrix 129 8.3 0.001 0.001 33.676 34.224 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.019 33.675 34.223 init_scf_loop 11 6.9 0.001 0.004 33.611 33.613 qs_ks_update_qs_env 129 7.6 0.001 0.001 30.297 30.809 multiply_cannon_metrocomm3 30324 15.6 0.090 0.096 16.319 29.635 prepare_preconditioner 11 7.9 0.000 0.001 29.197 29.288 make_preconditioner 11 8.9 0.000 0.002 29.197 29.287 qs_ot_get_derivative 118 11.6 0.001 0.002 28.536 29.199 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.842 28.423 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 22.296 23.532 apply_single 129 13.6 0.001 0.001 22.295 23.532 qs_ot_get_p 129 10.4 0.001 0.001 21.935 22.720 ot_diis_step 118 11.6 0.015 0.015 22.071 22.074 multiply_cannon_sync_h2d 30324 15.6 19.231 21.328 19.231 21.328 cp_fm_cholesky_invert 11 10.9 17.105 17.117 17.105 17.117 qs_ot_p2m_diag 83 11.4 0.187 0.215 17.067 17.108 make_m2s 5054 13.6 0.090 0.096 14.593 15.929 cp_dbcsr_syevd 83 12.4 0.006 0.006 15.893 15.894 make_images 5054 14.6 1.172 1.362 14.374 15.712 sum_up_and_integrate 129 10.3 0.117 0.136 14.979 15.012 integrate_v_rspace 129 11.3 0.003 0.004 14.862 14.901 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.481 14.516 calculate_rho_elec 129 8.7 0.088 0.105 14.481 14.515 cp_fm_diag_elpa 83 13.4 0.000 0.000 12.664 12.696 cp_fm_diag_elpa_base 83 14.4 12.401 12.493 12.658 12.693 init_scf_run 11 5.9 0.000 0.001 11.682 11.684 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.682 11.683 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 11.143 11.644 multiply_cannon_metrocomm4 27797 15.6 0.103 0.117 3.839 10.581 make_images_data 5054 15.6 0.063 0.073 8.876 10.560 mp_irecv_dv 70031 16.3 3.636 10.181 3.636 10.181 density_rs2pw 129 9.7 0.006 0.007 7.697 10.128 hybrid_alltoall_any 5240 16.5 0.347 1.539 7.537 9.920 dbcsr_mm_accdrv_process 62734 16.2 4.496 5.295 9.264 9.913 pw_transfer 1559 11.6 0.087 0.100 9.033 9.093 fft_wrap_pw1pw2 1301 12.7 0.011 0.012 8.806 8.870 wfi_extrapolate 11 7.9 0.001 0.001 8.434 8.434 rs_pw_transfer 1054 12.0 0.014 0.017 5.943 8.380 fft_wrap_pw1pw2_140 527 13.2 1.344 1.371 7.739 7.818 grid_integrate_task_list 129 12.3 7.223 7.597 7.223 7.597 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 6.543 7.300 cp_fm_cholesky_decompose 22 10.9 7.067 7.151 7.067 7.151 calculate_dm_sparse 129 9.5 0.001 0.001 6.523 6.647 mp_sum_l 8010 12.9 4.128 6.302 4.128 6.302 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.207 6.216 fft3d_ps 1301 14.7 2.837 3.011 6.096 6.161 grid_collocate_task_list 129 9.7 4.825 5.973 4.825 5.973 qs_ot_get_orbitals 118 10.6 0.001 0.001 5.440 5.504 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.315 5.471 mp_alltoall_d11v 2423 14.1 4.316 5.422 4.316 5.422 potential_pw2rs 129 12.3 0.016 0.019 5.116 5.138 mp_waitany 11836 13.9 2.550 5.067 2.550 5.067 mp_allgather_i34 2527 14.6 1.676 4.749 1.676 4.749 rs_pw_transfer_RS2PW_140 140 11.5 0.353 0.383 2.136 4.571 dbcsr_complete_redistribute 395 12.7 0.800 0.909 3.338 4.187 mp_sum_d 4499 12.1 2.719 4.110 2.719 4.110 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=194.334000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=795.181818, yerr=1.849726 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4047e199d09ff4d3c3dd9bcab791603060e8611a_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 2.928533E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 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 944.201728E+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 931531265168 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 58214. MP_Allreduce 11082 999. MP_Sync 87 MP_Alltoall 1712 9388896. MP_SendRecv 7936 75008. MP_ISendRecv 7936 75008. MP_Wait 21820 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 0.017 0.033 189.361 189.568 qs_mol_dyn_low 1 2.0 0.003 0.004 188.917 189.122 qs_forces 11 3.9 0.004 0.004 188.814 189.021 qs_energies 11 4.9 0.001 0.001 182.054 182.249 scf_env_do_scf 11 5.9 0.001 0.001 166.042 166.224 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 126.723 126.861 velocity_verlet 10 3.0 0.001 0.001 122.176 122.311 qs_scf_new_mos 117 7.6 0.001 0.001 86.864 87.203 qs_scf_loop_do_ot 117 8.6 0.001 0.001 86.864 87.202 dbcsr_multiply_generic 2507 12.6 0.183 0.190 83.263 84.604 ot_scf_mini 117 9.6 0.003 0.004 82.666 83.068 multiply_cannon 2507 13.6 0.501 0.524 62.521 66.278 multiply_cannon_loop 2507 14.6 0.860 0.886 58.898 62.567 ot_mini 117 10.6 0.001 0.001 44.058 44.417 init_scf_loop 11 6.9 0.000 0.000 39.207 39.250 mp_waitall_1 170520 16.6 26.030 37.410 26.030 37.410 prepare_preconditioner 11 7.9 0.000 0.000 35.064 35.124 make_preconditioner 11 8.9 0.000 0.000 35.064 35.124 make_full_inverse_cholesky 11 9.9 0.000 0.000 32.612 34.084 rebuild_ks_matrix 128 8.3 0.001 0.001 31.642 32.136 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.017 31.642 32.136 qs_ks_update_qs_env 128 7.6 0.001 0.001 28.475 28.930 multiply_cannon_metrocomm3 20056 15.6 0.058 0.061 15.573 26.704 multiply_cannon_multrec 20056 15.6 13.397 16.806 22.082 25.523 qs_ot_get_derivative 117 11.6 0.002 0.002 24.134 24.482 qs_ot_get_p 128 10.4 0.001 0.001 23.731 24.166 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 20.025 21.207 apply_single 128 13.6 0.001 0.001 20.025 21.206 ot_diis_step 117 11.6 0.018 0.018 19.797 19.819 qs_ot_p2m_diag 83 11.4 0.265 0.272 19.142 19.176 cp_dbcsr_syevd 83 12.4 0.005 0.006 17.976 17.998 cp_fm_cholesky_invert 11 10.9 17.165 17.189 17.165 17.189 make_m2s 5014 13.6 0.079 0.083 15.192 16.506 make_images 5014 14.6 1.189 1.280 14.956 16.271 multiply_cannon_sync_h2d 20056 15.6 14.170 16.010 14.170 16.010 sum_up_and_integrate 128 10.3 0.131 0.144 15.449 15.476 qs_rho_update_rho_low 128 7.7 0.001 0.001 15.341 15.392 calculate_rho_elec 