=== 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: 530234f606c3418e1a725f5149ed409c2f956214 ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, Cray-FFTW 3.3.8.10, # COSMA 2.6.2, ELPA 2022.11.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.1.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.1, # SIRIUS 7.3.2, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Author: Matthias Krack (12.01.2023) # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.1.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.1 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.3.2 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta_prefixed_scalapack.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/530234f606c3418e1a725f5149ed409c2f956214_performance_tests/01 job id: 44377235 --- 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/530234f606c3418e1a725f5149ed409c2f956214_performance_tests/02 job id: 44377236 --- 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/530234f606c3418e1a725f5149ed409c2f956214_performance_tests/03 job id: 44377238 --- 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/530234f606c3418e1a725f5149ed409c2f956214_performance_tests/04 job id: 44377239 --- 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/530234f606c3418e1a725f5149ed409c2f956214_performance_tests/05 job id: 44377241 --- 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/530234f606c3418e1a725f5149ed409c2f956214_performance_tests/06 job id: 44377242 --- 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/530234f606c3418e1a725f5149ed409c2f956214_performance_tests/07 job id: 44377244 --- 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/530234f606c3418e1a725f5149ed409c2f956214_performance_tests/08 job id: 44377245 --- 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/530234f606c3418e1a725f5149ed409c2f956214_performance_tests/09 job id: 44377247 --- 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/530234f606c3418e1a725f5149ed409c2f956214_performance_tests/10 job id: 44377249 --- 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/530234f606c3418e1a725f5149ed409c2f956214_performance_tests/11 job id: 44377250 --- 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/530234f606c3418e1a725f5149ed409c2f956214_performance_tests/12 job id: 44377252 --- 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/530234f606c3418e1a725f5149ed409c2f956214_performance_tests/13 job id: 44377253 --- 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/530234f606c3418e1a725f5149ed409c2f956214_performance_tests/14 job id: 44377254 --- 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/530234f606c3418e1a725f5149ed409c2f956214_performance_tests/15 job id: 44377255 --- 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/530234f606c3418e1a725f5149ed409c2f956214_performance_tests/16 job id: 44377256 --- 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/530234f606c3418e1a725f5149ed409c2f956214_performance_tests/17 job id: 44377257 --- 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/530234f606c3418e1a725f5149ed409c2f956214_performance_tests/18 job id: 44377258 --- 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/530234f606c3418e1a725f5149ed409c2f956214_performance_tests/19 job id: 44377259 --- 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/530234f606c3418e1a725f5149ed409c2f956214_performance_tests/20 job id: 44377260 --- 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/530234f606c3418e1a725f5149ed409c2f956214_performance_tests/21 job id: 44377261 --- 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/530234f606c3418e1a725f5149ed409c2f956214_performance_tests/22 job id: 44377263 --- 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/530234f606c3418e1a725f5149ed409c2f956214_performance_tests/23 job id: 44377265 --- 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/530234f606c3418e1a725f5149ed409c2f956214_performance_tests/24 job id: 44377266 --- 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/530234f606c3418e1a725f5149ed409c2f956214_performance_tests/25 job id: 44377267 --- 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/530234f606c3418e1a725f5149ed409c2f956214_performance_tests/26 job id: 44377268 --- Point --- name: 510 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/1r/12t) --- Point --- name: 511 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ === END TESTS (description) === ===== PLOTS (description) ===== ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2_mem", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md_mem", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md_mem", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md_mem", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls_mem", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" === END PLOTS (description) === ============ RESULTS ============ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/530234f606c3418e1a725f5149ed409c2f956214_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 344 9. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.024 0.040 132.973 132.974 farming_run 1 2.0 132.469 132.471 132.942 132.946 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.465360E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 228 1113141. MP_Allreduce 485 2282278. MP_Sync 27 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 8 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.026 0.046 114.861 114.862 qs_energies 1 2.0 0.000 0.000 114.623 114.628 mp2_main 1 3.0 0.000 0.000 112.749 112.754 mp2_gpw_main 1 4.0 0.024 0.035 111.920 111.925 mp2_ri_gpw_compute_in 1 5.0 0.172 0.173 93.302 93.686 mp2_ri_gpw_compute_in_loop 1 6.0 0.005 0.005 55.385 55.754 mp2_eri_3c_integrate_gpw 272 7.0 0.152 0.166 41.686 46.961 get_2c_integrals 1 6.0 0.000 0.001 37.137 37.731 integrate_v_rspace 273 8.0 0.439 0.450 25.063 30.199 pw_transfer 6555 10.6 0.374 0.389 27.338 27.671 grid_integrate_task_list 273 9.0 20.884 26.502 20.884 26.502 fft_wrap_pw1pw2 5465 11.4 0.045 0.048 25.986 26.362 fft_wrap_pw1pw2_100 2178 12.4 1.186 1.391 23.503 23.882 compute_2c_integrals 1 7.0 0.003 0.003 19.535 19.540 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.849 19.249 mp2_eri_2c_integrate_gpw 1 9.0 2.373 2.425 18.846 19.248 rpa_ri_compute_en 1 5.0 0.002 0.003 18.508 18.629 cp_fm_cholesky_decompose 12 8.2 17.542 18.121 17.542 18.121 cholesky_decomp 1 7.0 0.000 0.000 16.452 17.037 fft3d_s 5443 13.4 16.088 16.466 16.110 16.488 ao_to_mo_and_store_B_mult_1 272 7.0 10.856 15.553 10.856 15.553 calculate_wavefunction 272 8.0 5.435 5.568 12.538 13.190 rpa_num_int 1 6.0 0.000 0.001 10.517 10.526 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.501 10.511 calc_mat_Q 8 8.0 0.000 0.000 9.352 9.451 contract_S_to_Q 8 9.0 0.000 0.000 8.773 8.868 calc_potential_gpw 544 9.5 0.005 0.006 8.278 8.676 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.207 8.499 parallel_gemm_fm 14 9.1 0.000 0.000 8.355 8.428 parallel_gemm_fm_cosma 14 10.1 8.355 8.428 8.355 8.428 potential_pw2rs 545 10.0 0.107 0.109 7.652 8.302 collocate_single_gaussian 272 10.0 0.039 0.041 7.444 7.699 create_integ_mat 1 6.0 0.014 0.028 7.505 7.515 array2fm 1 7.0 0.000 0.000 6.612 6.988 pw_scatter_s 2720 13.7 4.422 4.598 4.422 4.598 pw_gather_s 2722 13.2 3.876 4.223 3.876 4.223 array2fm_buffer_send 1 8.0 2.931 3.155 2.931 3.155 pw_poisson_solve 545 10.5 1.114 1.148 2.174 2.411 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.925441, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2737.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/530234f606c3418e1a725f5149ed409c2f956214_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 205321. MP_Allreduce 344 10. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.026 0.038 394.776 394.779 farming_run 1 2.0 394.036 394.040 394.742 394.745 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827141120 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788822 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.223438E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 703 408373. MP_Allreduce 1821 23730. MP_Sync 38 MP_Alltoall 77 27779474. 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 0.012 0.046 209.141 209.143 qs_energies 1 2.0 0.000 0.000 208.917 208.925 scf_env_do_scf 1 3.0 0.000 0.000 105.949 105.949 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.060 105.068 rebuild_ks_matrix 4 6.0 0.000 0.000 105.058 105.067 qs_ks_build_kohn_sham_matrix 4 7.0 0.056 0.065 105.058 105.067 hfx_ks_matrix 4 8.0 0.001 0.001 104.670 104.676 integrate_four_center 4 9.0 0.144 0.452 104.670 104.676 mp2_main 1 3.0 0.000 0.000 102.674 102.683 mp2_gpw_main 1 4.0 0.033 0.048 101.739 101.750 integrate_four_center_main 4 10.0 0.063 0.423 96.612 99.190 integrate_four_center_bin 264 11.0 96.549 99.079 96.549 99.079 init_scf_loop 1 4.0 0.000 0.000 91.803 91.804 mp2_ri_gpw_compute_in 1 5.0 0.063 0.064 75.059 76.106 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.593 55.639 mp2_eri_3c_integrate_gpw 91 7.0 0.143 0.158 42.035 47.394 integrate_v_rspace 95 8.0 0.399 0.571 28.402 33.583 pw_transfer 2240 10.6 0.143 0.161 29.955 30.421 ao_to_mo_and_store_B_mult_1 91 7.0 10.872 29.939 10.872 29.939 fft_wrap_pw1pw2 1868 11.4 0.018 0.020 28.963 29.440 grid_integrate_task_list 95 9.0 23.680 29.073 23.680 29.073 mp2_ri_gpw_compute_en 1 5.0 0.054 0.063 26.531 28.222 fft_wrap_pw1pw2_100 730 12.4 1.294 1.435 26.679 27.141 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.833 1.918 24.841 24.851 get_2c_integrals 1 6.0 0.000 0.000 20.368 20.401 compute_2c_integrals 1 7.0 0.002 0.003 19.349 19.356 compute_2c_integrals_loop_lm 1 8.0 0.001 0.002 18.907 19.219 mp2_eri_2c_integrate_gpw 1 9.0 1.736 1.893 18.906 19.217 fft3d_s 1823 13.4 18.443 18.885 18.456 18.898 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.144 14.144 calculate_wavefunction 91 8.0 2.028 2.065 9.755 9.946 mp2_ri_gpw_compute_en_expansio 172 7.0 0.558 0.598 8.752 9.252 potential_pw2rs 186 10.0 0.033 0.035 8.676 9.205 local_gemm 172 8.0 8.195 8.688 8.195 8.688 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.274 8.609 mp2_ri_gpw_compute_en_comm 22 7.0 0.501 0.516 7.866 8.514 collocate_single_gaussian 91 10.0 0.017 0.020 7.900 8.161 calc_potential_gpw 182 9.5 0.002 0.002 7.912 8.158 mp_sendrecv_dm3 2068 8.0 5.898 6.540 5.898 6.540 mp2_ri_gpw_compute_en_ener 172 7.0 6.345 6.429 6.345 6.429 pw_gather_s 912 13.2 4.907 5.410 4.907 5.410 mp_sync 38 10.4 2.754 5.363 2.754 5.363 pw_scatter_s 910 13.7 3.923 4.414 3.923 4.414 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.722329, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1511.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/530234f606c3418e1a725f5149ed409c2f956214_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 451.883008E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62385. MP_Allreduce 10249 271. MP_Sync 580 MP_Alltoall 2083 578641. 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 0.009 0.026 51.446 51.447 qs_mol_dyn_low 1 2.0 0.003 0.012 51.250 51.259 qs_forces 11 3.9 0.002 0.002 51.186 51.186 qs_energies 11 4.9 0.004 0.004 49.708 49.724 scf_env_do_scf 11 5.9 0.000 0.001 43.823 43.823 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 41.786 41.787 dbcsr_multiply_generic 2286 12.5 0.093 0.096 32.763 33.190 qs_scf_new_mos 108 7.5 0.000 0.001 31.670 31.988 qs_scf_loop_do_ot 108 8.5 0.000 0.001 31.669 31.988 ot_scf_mini 108 9.5 0.002 0.002 30.018 30.222 multiply_cannon 2286 13.5 0.186 0.194 25.546 27.146 multiply_cannon_loop 2286 14.5 1.486 1.570 24.879 26.503 velocity_verlet 10 3.0 0.001 0.001 25.526 25.527 ot_mini 108 10.5 0.001 0.001 18.924 19.167 qs_ot_get_derivative 108 11.5 0.001 0.001 15.954 16.162 mp_waitall_1 245248 16.5 8.230 14.200 8.230 14.200 multiply_cannon_metrocomm3 54864 15.5 0.068 0.072 5.863 12.677 multiply_cannon_multrec 54864 15.5 4.234 6.433 7.463 10.741 rebuild_ks_matrix 119 8.3 0.000 0.000 7.871 8.011 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 7.870 8.011 multiply_cannon_sync_h2d 54864 15.5 5.912 7.351 5.912 7.351 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.933 7.063 mp_sum_l 7207 12.9 5.332 6.901 5.332 6.901 qs_ot_get_p 119 10.4 0.001 0.001 6.519 6.852 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.381 5.812 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.358 5.450 init_scf_run 11 5.9 0.000 0.001 4.672 4.672 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.672 4.672 sum_up_and_integrate 119 10.3 0.012 0.015 4.504 4.511 integrate_v_rspace 119 11.3 0.002 0.002 4.492 4.500 dbcsr_mm_accdrv_process 76910 16.1 1.153 1.836 3.151 4.428 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.200 4.338 calculate_rho_elec 119 8.7 0.011 0.017 4.200 4.337 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.694 3.774 multiply_cannon_metrocomm1 54864 15.5 0.052 0.056 1.732 3.245 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.890 3.076 apply_single 119 13.6 0.000 0.000 2.890 3.076 calculate_dm_sparse 119 9.5 0.000 0.001 2.893 3.041 cp_dbcsr_syevd 50 12.0 0.002 0.003 2.869 2.869 rs_pw_transfer 974 11.9 0.011 0.013 2.678 2.794 ot_diis_step 108 11.5 0.006 0.006 2.708 2.709 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.664 2.664 cp_fm_redistribute_end 50 14.0 2.425 2.642 2.430 2.643 cp_fm_diag_elpa_base 50 14.0 0.212 2.555 0.213 2.566 jit_kernel_multiply 13 15.8 1.936 2.514 1.936 2.514 calculate_first_density_matrix 1 7.0 0.008 0.034 2.491 2.500 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.414 2.496 density_rs2pw 119 9.7 0.004 0.005 2.208 2.316 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.229 2.231 acc_transpose_blocks 54864 15.5 0.230 0.252 1.717 2.160 wfi_extrapolate 11 7.9 0.001 0.001 2.115 2.115 grid_integrate_task_list 119 12.3 2.039 2.114 2.039 2.114 init_scf_loop 11 6.9 0.000 0.000 2.020 2.020 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.929 1.986 mp_sum_d 4125 12.0 1.342 1.968 1.342 1.968 pw_transfer 1439 11.6 0.053 0.058 1.782 1.851 potential_pw2rs 119 12.3 0.004 0.004 1.833 1.846 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.705 1.777 make_m2s 4572 13.5 0.054 0.055 1.573 1.623 mp_alltoall_d11v 2130 13.8 1.277 1.544 1.277 1.544 fft3d_ps 1201 14.6 0.370 0.473 1.475 1.543 make_images 4572 14.5 0.132 0.138 1.491 1.541 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.405 1.435 fft_wrap_pw1pw2_140 487 13.2 0.083 0.096 1.316 1.392 mp_waitany 12084 13.8 1.227 1.386 1.227 1.386 grid_collocate_task_list 119 9.7 1.292 1.367 1.292 1.367 dbcsr_dot_sd 1205 11.9 0.051 0.062 0.731 1.119 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=51.447000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=430.818182, yerr=0.385695 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/530234f606c3418e1a725f5149ed409c2f956214_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 488.972288E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2194560 MPI messages size (bytes): total size 310.646604E+09 min size 0.000000E+00 max size 1.145520E+06 average size 141.553031E+03 MPI breakdown and total messages size (bytes): size <= 128 724648 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 281952 4619501568 32768 < size <= 131072 494448 39143342080 131072 < size <= 4194304 440000 264807943488 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62664. MP_Allreduce 10226 305. MP_Sync 104 MP_Alltoall 2060 1063903. 