128 8.7 0.130 0.145 15.340 15.392 integrate_v_rspace 128 11.3 0.004 0.004 15.317 15.347 cp_fm_diag_elpa 83 13.4 0.000 0.000 14.386 14.407 cp_fm_diag_elpa_base 83 14.4 13.934 14.120 14.382 14.403 make_images_data 5014 15.6 0.058 0.065 9.203 10.982 init_scf_run 11 5.9 0.000 0.001 10.932 10.943 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.932 10.943 density_rs2pw 128 9.7 0.006 0.006 8.118 10.319 hybrid_alltoall_any 5200 16.5 0.434 1.988 7.999 9.997 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.628 9.880 pw_transfer 1547 11.6 0.087 0.104 9.327 9.441 multiply_cannon_metrocomm4 17549 15.6 0.061 0.071 3.483 9.375 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 9.102 9.224 mp_irecv_dv 50230 16.2 3.362 9.125 3.362 9.125 cp_fm_cholesky_decompose 22 10.9 8.648 8.685 8.648 8.685 rs_pw_transfer 1046 11.9 0.013 0.014 6.427 8.684 dbcsr_mm_accdrv_process 41502 16.2 4.469 5.297 8.150 8.297 grid_integrate_task_list 128 12.3 7.267 8.086 7.267 8.086 wfi_extrapolate 11 7.9 0.001 0.001 7.915 7.923 fft_wrap_pw1pw2_140 523 13.2 1.408 1.437 7.790 7.914 cp_fm_upper_to_full 105 14.8 5.825 7.457 5.825 7.457 dbcsr_complete_redistribute 395 12.7 1.172 1.202 4.828 6.709 fft3d_ps 1291 14.7 2.738 2.974 6.230 6.326 mp_alltoall_d11v 2415 14.1 4.552 6.249 4.552 6.249 calculate_dm_sparse 128 9.5 0.001 0.001 5.815 5.905 grid_collocate_task_list 128 9.7 4.981 5.793 4.981 5.793 potential_pw2rs 128 12.3 0.020 0.022 5.470 5.520 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.495 5.505 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.809 5.496 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.510 5.392 mp_sum_l 7950 12.9 3.730 5.172 3.730 5.172 mp_allgather_i34 2507 14.6 2.137 4.931 2.137 4.931 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.530 4.676 mp_waitany 11748 13.9 2.431 4.625 2.431 4.625 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.235 4.271 transfer_fm_to_dbcsr 11 9.9 0.020 0.029 2.432 4.247 rs_pw_transfer_RS2PW_140 139 11.5 0.331 0.356 1.960 4.158 mp_alltoall_i22 716 14.1 2.076 4.128 2.076 4.128 rs_gather_matrices 128 12.3 0.619 0.686 2.390 3.953 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=189.568000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=896.454545, yerr=9.158837 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4047e199d09ff4d3c3dd9bcab791603060e8611a_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 4.353791E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 5977344 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.136128E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1143192 MPI messages size (bytes): total size 2.023815E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.770320E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 319024 36042702848 131072 < size <= 4194304 715736 785529176064 4194304 < size <= 16777216 70320 665379475120 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 58210. MP_Allreduce 11082 1082. MP_Sync 87 MP_Alltoall 1712 12503107. MP_SendRecv 5888 75008. MP_ISendRecv 5888 75008. MP_Wait 22442 MP_ISend 14952 244818. MP_IRecv 14952 244818. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.033 0.064 204.625 204.626 qs_mol_dyn_low 1 2.0 0.003 0.004 204.138 204.151 qs_forces 11 3.9 0.004 0.004 204.023 204.028 qs_energies 11 4.9 0.002 0.007 196.633 196.640 scf_env_do_scf 11 5.9 0.001 0.002 179.389 179.396 velocity_verlet 10 3.0 0.001 0.001 136.117 136.120 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 126.756 126.758 qs_scf_new_mos 117 7.6 0.001 0.001 87.820 88.170 qs_scf_loop_do_ot 117 8.6 0.001 0.001 87.819 88.169 ot_scf_mini 117 9.6 0.003 0.004 83.322 83.610 dbcsr_multiply_generic 2507 12.6 0.189 0.194 80.742 81.576 multiply_cannon 2507 13.6 0.555 0.585 54.907 58.387 multiply_cannon_loop 2507 14.6 1.188 1.215 50.931 53.017 init_scf_loop 11 6.9 0.001 0.003 52.498 52.499 prepare_preconditioner 11 7.9 0.000 0.000 48.266 48.292 make_preconditioner 11 8.9 0.000 0.002 48.266 48.292 make_full_inverse_cholesky 11 9.9 0.012 0.021 41.566 46.817 ot_mini 117 10.6 0.001 0.001 43.692 43.989 rebuild_ks_matrix 128 8.3 0.001 0.001 30.812 31.117 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.021 30.811 31.117 multiply_cannon_multrec 30084 15.6 14.177 19.296 26.137 30.911 mp_waitall_1 147882 16.7 18.146 28.921 18.146 28.921 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.735 28.009 qs_ot_get_p 128 10.4 0.001 0.001 25.653 25.977 qs_ot_get_derivative 117 11.6 0.002 0.002 23.999 24.301 make_m2s 5014 13.6 0.096 0.100 21.169 22.160 make_images 5014 14.6 1.952 2.264 20.857 21.845 qs_ot_p2m_diag 83 11.4 0.342 0.389 21.176 21.233 cp_dbcsr_syevd 83 12.4 0.005 0.006 19.793 19.795 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 19.076 19.669 apply_single 128 13.6 0.001 0.001 19.076 19.668 cp_fm_cholesky_invert 11 10.9 19.642 19.655 19.642 19.655 ot_diis_step 117 11.6 0.018 0.019 19.564 19.566 cp_fm_upper_to_full 105 14.8 11.897 17.592 11.897 17.592 cp_fm_diag_elpa 83 13.4 0.000 0.000 16.175 16.188 cp_fm_diag_elpa_base 83 14.4 15.112 15.488 16.169 16.183 sum_up_and_integrate 128 10.3 0.140 0.154 15.531 15.561 integrate_v_rspace 128 11.3 0.004 0.005 15.391 15.426 multiply_cannon_metrocomm3 30084 15.6 0.047 0.049 6.256 15.420 qs_rho_update_rho_low 128 7.7 0.001 0.001 15.377 15.413 calculate_rho_elec 128 8.7 0.173 0.189 15.376 15.412 dbcsr_complete_redistribute 395 12.7 1.509 1.630 9.499 13.445 make_images_data 5014 15.6 0.060 0.065 11.439 13.186 multiply_cannon_sync_h2d 30084 15.6 11.676 12.624 11.676 12.624 hybrid_alltoall_any 5200 16.5 0.527 2.202 10.006 12.167 dbcsr_mm_accdrv_process 62264 16.2 7.309 8.287 11.530 12.089 copy_fm_to_dbcsr 209 11.7 0.002 0.003 8.009 11.946 init_scf_run 11 5.9 0.000 0.001 11.265 11.267 