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 0.040 0.078 38.751 38.751 qs_mol_dyn_low 1 2.0 0.003 0.007 38.336 38.346 qs_forces 11 3.9 0.012 0.014 38.227 38.227 qs_energies 11 4.9 0.002 0.005 36.516 36.528 scf_env_do_scf 11 5.9 0.001 0.001 31.406 31.406 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 28.974 28.975 dbcsr_multiply_generic 2286 12.5 0.099 0.101 21.136 21.495 qs_scf_new_mos 108 7.5 0.001 0.001 19.717 19.961 qs_scf_loop_do_ot 108 8.5 0.001 0.001 19.717 19.961 ot_scf_mini 108 9.5 0.002 0.003 18.827 19.001 velocity_verlet 10 3.0 0.001 0.001 18.204 18.206 multiply_cannon 2286 13.5 0.209 0.220 16.231 17.817 multiply_cannon_loop 2286 14.5 0.903 0.979 15.123 16.470 ot_mini 108 10.5 0.001 0.001 11.627 11.863 mp_waitall_1 200699 16.5 5.577 10.752 5.577 10.752 multiply_cannon_metrocomm3 27432 15.5 0.068 0.070 4.100 9.457 qs_ot_get_derivative 108 11.5 0.001 0.002 9.174 9.351 multiply_cannon_multrec 27432 15.5 1.978 4.554 5.857 8.628 rebuild_ks_matrix 119 8.3 0.000 0.000 7.266 7.400 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.017 7.266 7.399 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.412 6.533 dbcsr_mm_accdrv_process 47894 16.0 3.009 5.073 3.809 5.633 qs_ot_get_p 119 10.4 0.001 0.001 4.325 4.556 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.541 4.389 sum_up_and_integrate 119 10.3 0.024 0.026 4.228 4.236 integrate_v_rspace 119 11.3 0.002 0.002 4.205 4.213 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.043 4.114 apply_single 119 13.6 0.000 0.000 3.043 4.114 mp_sum_l 7207 12.9 2.013 3.972 2.013 3.972 init_scf_run 11 5.9 0.000 0.001 3.898 3.898 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.898 3.898 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.814 3.849 calculate_rho_elec 119 8.7 0.021 0.024 3.813 3.848 rs_pw_transfer 974 11.9 0.010 0.011 2.569 3.025 qs_ot_p2m_diag 50 11.0 0.009 0.013 2.915 2.933 multiply_cannon_sync_h2d 27432 15.5 2.194 2.886 2.194 2.886 make_m2s 4572 13.5 0.052 0.054 2.446 2.672 density_rs2pw 119 9.7 0.004 0.004 2.140 2.607 make_images 4572 14.5 0.200 0.238 2.357 2.580 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.485 2.486 init_scf_loop 11 6.9 0.001 0.003 2.411 2.412 ot_diis_step 108 11.5 0.011 0.011 2.404 2.404 calculate_first_density_matrix 1 7.0 0.003 0.015 2.385 2.386 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.103 2.194 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.160 2.161 cp_fm_redistribute_end 50 14.0 1.789 2.137 1.793 2.139 calculate_dm_sparse 119 9.5 0.000 0.000 2.028 2.105 cp_fm_diag_elpa_base 50 14.0 0.331 2.031 0.344 2.084 jit_kernel_multiply 11 16.1 0.748 2.079 0.748 2.079 pw_transfer 1439 11.6 0.065 0.072 2.023 2.058 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.931 1.969 potential_pw2rs 119 12.3 0.006 0.006 1.952 1.962 grid_integrate_task_list 119 12.3 1.837 1.947 1.837 1.947 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 1.905 1.910 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.696 1.737 fft3d_ps 1201 14.6 0.513 0.567 1.630 1.661 fft_wrap_pw1pw2_140 487 13.2 0.079 0.085 1.518 1.559 make_images_data 4572 15.5 0.045 0.051 1.129 1.529 prepare_preconditioner 11 7.9 0.000 0.000 1.496 1.522 make_preconditioner 11 8.9 0.000 0.001 1.496 1.522 acc_transpose_blocks 27432 15.5 0.110 0.115 1.200 1.509 wfi_extrapolate 11 7.9 0.001 0.001 1.462 1.462 hybrid_alltoall_any 4725 16.4 0.051 0.112 0.980 1.454 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.395 1.451 grid_collocate_task_list 119 9.7 1.227 1.369 1.227 1.369 mp_allgather_i34 2286 14.5 0.549 1.304 0.549 1.304 mp_alltoall_d11v 2130 13.8 1.165 1.297 1.165 1.297 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.246 1.253 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.162 1.209 mp_sum_d 4125 12.0 0.581 1.041 0.581 1.041 mp_waitany 5720 13.7 0.518 1.011 0.518 1.011 rs_pw_transfer_RS2PW_140 130 11.5 0.139 0.149 0.543 0.995 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 0.947 0.957 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.913 0.926 acc_transpose_blocks_kernels 27432 16.5 0.182 0.270 0.668 0.887 mp_alltoall_z22v 1201 16.6 0.719 0.823 0.719 0.823 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=38.751000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=465.090909, yerr=1.928473 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/530234f606c3418e1a725f5149ed409c2f956214_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 527.130624E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62660. MP_Allreduce 10225 303. MP_Sync 104 MP_Alltoall 1821 1603749. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_comm_split 50 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.036 32.143 32.144 qs_mol_dyn_low 1 2.0 0.003 0.004 31.897 31.906 qs_forces 11 3.9 0.002 0.002 31.835 31.836 qs_energies 11 4.9 0.001 0.003 30.257 30.260 scf_env_do_scf 11 5.9 0.001 0.002 25.573 25.573 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 23.015 23.015 dbcsr_multiply_generic 2286 12.5 0.095 0.098 16.192 16.295 velocity_verlet 10 3.0 0.001 0.001 15.326 15.327 qs_scf_new_mos 108 7.5 0.001 0.001 14.692 14.705 qs_scf_loop_do_ot 108 8.5 0.001 0.001 14.691 14.704 ot_scf_mini 108 9.5 0.002 0.003 13.973 13.981 multiply_cannon 2286 13.5 0.198 0.206 12.993 13.758 multiply_cannon_loop 2286 14.5 0.638 0.666 12.208 13.001 ot_mini 108 10.5 0.001 0.001 8.625 8.639 qs_ot_get_derivative 108 11.5 0.001 0.001 7.139 7.148 multiply_cannon_multrec 18288 15.5 1.920 3.003 6.752 7.128 rebuild_ks_matrix 119 8.3 0.000 0.000 6.436 6.450 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 6.436 6.449 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.689 5.702 dbcsr_mm_accdrv_process 38222 16.0 4.076 5.487 4.747 5.584 sum_up_and_integrate 119 10.3 0.031 0.032 3.997 4.002 integrate_v_rspace 119 11.3 0.002 0.003 3.965 3.975 mp_waitall_1 158411 16.6 2.531 3.651 2.531 3.651 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.537 3.546 calculate_rho_elec 119 8.7 0.030 0.031 3.536 3.545 init_scf_run 11 5.9 0.000 0.001 3.502 3.502 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.502 3.502 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.672 3.277 qs_ot_get_p 119 10.4 0.001 0.001 3.199 3.216 init_scf_loop 11 6.9 0.000 0.002 2.540 2.540 rs_pw_transfer 974 11.9 0.009 0.010 2.287 2.539 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.011 2.336 apply_single 119 13.6 0.000 0.000 2.010 2.335 multiply_cannon_metrocomm3 18288 15.5 0.044 0.045 1.361 2.333 calculate_first_density_matrix 1 7.0 0.000 0.000 2.280 2.281 density_rs2pw 119 9.7 0.004 0.004 2.051 2.279 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.122 2.127 pw_transfer 1439 11.6 0.066 0.072 2.029 2.037 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.935 1.946 grid_integrate_task_list 119 12.3 1.798 1.910 1.798 1.910 make_m2s 4572 13.5 0.044 0.045 1.751 1.898 jit_kernel_multiply 10 16.3 0.620 1.888 0.620 1.888 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.858 1.858 calculate_dm_sparse 119 9.5 0.000 0.000 1.802 1.814 make_images 4572 14.5 0.190 0.201 1.666 1.811 potential_pw2rs 119 12.3 0.007 0.008 1.761 1.766 prepare_preconditioner 11 7.9 0.000 0.000 1.737 1.739 make_preconditioner 11 8.9 0.000 0.001 1.737 1.739 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.696 1.697 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.588 1.674 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.628 1.628 fft3d_ps 1201 14.6 0.524 0.541 1.615 1.628 cp_fm_redistribute_end 50 14.0 1.215 1.608 1.216 1.609 multiply_cannon_sync_h2d 18288 15.5 1.393 1.606 1.393 1.606 cp_fm_diag_elpa_base 50 14.0 0.377 1.525 0.391 1.573 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.562 1.568 fft_wrap_pw1pw2_140 487 13.2 0.090 0.095 1.530 1.541 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.533 1.540 mp_sum_l 7207 12.9 1.070 1.475 1.070 1.475 ot_diis_step 108 11.5 0.011 0.011 1.466 1.466 grid_collocate_task_list 119 9.7 1.216 1.356 1.216 1.356 acc_transpose_blocks 18288 15.5 0.076 0.078 1.252 1.276 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.175 1.180 wfi_extrapolate 11 7.9 0.001 0.001 1.173 1.173 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 0.958 0.960 make_images_data 4572 15.5 0.045 0.048 0.766 0.925 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.868 0.889 hybrid_alltoall_any 4725 16.4 0.055 0.114 0.651 0.848 acc_transpose_blocks_kernels 18288 16.5 0.210 0.219 0.810 0.822 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.819 0.820 mp_waitany 9880 13.7 0.523 0.801 0.523 0.801 mp_alltoall_z22v 1201 16.6 0.690 0.787 0.690 0.787 rs_pw_transfer_RS2PW_140 130 11.5 0.119 0.122 0.513 0.766 mp_alltoall_d11v 2130 13.8 0.660 0.759 0.660 0.759 cp_fm_cholesky_invert 11 10.9 0.723 0.727 0.723 0.727 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.660 0.725 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=32.144000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=498.727273, yerr=4.158035 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/530234f606c3418e1a725f5149ed409c2f956214_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 547.098624E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62659. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_comm_split 50 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.035 35.291 35.291 qs_mol_dyn_low 1 2.0 0.006 0.016 34.988 34.997 qs_forces 11 3.9 0.012 0.038 34.923 34.927 qs_energies 11 4.9 0.001 0.004 33.183 33.197 scf_env_do_scf 11 5.9 0.001 0.003 28.175 28.176 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 24.736 24.737 dbcsr_multiply_generic 2286 12.5 0.098 0.100 17.963 18.077 velocity_verlet 10 3.0 0.001 0.001 17.912 17.914 qs_scf_new_mos 108 7.5 0.001 0.001 16.117 16.171 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.116 16.170 ot_scf_mini 108 9.5 0.003 0.004 15.184 15.239 multiply_cannon 2286 13.5 0.223 0.253 14.377 14.819 multiply_cannon_loop 2286 14.5 0.939 0.970 13.418 13.754 ot_mini 108 10.5 0.001 0.001 9.172 9.235 multiply_cannon_multrec 27432 15.5 2.334 3.027 8.594 8.953 qs_ot_get_derivative 108 11.5 0.001 0.001 7.351 7.405 dbcsr_mm_accdrv_process 47916 15.9 5.257 6.597 6.165 7.366 rebuild_ks_matrix 119 8.3 0.000 0.000 6.680 6.743 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 6.680 6.743 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.933 5.987 sum_up_and_integrate 119 10.3 0.036 0.039 3.879 3.889 integrate_v_rspace 119 11.3 0.002 0.002 3.843 3.854 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.633 3.665 calculate_rho_elec 119 8.7 0.040 0.046 3.632 3.664 init_scf_run 11 5.9 0.000 0.001 3.611 3.612 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.611 3.612 qs_ot_get_p 119 10.4 0.001 0.001 3.395 3.466 init_scf_loop 11 6.9 0.001 0.003 3.416 3.416 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.758 3.156 prepare_preconditioner 11 7.9 0.000 0.000 2.549 2.558 make_preconditioner 11 8.9 0.002 0.006 2.549 2.558 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.157 2.483 mp_waitall_1 137007 16.6 1.780 2.357 1.780 2.357 make_m2s 4572 13.5 0.054 0.055 2.199 2.310 rs_pw_transfer 974 11.9 0.009 0.009 2.072 2.260 calculate_first_density_matrix 1 7.0 0.000 0.002 2.232 2.234 density_rs2pw 119 9.7 0.004 0.004 2.045 2.232 make_images 4572 14.5 0.271 0.331 2.092 2.202 pw_transfer 1439 11.6 0.066 0.071 2.150 2.188 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.046 2.176 apply_single 119 13.6 0.000 0.000 2.046 2.176 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.108 2.117 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.057 2.100 calculate_dm_sparse 119 9.5 0.000 0.000 2.024 2.085 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.923 1.953 grid_integrate_task_list 119 12.3 1.820 1.911 1.820 1.911 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.801 1.802 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.785 1.785 ot_diis_step 108 11.5 0.012 0.012 1.781 1.781 fft3d_ps 1201 14.6 0.559 0.608 1.732 1.772 fft_wrap_pw1pw2_140 487 13.2 0.088 0.095 1.683 1.730 jit_kernel_multiply 9 15.9 0.849 1.680 0.849 1.680 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.630 1.647 potential_pw2rs 119 12.3 0.008 0.009 1.632 1.637 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.524 1.525 cp_fm_redistribute_end 50 14.0 1.008 1.496 1.009 1.497 acc_transpose_blocks 27432 15.5 0.113 0.116 1.454 1.472 cp_fm_diag_elpa_base 50 14.0 0.465 1.426 0.485 1.465 grid_collocate_task_list 119 9.7 1.222 1.360 1.222 1.360 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.735 1.338 wfi_extrapolate 11 7.9 0.001 0.001 1.328 1.328 mp_sum_l 7207 12.9 0.966 1.249 0.966 1.249 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.001 1.213 1.225 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.131 1.145 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.100 1.122 cp_fm_upper_to_full 72 13.5 0.792 1.100 0.792 1.100 multiply_cannon_sync_h2d 27432 15.5 0.996 1.072 0.996 1.072 dbcsr_complete_redistribute 329 12.2 0.117 0.138 0.790 1.064 make_images_data 4572 15.5 0.045 0.049 0.859 0.976 hybrid_alltoall_any 4725 16.4 0.062 0.153 0.748 0.920 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.793 0.872 mp_alltoall_d11v 2130 13.8 0.743 0.846 0.743 0.846 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.838 0.844 acc_transpose_blocks_kernels 27432 16.5 0.267 0.275 0.829 0.839 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.564 0.825 mp_alltoall_z22v 1201 16.6 0.756 0.794 0.756 0.794 cp_fm_cholesky_invert 11 10.9 0.754 0.757 0.754 0.757 mp_alltoall_i22 627 13.8 0.430 0.719 0.430 0.719 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=35.291000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=520.363636, yerr=2.346089 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/530234f606c3418e1a725f5149ed409c2f956214_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 604.237824E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62658. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_comm_split 50 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.032 28.552 28.553 qs_mol_dyn_low 1 2.0 0.003 0.004 28.169 28.176 qs_forces 11 3.9 0.002 0.002 28.108 28.108 qs_energies 11 4.9 0.001 0.001 26.392 26.394 scf_env_do_scf 11 5.9 0.000 0.001 21.471 21.472 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 18.976 18.977 velocity_verlet 10 3.0 0.001 0.001 14.400 14.402 dbcsr_multiply_generic 2286 12.5 0.091 0.092 12.305 12.375 qs_scf_new_mos 108 7.5 0.001 0.001 10.893 10.921 qs_scf_loop_do_ot 108 8.5 0.001 0.001 10.892 10.920 multiply_cannon 2286 13.5 0.229 0.237 9.828 10.269 ot_scf_mini 108 9.5 0.002 0.002 10.225 10.253 multiply_cannon_loop 2286 14.5 0.331 0.343 8.927 9.108 multiply_cannon_multrec 9144 15.5 1.695 1.960 6.070 6.299 rebuild_ks_matrix 119 8.3 0.000 0.000 6.054 6.078 