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.265 11.267 transfer_fm_to_dbcsr 11 9.9 0.001 0.010 6.677 10.510 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 10.012 10.238 mp_alltoall_i22 716 14.1 5.853 9.901 5.853 9.901 pw_transfer 1547 11.6 0.087 0.103 9.517 9.604 cp_fm_cholesky_decompose 22 10.9 9.450 9.550 9.450 9.550 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 9.290 9.384 density_rs2pw 128 9.7 0.006 0.006 7.758 9.137 fft_wrap_pw1pw2_140 523 13.2 1.559 1.595 8.055 8.151 wfi_extrapolate 11 7.9 0.001 0.001 8.095 8.095 grid_integrate_task_list 128 12.3 7.494 7.890 7.494 7.890 rs_pw_transfer 1046 11.9 0.013 0.014 5.816 7.256 multiply_cannon_metrocomm4 25070 15.6 0.079 0.090 2.797 6.726 mp_irecv_dv 76098 16.2 2.651 6.459 2.651 6.459 calculate_dm_sparse 128 9.5 0.001 0.001 6.208 6.297 fft3d_ps 1291 14.7 2.832 2.920 6.220 6.286 mp_alltoall_d11v 2415 14.1 5.211 5.994 5.211 5.994 grid_collocate_task_list 128 9.7 5.153 5.870 5.153 5.870 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.523 5.568 potential_pw2rs 128 12.3 0.023 0.023 5.252 5.271 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.638 4.723 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.409 4.480 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 4.453 4.454 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.352 4.413 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=204.626000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1076.636364, yerr=12.212309 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4047e199d09ff4d3c3dd9bcab791603060e8611a_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 5.865089E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 1960712 0.0% 0.0% 100.0% average stack size 0.0 0.0 3445.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.516503E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 240672 MPI messages size (bytes): total size 1.331455E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.532238E+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 113904 59718500352 4194304 < size <= 16777216 104976 550376570880 16777216 < size 20208 721350232272 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8931 51. MP_Alltoall 9654 799394. MP_ISend 40068 2102573. MP_IRecv 40068 2101676. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4020 57953. MP_Allreduce 11127 1165. MP_Sync 87 MP_Alltoall 1712 18838222. MP_SendRecv 3840 122880. MP_ISendRecv 3840 122880. MP_Wait 16122 MP_ISend 10680 423556. MP_IRecv 10680 423556. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.033 175.715 175.715 qs_mol_dyn_low 1 2.0 0.003 0.003 175.100 175.112 qs_forces 11 3.9 0.004 0.004 174.959 174.967 qs_energies 11 4.9 0.001 0.001 167.360 167.370 scf_env_do_scf 11 5.9 0.001 0.001 149.438 149.448 velocity_verlet 10 3.0 0.001 0.001 114.830 114.835 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 113.332 113.333 dbcsr_multiply_generic 2507 12.6 0.182 0.194 74.563 74.942 qs_scf_new_mos 117 7.6 0.001 0.001 74.504 74.605 qs_scf_loop_do_ot 117 8.6 0.001 0.001 74.504 74.604 ot_scf_mini 117 9.6 0.003 0.004 70.067 70.146 multiply_cannon 2507 13.6 0.580 0.647 54.610 58.804 multiply_cannon_loop 2507 14.6 0.445 0.457 49.964 50.684 ot_mini 117 10.6 0.001 0.001 39.577 39.646 init_scf_loop 11 6.9 0.000 0.000 35.953 35.955 mp_waitall_1 125778 16.7 25.934 33.377 25.934 33.377 prepare_preconditioner 11 7.9 0.000 0.000 32.012 32.037 make_preconditioner 11 8.9 0.000 0.000 32.012 32.037 rebuild_ks_matrix 128 8.3 0.001 0.001 30.059 30.154 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.018 30.058 30.153 make_full_inverse_cholesky 11 9.9 0.014 0.028 29.807 30.058 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.257 27.350 multiply_cannon_multrec 10028 15.6 10.414 14.476 18.025 21.194 ot_diis_step 117 11.6 0.020 0.020 19.803 19.803 qs_ot_get_derivative 117 11.6 0.002 0.002 19.708 19.790 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.385 19.641 apply_single 128 13.6 0.001 0.001 19.385 19.641 multiply_cannon_metrocomm3 10028 15.6 0.023 0.025 12.125 19.108 make_m2s 5014 13.6 0.066 0.070 16.286 18.523 make_images 5014 14.6 2.289 2.800 15.976 18.205 cp_fm_cholesky_invert 11 10.9 18.137 18.142 18.137 18.142 qs_ot_get_p 128 10.4 0.001 0.001 17.585 17.657 qs_rho_update_rho_low 128 7.7 0.001 0.001 15.488 15.524 calculate_rho_elec 128 8.7 0.256 0.267 15.487 15.524 sum_up_and_integrate 128 10.3 0.181 0.191 15.115 15.168 integrate_v_rspace 128 11.3 0.004 0.004 14.933 14.997 qs_ot_p2m_diag 83 11.4 0.495 0.500 13.860 13.876 cp_dbcsr_syevd 83 12.4 0.005 0.006 12.731 12.732 make_images_data 5014 15.6 0.050 0.060 9.963 12.583 multiply_cannon_sync_h2d 10028 15.6 11.618 12.227 11.618 12.227 hybrid_alltoall_any 5200 16.5 0.820 3.623 9.766 12.084 init_scf_run 11 5.9 0.000 0.001 11.044 11.046 scf_env_initial_rho_setup 11 6.9 0.001 0.004 11.044 11.045 cp_fm_diag_elpa 83 13.4 0.000 0.000 9.690 9.702 cp_fm_diag_elpa_base 83 14.4 9.447 9.523 9.684 9.696 pw_transfer 1547 11.6 0.086 0.097 9.555 9.586 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 9.331 9.367 fft_wrap_pw1pw2_140 523 13.2 1.914 1.968 8.253 8.295 grid_integrate_task_list 128 12.3 7.761 8.268 7.761 8.268 cp_fm_cholesky_decompose 22 10.9 8.089 8.188 8.089 8.188 dbcsr_mm_accdrv_process 20762 16.1 2.743 3.524 7.237 8.038 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 7.879 7.940 density_rs2pw 128 9.7 0.006 0.006 6.971 7.856 wfi_extrapolate 11 7.9 0.001 0.001 7.737 7.737 multiply_cannon_metrocomm1 10028 15.6 0.028 0.029 4.570 7.694 mp_allgather_i34 2507 14.6 2.844 7.253 2.844 7.253 mp_alltoall_d11v 2415 14.1 5.359 6.458 5.359 6.458 calculate_dm_sparse 128 9.5 0.001 0.001 6.183 6.253 grid_collocate_task_list 128 9.7 5.466 6.185 5.466 6.185 fft3d_ps 1291 14.7 2.761 2.829 5.810 5.856 dbcsr_complete_redistribute 395 12.7 2.126 2.206 5.391 5.757 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.543 5.582 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.305 5.306 rs_pw_transfer 1046 11.9 0.012 0.013 4.203 5.118 potential_pw2rs 128 12.3 0.026 0.026 4.454 4.468 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.390 4.422 multiply_cannon_metrocomm4 7521 15.6 0.023 0.026 1.843 4.209 mp_irecv_dv 28860 15.9 1.807 4.139 1.807 4.139 copy_fm_to_dbcsr 209 11.7 0.001 0.002 3.514 3.894 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.564 3.838 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.644 3.767 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.709 3.718 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.687 3.716 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=175.715000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1412.909091, yerr=54.636742 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4047e199d09ff4d3c3dd9bcab791603060e8611a_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 3.131769E+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 3997 59213. MP_Allreduce 11070 1504. MP_Sync 86 MP_Alltoall 1700 36954383. MP_SendRecv 1778 218624. MP_ISendRecv 1778 218624. MP_Wait 9728 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 0.029 0.051 299.135 299.136 qs_mol_dyn_low 1 2.0 0.003 0.004 298.333 298.345 qs_forces 11 3.9 0.012 0.024 298.238 298.241 qs_energies 11 4.9 0.001 0.001 288.869 288.889 scf_env_do_scf 11 5.9 0.001 0.001 266.037 266.053 velocity_verlet 10 3.0 0.001 0.001 213.967 213.976 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 141.402 141.403 init_scf_loop 11 6.9 0.000 0.000 124.382 124.383 prepare_preconditioner 11 7.9 0.000 0.000 119.259 119.282 make_preconditioner 11 8.9 0.000 0.000 119.259 119.282 make_full_inverse_cholesky 11 9.9 0.036 0.039 95.182 116.366 qs_scf_new_mos 116 7.6 0.001 0.001 91.085 91.167 qs_scf_loop_do_ot 116 8.6 0.001 0.001 91.084 91.166 ot_scf_mini 116 9.6 0.004 0.004 86.324 86.364 dbcsr_multiply_generic 2485 12.5 0.211 0.224 84.732 85.363 cp_fm_upper_to_full 104 14.8 52.455 75.416 52.455 75.416 multiply_cannon 2485 13.5 0.699 0.747 59.512 60.371 multiply_cannon_loop 2485 14.5 0.467 0.475 55.565 57.139 ot_mini 116 10.6 0.001 0.001 45.233 45.267 dbcsr_complete_redistribute 393 12.7 3.991 4.031 29.675 42.809 copy_fm_to_dbcsr 208 11.6 0.001 0.002 26.332 39.512 rebuild_ks_matrix 127 8.3 0.001 0.001 38.127 38.214 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.017 38.127 38.214 transfer_fm_to_dbcsr 11 9.9 0.027 0.032 24.030 37.061 mp_alltoall_i22 712 14.1 21.812 35.184 21.812 35.184 qs_ks_update_qs_env 127 7.6 0.001 0.001 35.096 35.181 cp_fm_cholesky_invert 11 10.9 33.507 33.513 33.507 33.513 mp_waitall_1 102768 16.8 29.005 32.746 29.005 32.746 qs_ot_get_p 127 10.4 0.001 0.001 25.602 25.680 qs_ot_get_derivative 116 11.6 0.002 0.002 24.510 24.553 make_m2s 4970 13.5 0.074 0.076 20.549 21.786 qs_ot_p2m_diag 82 11.4 0.867 0.872 21.534 21.563 make_images 4970 14.5 3.743 3.919 20.070 21.311 ot_diis_step 116 11.6 0.022 0.022 20.688 20.689 qs_rho_update_rho_low 127 7.7 0.001 0.001 20.389 20.398 calculate_rho_elec 127 8.7 0.477 0.478 20.389 20.397 multiply_cannon_metrocomm3 9940 15.5 0.023 0.025 18.874 20.284 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.945 20.172 apply_single 127 13.6 0.001 0.001 19.945 20.172 cp_dbcsr_syevd 82 12.4 0.006 0.006 19.821 19.824 sum_up_and_integrate 127 10.3 0.319 0.321 19.543 19.618 integrate_v_rspace 127 11.3 0.004 0.004 19.223 19.298 multiply_cannon_multrec 9940 15.5 10.488 12.381 18.513 18.599 cp_fm_diag_elpa 82 13.4 0.000 0.000 16.657 16.658 cp_fm_diag_elpa_base 82 14.4 12.376 13.934 16.653 16.653 multiply_cannon_sync_h2d 9940 15.5 15.528 15.538 15.528 15.538 hybrid_alltoall_any 5155 16.4 1.292 3.037 11.394 13.428 make_images_data 4970 15.5 0.057 0.061 11.308 13.351 init_scf_run 11 5.9 0.000 0.001 12.909 12.909 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.908 12.909 pw_transfer 1535 11.6 0.092 0.094 11.380 11.386 fft_wrap_pw1pw2 1281 12.7 0.011 0.011 11.146 11.150 fft_wrap_pw1pw2_140 519 13.2 3.053 3.083 9.896 9.905 dbcsr_mm_accdrv_process 20590 16.0 3.939 5.819 7.787 9.652 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.459 9.499 mp_alltoall_d11v 2401 14.1 8.297 9.213 8.297 9.213 wfi_extrapolate 11 7.9 0.001 0.001 9.159 9.159 cp_fm_cholesky_decompose 22 10.9 8.864 8.924 8.864 8.924 grid_integrate_task_list 127 12.3 8.526 8.703 8.526 8.703 density_rs2pw 127 9.7 0.005 0.006 7.953 8.127 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 7.922 7.931 calculate_dm_sparse 127 9.5 0.001 0.001 7.001 7.062 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.526 6.577 grid_collocate_task_list 127 9.7 6.390 6.458 6.390 6.458 fft3d_ps 1281 14.7 2.822 2.833 6.259 6.287 rs_scatter_matrices 138 9.7 3.665 4.482 6.016 6.253 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.019 6.075 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=299.136000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2793.000000, yerr=163.261642 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4047e199d09ff4d3c3dd9bcab791603060e8611a_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 1.259250E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255669. MP_Allreduce 3139 6114. MP_Sync 4 MP_Alltoall 54 MP_SendRecv 285 19200. MP_ISendRecv 285 19200. MP_Wait 1017 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.031 85.086 85.087 qs_energies 1 2.0 0.000 0.000 84.584 84.590 ls_scf 1 3.0 0.000 0.000 83.697 83.702 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.710 72.870 multiply_cannon 111 7.7 0.017 0.020 55.858 57.246 multiply_cannon_loop 111 8.7 0.207 