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 6.053 6.077 ot_mini 108 10.5 0.001 0.001 5.706 5.741 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.396 5.418 qs_ot_get_derivative 108 11.5 0.001 0.001 4.408 4.437 dbcsr_mm_accdrv_process 12550 15.8 3.230 4.095 4.273 4.355 sum_up_and_integrate 119 10.3 0.038 0.042 3.701 3.710 integrate_v_rspace 119 11.3 0.003 0.003 3.663 3.672 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.565 3.575 calculate_rho_elec 119 8.7 0.059 0.061 3.565 3.575 init_scf_run 11 5.9 0.000 0.001 3.456 3.456 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.456 3.456 qs_ot_get_p 119 10.4 0.001 0.001 2.738 2.790 init_scf_loop 11 6.9 0.000 0.000 2.475 2.476 calculate_first_density_matrix 1 7.0 0.000 0.000 2.306 2.309 pw_transfer 1439 11.6 0.066 0.069 2.183 2.192 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.089 2.100 density_rs2pw 119 9.7 0.004 0.004 1.897 2.034 jit_kernel_multiply 10 15.8 1.005 1.970 1.005 1.970 grid_integrate_task_list 119 12.3 1.854 1.928 1.854 1.928 make_m2s 4572 13.5 0.034 0.035 1.717 1.878 mp_waitall_1 115863 16.7 1.365 1.866 1.365 1.866 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.844 1.846 calculate_dm_sparse 119 9.5 0.000 0.000 1.781 1.801 make_images 4572 14.5 0.268 0.301 1.628 1.787 rs_pw_transfer 974 11.9 0.008 0.008 1.658 1.785 fft3d_ps 1201 14.6 0.563 0.575 1.749 1.759 fft_wrap_pw1pw2_140 487 13.2 0.087 0.090 1.730 1.741 prepare_preconditioner 11 7.9 0.000 0.000 1.722 1.726 make_preconditioner 11 8.9 0.000 0.000 1.722 1.726 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.678 1.680 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.651 1.651 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.607 1.633 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.562 1.581 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.495 1.502 potential_pw2rs 119 12.3 0.010 0.010 1.428 1.430 grid_collocate_task_list 119 9.7 1.271 1.388 1.271 1.388 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.375 1.375 cp_fm_redistribute_end 50 14.0 0.686 1.352 0.687 1.353 cp_fm_diag_elpa_base 50 14.0 0.621 1.281 0.665 1.338 ot_diis_step 108 11.5 0.012 0.013 1.286 1.286 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.257 1.257 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.206 1.212 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.163 1.183 apply_single 119 13.6 0.000 0.000 1.163 1.183 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.121 1.135 wfi_extrapolate 11 7.9 0.001 0.001 1.099 1.099 hybrid_alltoall_any 4725 16.4 0.063 0.176 0.783 1.027 make_images_data 4572 15.5 0.039 0.042 0.811 0.999 acc_transpose_blocks 9144 15.5 0.039 0.040 0.974 0.981 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.868 0.915 cp_fm_cholesky_invert 11 10.9 0.867 0.869 0.867 0.869 mp_alltoall_d11v 2130 13.8 0.759 0.858 0.759 0.858 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.780 0.783 multiply_cannon_sync_h2d 9144 15.5 0.715 0.782 0.715 0.782 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.693 0.740 acc_transpose_blocks_kernels 9144 16.5 0.117 0.120 0.734 0.736 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.715 0.723 mp_alltoall_z22v 1201 16.6 0.632 0.685 0.632 0.685 mp_allgather_i34 2286 14.5 0.237 0.659 0.237 0.659 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.314 0.647 jit_kernel_transpose 5 15.6 0.617 0.620 0.617 0.620 yz_to_x 606 15.1 0.265 0.275 0.591 0.608 x_to_yz 595 16.2 0.276 0.287 0.582 0.587 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=28.553000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=568.181818, yerr=6.766471 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/530234f606c3418e1a725f5149ed409c2f956214_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 750.608384E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63729. MP_Allreduce 10074 433. MP_Sync 54 MP_Alltoall 1582 7383731. MP_SendRecv 2499 189067. MP_ISendRecv 2499 189067. MP_Wait 6399 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.027 0.049 41.657 41.658 qs_mol_dyn_low 1 2.0 0.004 0.009 41.225 41.234 qs_forces 11 3.9 0.014 0.015 41.118 41.119 qs_energies 11 4.9 0.001 0.001 39.105 39.108 scf_env_do_scf 11 5.9 0.001 0.001 33.275 33.275 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 25.509 25.511 velocity_verlet 10 3.0 0.001 0.001 23.337 23.349 dbcsr_multiply_generic 2286 12.5 0.099 0.101 17.399 17.584 qs_scf_new_mos 108 7.5 0.001 0.001 15.640 15.735 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.639 15.734 ot_scf_mini 108 9.5 0.002 0.002 14.559 14.658 multiply_cannon 2286 13.5 0.301 0.304 13.641 14.539 multiply_cannon_loop 2286 14.5 0.343 0.349 12.400 13.343 ot_mini 108 10.5 0.001 0.001 8.680 8.795 multiply_cannon_multrec 9144 15.5 3.361 4.721 8.618 8.704 init_scf_loop 11 6.9 0.000 0.000 7.739 7.743 rebuild_ks_matrix 119 8.3 0.000 0.000 7.358 7.502 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.358 7.502 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.654 6.784 prepare_preconditioner 11 7.9 