0.221 52.424 54.009 ls_scf_main 1 4.0 0.000 0.000 52.361 52.373 density_matrix_trs4 2 5.0 0.002 0.003 46.811 46.896 ls_scf_init_scf 1 4.0 0.000 0.000 28.291 28.292 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.158 27.209 mp_waitall_1 11031 10.9 22.545 26.012 22.545 26.012 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.066 25.082 multiply_cannon_multrec 2664 9.7 8.157 8.915 15.551 17.282 multiply_cannon_sync_h2d 2664 9.7 13.514 15.313 13.514 15.313 make_m2s 222 7.7 0.008 0.011 13.053 13.593 make_images 222 8.7 0.099 0.110 13.031 13.573 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 9.730 12.714 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.490 8.584 make_images_data 222 9.7 0.004 0.005 7.659 8.158 dbcsr_mm_accdrv_process 4760 10.4 0.506 0.615 7.014 8.011 hybrid_alltoall_any 227 10.6 0.215 1.834 6.549 7.800 dbcsr_mm_accdrv_process_sort 4760 11.4 6.305 7.212 6.305 7.212 calculate_norms 4752 9.8 5.484 6.164 5.484 6.164 mp_sum_l 887 5.1 3.261 5.230 3.261 5.230 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.032 5.162 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.499 3.982 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.060 3.573 mp_irecv_dv 6231 10.9 2.043 3.551 2.043 3.551 make_images_sizes 222 9.7 0.000 0.000 0.690 3.364 mp_alltoall_i44 222 10.7 0.689 3.364 0.689 3.364 arnoldi_extremal 4 6.8 0.000 0.000 3.222 3.250 arnoldi_normal_ev 4 7.8 0.001 0.003 3.222 3.250 build_subspace 16 8.4 0.009 0.012 3.128 3.131 ls_scf_post 1 4.0 0.000 0.000 3.045 3.051 ls_scf_store_result 1 5.0 0.000 0.000 2.865 2.907 dbcsr_special_finalize 555 9.7 0.005 0.006 2.353 2.785 dbcsr_merge_single_wm 555 10.7 0.459 0.601 2.344 2.776 make_images_pack 222 9.7 2.205 2.622 2.207 2.624 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.332 2.563 dbcsr_sort_data 658 11.4 2.142 2.479 2.142 2.479 dbcsr_matrix_vector_mult_local 304 10.0 2.070 2.462 2.072 2.464 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.328 2.420 buffer_matrices_ensure_size 222 8.7 1.758 2.089 1.758 2.089 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.786 1.788 rebuild_ks_matrix 3 7.3 0.000 0.000 1.777 1.778 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.777 1.778 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.087000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1131.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4047e199d09ff4d3c3dd9bcab791603060e8611a_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 2.102469E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3138 10075. MP_Sync 4 MP_Alltoall 47 15335933. MP_SendRecv 141 57600. MP_ISendRecv 141 57600. MP_Wait 687 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.028 0.041 91.723 91.724 qs_energies 1 2.0 0.000 0.000 91.274 91.279 ls_scf 1 3.0 0.000 0.000 89.957 89.962 dbcsr_multiply_generic 111 6.7 0.015 0.016 75.981 76.291 multiply_cannon 111 7.7 0.028 0.040 53.544 57.466 ls_scf_main 1 4.0 0.000 0.000 55.799 55.803 multiply_cannon_loop 111 8.7 0.115 0.122 50.225 53.699 density_matrix_trs4 2 5.0 0.002 0.003 50.034 50.234 mp_waitall_1 9105 10.9 21.572 30.775 21.572 30.775 ls_scf_init_scf 1 4.0 0.000 0.000 30.650 30.650 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.439 29.533 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.048 27.058 multiply_cannon_multrec 1332 9.7 13.107 16.370 22.412 26.839 multiply_cannon_metrocomm3 1332 9.7 0.006 0.007 11.954 21.033 make_m2s 222 7.7 0.006 0.007 15.599 16.283 make_images 222 8.7 1.581 1.934 15.569 16.253 dbcsr_mm_accdrv_process 4041 10.4 0.282 0.451 8.903 10.509 dbcsr_mm_accdrv_process_sort 4041 11.4 8.498 10.060 8.498 10.060 make_images_data 222 9.7 0.004 0.004 9.000 9.884 hybrid_alltoall_any 227 10.6 0.520 2.442 8.397 9.638 mp_sum_l 887 5.1 5.527 8.785 5.527 8.785 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.203 7.635 mp_irecv_dv 3311 11.0 3.183 7.578 3.183 7.578 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.265 6.812 calculate_norms 2376 9.8 5.988 6.656 5.988 6.656 multiply_cannon_sync_h2d 1332 9.7 4.850 5.925 4.850 5.925 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.040 5.249 arnoldi_extremal 4 6.8 0.000 0.000 4.640 4.655 arnoldi_normal_ev 4 7.8 0.001 0.005 4.639 4.655 build_subspace 16 8.4 0.014 0.021 4.384 4.386 ls_scf_post 1 4.0 0.000 0.000 3.508 3.513 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.145 3.380 ls_scf_store_result 1 5.0 0.000 0.000 3.223 3.332 dbcsr_matrix_vector_mult_local 304 10.0 2.750 3.242 2.751 3.244 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.249 3.184 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.632 2.712 mp_allgather_i34 111 8.7 1.009 2.522 1.009 2.522 make_images_pack 222 9.7 2.023 2.438 2.026 2.440 dbcsr_sort_data 436 11.2 1.833 2.081 1.833 2.081 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.877 1.879 rebuild_ks_matrix 3 7.3 0.000 0.000 1.864 1.866 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.864 1.866 dbcsr_data_new 4174 10.1 1.615 1.855 1.615 1.855 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.724000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1748.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4047e199d09ff4d3c3dd9bcab791603060e8611a_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 2.683269E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265470. MP_Allreduce 3138 10896. MP_Sync 4 MP_Alltoall 47 23526250. MP_SendRecv 93 57600. MP_ISendRecv 93 57600. MP_Wait 639 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.034 0.056 93.419 93.420 qs_energies 1 2.0 0.000 0.000 92.771 92.774 ls_scf 1 3.0 0.000 0.001 91.349 91.353 dbcsr_multiply_generic 111 6.7 0.016 0.016 76.023 76.241 ls_scf_main 1 4.0 0.000 0.001 57.139 57.143 multiply_cannon 111 7.7 0.035 0.072 52.350 56.555 multiply_cannon_loop 111 8.7 0.100 0.107 48.802 52.232 density_matrix_trs4 2 5.0 0.002 0.004 51.086 51.309 mp_waitall_1 7281 11.0 23.308 32.658 23.308 32.658 ls_scf_init_scf 1 4.0 0.000 0.002 30.648 30.651 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.440 29.515 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 27.052 27.074 multiply_cannon_multrec 888 9.7 12.627 15.404 21.242 24.744 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.647 21.862 make_m2s 222 7.7 0.006 0.007 17.156 18.333 make_images 222 8.7 1.970 2.278 17.117 18.295 make_images_data 222 9.7 0.003 0.004 9.756 10.807 hybrid_alltoall_any 227 10.6 0.620 2.881 9.377 10.805 dbcsr_mm_accdrv_process 3754 10.4 0.234 0.420 8.163 9.431 dbcsr_mm_accdrv_process_sort 3754 11.4 7.793 9.011 7.793 9.011 mp_sum_l 887 5.1 4.978 7.868 4.978 7.868 multiply_cannon_sync_h2d 888 9.7 6.045 7.333 6.045 7.333 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.525 7.014 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.431 6.803 mp_irecv_dv 2335 11.1 2.417 6.755 2.417 6.755 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.767 6.132 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.971 5.177 arnoldi_extremal 4 6.8 0.000 0.000 5.062 5.083 arnoldi_normal_ev 4 7.8 0.001 0.005 5.062 5.083 build_subspace 16 8.4 0.014 0.020 4.759 4.768 calculate_norms 1584 9.8 4.301 4.597 4.301 4.597 mp_allgather_i34 111 8.7 1.385 3.790 1.385 3.790 dbcsr_matrix_vector_mult 304 9.0 0.009 0.021 3.419 3.756 dbcsr_matrix_vector_mult_local 304 10.0 3.003 3.578 3.005 3.580 ls_scf_post 1 4.0 0.000 0.001 3.562 3.565 ls_scf_store_result 1 5.0 0.000 0.000 3.290 3.383 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.874 2.979 make_images_sizes 222 9.7 0.000 0.000 1.083 2.226 mp_alltoall_i44 222 10.7 1.082 2.226 1.082 2.226 dbcsr_sort_data 325 11.1 1.871 2.134 1.871 2.134 make_images_pack 222 9.7 1.812 2.104 1.815 2.107 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.913 1.915 rebuild_ks_matrix 3 7.3 0.000 0.000 1.895 1.896 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.895 1.896 dbcsr_data_release 9322 10.9 1.286 1.886 1.286 1.886 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=93.420000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2190.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4047e199d09ff4d3c3dd9bcab791603060e8611a_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 3.394953E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3138 13030. MP_Sync 4 MP_Alltoall 47 30278988. MP_SendRecv 69 86400. MP_ISendRecv 69 86400. MP_Wait 531 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.033 0.047 97.197 97.198 qs_energies 1 2.0 0.000 0.000 96.574 96.580 ls_scf 1 3.0 0.000 0.000 94.911 94.918 dbcsr_multiply_generic 111 6.7 0.017 0.017 78.457 78.718 ls_scf_main 1 4.0 0.000 0.000 58.713 58.714 multiply_cannon 111 7.7 0.062 0.159 51.851 56.752 density_matrix_trs4 2 5.0 0.002 0.003 52.528 52.627 multiply_cannon_loop 111 8.7 0.114 0.125 46.800 49.854 ls_scf_init_scf 1 4.0 0.000 0.000 32.943 32.944 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.695 31.779 mp_waitall_1 6369 11.0 22.693 29.419 22.693 29.419 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.246 29.256 multiply_cannon_multrec 1332 9.7 14.207 17.575 22.224 25.675 make_m2s 222 7.7 0.006 0.007 20.970 22.440 make_images 222 8.7 3.146 3.595 20.920 22.392 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.335 17.887 make_images_data 222 9.7 0.004 0.004 11.654 13.297 hybrid_alltoall_any 227 10.6 0.793 3.736 10.999 12.891 dbcsr_mm_accdrv_process 3641 10.4 0.275 0.811 7.647 9.178 dbcsr_mm_accdrv_process_sort 3641 11.4 7.225 8.687 7.225 8.687 mp_sum_l 887 5.1 4.067 7.079 4.067 7.079 multiply_cannon_sync_h2d 1332 9.7 5.524 6.126 5.524 6.126 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.064 5.931 mp_irecv_dv 3229 10.9 2.041 5.851 2.041 5.851 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.168 5.564 arnoldi_extremal 4 6.8 0.000 0.000 5.322 5.335 arnoldi_normal_ev 4 7.8 0.001 0.005 5.322 5.335 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.527 5.013 build_subspace 16 8.4 0.014 0.021 4.971 4.982 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.535 4.696 mp_allgather_i34 111 8.7 2.160 4.580 2.160 4.580 calculate_norms 2376 9.8 4.191 4.512 4.191 4.512 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.630 3.898 dbcsr_matrix_vector_mult_local 304 10.0 3.156 3.646 3.158 3.647 dbcsr_sort_data 658 11.4 3.071 3.396 3.071 3.396 ls_scf_post 1 4.0 0.000 0.000 3.255 3.260 dbcsr_special_finalize 555 9.7 0.006 0.007 2.832 3.214 dbcsr_merge_single_wm 555 10.7 0.538 0.670 2.824 3.205 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.076 3.108 ls_scf_store_result 1 5.0 0.000 0.000 2.998 3.061 dbcsr_data_release 10477 10.7 1.593 2.403 1.593 2.403 dbcsr_finalize 304 7.8 0.049 0.061 1.791 1.978 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.954 1.956 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.198000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2717.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4047e199d09ff4d3c3dd9bcab791603060e8611a_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 4.711481E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265558. MP_Allreduce 3129 15263. MP_Sync 4 MP_Alltoall 47 46208988. MP_SendRecv 45 115200. MP_ISendRecv 45 115200. MP_Wait 528 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.042 0.055 99.254 99.255 qs_energies 1 2.0 0.000 0.000 98.446 98.452 ls_scf 1 3.0 0.000 0.000 96.537 96.548 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.013 78.217 ls_scf_main 1 4.0 0.000 0.000 62.126 62.127 multiply_cannon 111 7.7 0.064 0.190 55.359 60.982 density_matrix_trs4 2 5.0 0.002 0.002 54.926 55.025 multiply_cannon_loop 111 8.7 0.069 0.076 50.978 52.651 mp_waitall_1 5436 11.0 26.723 32.788 26.723 32.788 ls_scf_init_scf 1 4.0 0.000 0.000 30.795 30.800 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.486 29.544 matrix_sqrt_Newton_Schulz 2 6.5 0.002 0.004 27.344 27.354 multiply_cannon_multrec 444 9.7 13.905 16.517 21.040 23.111 make_m2s 222 7.7 0.004 0.005 17.863 20.208 make_images 222 8.7 3.723 4.403 17.801 20.147 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.096 16.349 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.389 14.793 make_images_data 222 9.7 0.003 0.004 10.040 12.410 hybrid_alltoall_any 227 10.6 0.790 3.764 9.875 12.326 dbcsr_mm_accdrv_process 3003 10.4 0.183 0.378 6.840 7.984 dbcsr_mm_accdrv_process_sort 3003 11.4 6.515 7.640 6.515 7.640 multiply_cannon_sync_h2d 444 9.7 6.545 7.516 6.545 7.516 mp_allgather_i34 111 8.7 2.611 6.980 2.611 6.980 arnoldi_extremal 4 6.8 0.000 0.000 5.749 5.754 arnoldi_normal_ev 4 7.8 0.002 0.005 5.749 5.754 build_subspace 16 8.4 0.015 0.019 5.363 5.374 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.569 4.747 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.712 4.456 mp_sum_l 887 5.1 2.827 4.454 2.827 4.454 mp_irecv_dv 1241 11.2 1.694 4.431 1.694 4.431 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.103 4.319 dbcsr_matrix_vector_mult_local 304 10.0 3.635 4.123 3.637 4.125 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.604 3.699 calculate_norms 792 9.8 3.561 3.672 3.561 3.672 ls_scf_post 1 4.0 0.000 0.000 3.616 3.621 make_images_sizes 222 9.7 0.000 0.000 1.168 3.521 mp_alltoall_i44 222 10.7 1.168 3.521 1.168 3.521 ls_scf_store_result 1 5.0 0.000 0.000 3.396 3.438 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.928 3.146 dbcsr_finalize 304 7.8 0.062 0.077 2.202 2.278 dbcsr_merge_all 275 8.9 0.472 0.524 2.051 2.117 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.084 2.086 rebuild_ks_matrix 3 7.3 0.000 0.000 2.052 2.054 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.052 2.054 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=99.255000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3670.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4047e199d09ff4d3c3dd9bcab791603060e8611a_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 8.694411E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284111. MP_Allreduce 3123 21388. MP_Sync 4 MP_Alltoall 47 88727262. MP_SendRecv 42 732600. MP_ISendRecv 42 732600. MP_Wait 267 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.086 0.119 108.008 108.009 qs_energies 1 2.0 0.000 0.000 106.583 106.602 ls_scf 1 3.0 0.000 0.000 103.678 103.696 dbcsr_multiply_generic 111 6.7 0.023 0.028 77.475 77.605 ls_scf_main 1 4.0 0.000 0.000 65.857 65.858 density_matrix_trs4 2 5.0 0.002 0.003 56.649 56.701 multiply_cannon 111 7.7 0.162 0.358 49.986 52.019 multiply_cannon_loop 111 8.7 0.067 0.070 46.492 47.432 ls_scf_init_scf 1 4.0 0.000 0.000 34.133 34.133 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.555 32.576 matrix_sqrt_Newton_Schulz 2 6.5 0.002 0.005 29.736 29.742 mp_waitall_1 4527 11.1 22.223 25.986 22.223 25.986 make_m2s 222 7.7 0.005 0.005 23.888 24.921 make_images 222 8.7 4.592 5.003 23.782 24.812 multiply_cannon_multrec 444 9.7 17.819 18.594 22.536 23.247 make_images_data 222 9.7 0.003 0.003 13.126 15.597 hybrid_alltoall_any 227 10.6 1.662 3.634 12.928 15.538 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.487 10.938 multiply_cannon_sync_h2d 444 9.7 8.843 8.891 8.843 8.891 arnoldi_extremal 4 6.8 0.000 0.000 7.172 7.184 arnoldi_normal_ev 4 7.8 0.003 0.009 7.172 7.184 build_subspace 16 8.4 0.026 0.036 6.624 6.638 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.312 5.442 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.196 5.281 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.978 5.231 dbcsr_matrix_vector_mult_local 304 10.0 4.897 5.181 4.899 5.184 dbcsr_mm_accdrv_process 1814 10.4 0.331 0.436 4.547 4.713 dbcsr_mm_accdrv_process_sort 1814 11.4 4.191 4.329 4.191 4.329 ls_scf_post 1 4.0 0.000 0.000 3.688 3.706 mp_allgather_i34 111 8.7 1.146 3.637 1.146 3.637 make_images_sizes 222 9.7 0.000 0.000 1.422 3.484 mp_alltoall_i44 222 10.7 1.422 3.483 1.422 3.483 ls_scf_store_result 1 5.0 0.000 0.000 3.413 3.422 calculate_norms 792 9.8 3.245 3.279 3.245 3.279 dbcsr_finalize 304 7.8 0.082 0.089 3.085 3.183 dbcsr_merge_all 275 8.9 0.891 0.918 2.870 2.963 dbcsr_complete_redistribute 5 7.6 1.447 1.481 2.778 2.893 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.874 2.874 dbcsr_data_release 12724 10.6 2.327 2.852 2.327 2.852 matrix_ls_to_qs 2 6.0 0.000 0.000 2.413 2.542 dbcsr_sort_data 325 11.1 2.443 2.519 2.443 2.519 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.466 2.468 rebuild_ks_matrix 3 7.3 0.000 0.000 2.400 2.401 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.400 2.401 dbcsr_new_transposed 4 7.5 0.243 0.252 2.264 2.286 dbcsr_frobenius_norm 74 6.6 2.052 2.125 2.188 2.231 dbcsr_add_d 103 6.2 0.000 0.000 2.131 2.202 dbcsr_add_anytype 103 7.2 0.859 0.892 2.131 2.202 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=108.009000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6811.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/4047e199d09ff4d3c3dd9bcab791603060e8611a_performance_tests/27/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 7009386627072 0.0% 0.0% 100.0% flops 9 x 9 x 32 7335108845568 0.0% 0.0% 100.0% flops 9 x 22 x 32 9866241589248 0.0% 0.0% 100.0% flops 22 x 9 x 32 9884108906496 0.0% 0.0% 100.0% flops 22 x 22 x 32 13354440523776 0.0% 0.0% 100.0% flops 32 x 32 x 9 20607185977344 0.0% 0.0% 100.0% flops 32 x 32 x 22 25186560638976 0.0% 0.0% 100.0% flops 9 x 32 x 32 28458319085568 0.0% 0.0% 100.0% flops 22 x 32 x 32 34782389993472 0.0% 0.0% 100.0% flops 9 x 32 x 9 42881542373376 0.0% 0.0% 100.0% flops 22 x 32 x 9 55680402235392 0.0% 0.0% 100.0% flops 9 x 32 x 22 55680402235392 0.0% 0.0% 100.0% flops 22 x 32 x 22 72328573419520 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 383.054662E+12 0.0% 0.0% 100.0% flops max/rank 733.641090E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 26899403712 0.0% 0.0% 100.0% number of processed stacks 118860288 0.0% 0.0% 100.0% average stack size 0.0 0.0 226.3 marketing flops 780.439111E+12 ------------------------------------------------------------------------------- # multiplications 1445 max memory usage/rank 590.127104E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 102097920 MPI messages size (bytes): total size 37.227590E+12 min size 0.000000E+00 max size 4.551360E+06 average size 364.626312E+03 MPI breakdown and total messages size (bytes): size <= 128 731472 0 128 < size <= 8192 11922720 97670922240 8192 < size <= 32768 24718992 614677610496 32768 < size <= 131072 20000256 1970081366016 131072 < size <= 4194304 42515668 24886801223040 4194304 < size <= 16777216 2208812 9656099886720 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4640 76559. MP_Allreduce 13232 2081. MP_Sync 1064 MP_Alltoall 2588 929406232. MP_SendRecv 168740 11136. MP_ISendRecv 92040 11136. MP_Wait 102830 MP_comm_split 40 MP_ISend 26090 85106. MP_IRecv 37890 59644. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.030 0.068 270.474 270.476 qs_mol_dyn_low 1 2.0 0.004 0.026 269.031 269.053 qs_forces 5 3.8 0.014 0.031 268.925 268.927 qs_energies 5 4.8 0.001 0.024 265.605 265.632 scf_env_do_scf 5 5.8 0.001 0.013 248.018 248.022 scf_env_do_scf_inner_loop 105 6.6 0.002 0.013 215.294 215.305 qs_scf_new_mos 105 7.6 0.000 0.001 167.247 167.445 qs_scf_loop_do_ot 105 8.6 0.001 0.001 167.247 167.444 ot_scf_mini 105 9.6 0.003 0.003 157.389 157.546 dbcsr_multiply_generic 1445 12.2 0.130 0.138 136.525 137.140 multiply_cannon 1445 13.2 0.274 0.283 116.275 118.981 multiply_cannon_loop 1445 14.2 2.398 2.513 113.897 115.819 velocity_verlet 4 3.0 0.000 0.001 115.019 115.020 qs_ot_get_p 112 10.4 0.001 0.001 66.159 66.411 ot_mini 105 10.6 0.001 0.001 61.077 61.202 qs_ot_p2m_diag 40 11.0 0.020 0.031 54.477 54.623 cp_dbcsr_syevd 40 12.0 0.002 0.002 51.184 51.187 cp_fm_syevd 40 13.0 0.000 0.000 45.149 45.303 multiply_cannon_multrec 69360 15.2 31.543 36.362 39.475 44.891 mp_waitall_1 488190 16.1 36.253 44.294 36.253 44.294 cp_fm_redistribute_end 40 14.0 19.884 39.668 19.893 39.674 cp_fm_syevd_base 40 14.0 19.764 39.545 19.764 39.545 qs_ot_get_derivative 55 11.6 0.001 0.001 39.393 39.525 multiply_cannon_sync_h2d 69360 15.2 30.444 34.859 30.444 34.859 multiply_cannon_metrocomm3 69360 15.2 0.192 0.201 25.168 34.687 init_scf_loop 7 6.6 0.000 0.018 32.686 32.689 rebuild_ks_matrix 110 8.4 0.000 0.000 32.022 32.234 qs_ks_build_kohn_sham_matrix 110 9.4 0.011 0.014 32.022 32.233 qs_ks_update_qs_env 112 7.6 0.001 0.001 29.387 29.577 prepare_preconditioner 7 7.6 0.000 0.001 27.222 27.251 make_preconditioner 7 8.6 0.000 0.000 27.222 27.251 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 23.026 23.308 apply_single 62 13.6 0.000 0.000 23.026 23.308 qs_rho_update_rho_low 110 7.6 0.001 0.001 22.705 23.033 calculate_rho_elec 110 8.6 0.030 0.034 22.704 23.033 ot_new_cg_direction 55 11.6 0.001 0.001 20.979 20.980 rs_pw_transfer 690 11.5 0.010 0.012 18.127 19.348 make_full_inverse_cholesky 7 9.6 0.000 0.000 19.115 19.179 density_rs2pw 110 9.6 0.006 0.008 16.380 17.490 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.932 15.118 init_scf_run 5 5.8 0.000 0.001 14.789 14.790 scf_env_initial_rho_setup 5 6.8 0.002 0.003 14.789 14.790 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 14.143 14.267 pw_transfer 1645 12.4 0.086 0.115 12.954 13.203 fft_wrap_pw1pw2 1425 13.5 0.014 0.018 12.809 13.064 mp_sum_l 4764 12.2 12.091 12.974 12.091 12.974 cp_fm_cholesky_invert 7 10.6 12.853 12.870 12.853 12.870 calculate_dm_sparse 110 9.5 0.000 0.001 11.512 11.651 sum_up_and_integrate 60 10.3 0.028 0.031 11.182 11.194 integrate_v_rspace 60 11.3 0.001 0.003 11.154 11.167 check_diag 80 13.5 8.586 8.858 10.793 10.942 fft_wrap_pw1pw2_240 915 15.0 1.238 1.338 10.630 10.875 qs_vxc_create 110 10.4 0.002 0.004 10.745 10.809 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 10.712 10.782 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.002 9.833 9.850 acc_transpose_blocks 69360 15.2 0.366 0.383 9.114 9.514 calculate_first_density_matrix 1 7.0 0.000 0.005 9.453 9.470 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 9.305 9.364 fft3d_pb 915 16.0 2.405 2.764 8.637 8.996 dbcsr_mm_accdrv_process 154766 15.8 4.104 4.266 7.801 8.583 make_m2s 2890 13.2 0.079 0.086 7.709 8.327 make_images 2890 14.2 0.240 0.259 7.601 8.217 make_full_single_inverse 7 9.6 0.001 0.001 7.737 7.767 xc_rho_set_and_dset_create 110 12.4 0.076 0.096 7.448 7.705 multiply_cannon_metrocomm1 69360 15.2 0.091 0.097 4.440 7.485 acc_transpose_blocks_kernels 69360 16.2 0.860 0.905 7.161 7.477 xc_vxc_pw_create 60 11.3 0.039 0.049 7.229 7.292 potential_pw2rs 60 12.3 0.003 0.004 7.118 7.166 rs_pw_transfer_RS2PW_30 110 11.6 1.588 1.675 6.453 6.806 jit_kernel_transpose 5 15.0 6.300 6.594 6.300 6.594 xc_pw_derive 510 13.4 0.006 0.007 6.436 6.501 mp_sendrecv_dv 168740 12.6 6.071 6.242 6.071 6.242 mp_alltoall_z22v 2340 17.7 5.692 6.238 5.692 6.238 mp_waitany 7680 13.5 4.445 5.776 4.445 5.776 cp_fm_cholesky_decompose 14 10.2 5.693 5.702 5.693 5.702 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=270.476000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=559.400000, yerr=3.382307 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 4047e199d09ff4d3c3dd9bcab791603060e8611a Summary: empty Status: OK