0.000 0.000 6.766 6.779 make_preconditioner 11 8.9 0.000 0.000 6.765 6.779 qs_ot_get_derivative 108 11.5 0.001 0.001 6.673 6.774 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.324 6.660 dbcsr_mm_accdrv_process 12550 15.8 4.123 5.609 5.130 6.448 cp_fm_upper_to_full 72 14.2 3.152 4.537 3.152 4.537 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.348 4.409 calculate_rho_elec 119 8.7 0.118 0.121 4.348 4.409 sum_up_and_integrate 119 10.3 0.065 0.067 4.041 4.047 integrate_v_rspace 119 11.3 0.003 0.003 3.976 3.983 init_scf_run 11 5.9 0.000 0.001 3.787 3.787 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.786 3.786 qs_ot_get_p 119 10.4 0.001 0.001 3.169 3.306 mp_waitall_1 94719 16.7 2.285 3.277 2.285 3.277 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.446 2.855 pw_transfer 1439 11.6 0.069 0.069 2.840 2.845 dbcsr_complete_redistribute 329 12.2 0.286 0.293 1.969 2.778 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.741 2.746 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.664 2.472 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.157 2.438 apply_single 119 13.6 0.000 0.000 2.157 2.438 make_m2s 4572 13.5 0.038 0.038 2.213 2.370 fft3d_ps 1201 14.6 0.596 0.611 2.365 2.368 calculate_first_density_matrix 1 7.0 0.000 0.000 2.283 2.345 calculate_dm_sparse 119 9.5 0.000 0.000 2.281 2.339 fft_wrap_pw1pw2_140 487 13.2 0.095 0.096 2.309 2.318 mp_alltoall_i22 627 13.8 1.413 2.263 1.413 2.263 make_images 4572 14.5 0.354 0.386 2.092 2.250 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.437 2.237 density_rs2pw 119 9.7 0.004 0.004 2.205 2.226 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 1.302 2.212 grid_integrate_task_list 119 12.3 2.059 2.083 2.059 2.083 ot_diis_step 108 11.5 0.014 0.014 1.982 1.982 qs_ot_p2m_diag 50 11.0 0.043 0.044 1.927 1.927 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.861 1.862 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.786 1.787 mp_sum_l 7207 12.9 1.007 1.745 1.007 1.745 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.663 1.714 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.614 1.659 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.655 1.655 grid_collocate_task_list 119 9.7 1.491 1.508 1.491 1.508 jit_kernel_multiply 8 15.5 0.980 1.493 0.980 1.493 rs_pw_transfer 974 11.9 0.009 0.009 1.431 1.465 potential_pw2rs 119 12.3 0.014 0.015 1.436 1.439 cp_fm_cholesky_invert 11 10.9 1.414 1.417 1.414 1.417 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.346 1.404 wfi_extrapolate 11 7.9 0.001 0.001 1.379 1.379 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.369 1.369 cp_fm_diag_elpa_base 50 14.0 1.227 1.278 1.367 1.368 hybrid_alltoall_any 4725 16.4 0.087 0.148 1.071 1.268 make_images_data 4572 15.5 0.043 0.046 1.015 1.220 mp_alltoall_d11v 2130 13.8 1.123 1.191 1.123 1.191 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.132 1.152 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.113 1.129 multiply_cannon_sync_h2d 9144 15.5 1.043 1.048 1.043 1.048 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.989 1.044 yz_to_x 606 15.1 0.460 0.473 0.945 0.955 qs_create_task_list 11 7.9 0.000 0.001 0.937 0.951 generate_qs_task_list 11 8.9 0.371 0.391 0.936 0.950 acc_transpose_blocks 9144 15.5 0.038 0.038 0.929 0.933 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.917 0.930 mp_alltoall_z22v 1201 16.6 0.817 0.861 0.817 0.861 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=41.658000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=704.363636, yerr=13.452088 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/530234f606c3418e1a725f5149ed409c2f956214_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 502.243328E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65372. MP_Allreduce 9840 486. MP_Sync 100 MP_Alltoall 1938 1379060. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_comm_split 48 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.030 82.930 82.931 qs_mol_dyn_low 1 2.0 0.003 0.004 82.685 82.697 qs_forces 11 3.9 0.002 0.002 82.612 82.614 qs_energies 11 4.9 0.002 0.010 79.767 79.776 scf_env_do_scf 11 5.9 0.000 0.001 70.933 70.935 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 65.445 65.445 dbcsr_multiply_generic 2055 12.4 0.106 0.109 51.828 52.120 qs_scf_new_mos 99 7.5 0.000 0.001 47.982 48.118 qs_scf_loop_do_ot 99 8.5 0.000 0.001 47.982 48.117 ot_scf_mini 99 9.5 0.002 0.002 45.547 45.636 multiply_cannon 2055 13.4 0.179 0.186 42.395 43.261 multiply_cannon_loop 2055 14.4 1.544 1.587 41.437 42.323 velocity_verlet 10 3.0 0.001 0.001 41.858 41.859 ot_mini 99 10.5 0.001 0.001 27.456 27.546 qs_ot_get_derivative 99 11.5 0.001 0.001 20.648 20.748 multiply_cannon_multrec 49320 15.4 12.091 12.913 17.120 17.725 rebuild_ks_matrix 110 8.3 0.000 0.001 14.589 14.729 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.021 14.589 14.728 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.802 12.928 mp_waitall_1 220248 16.4 11.399 12.535 11.399 12.535 multiply_cannon_sync_h2d 49320 15.4 9.954 10.540 9.954 10.540 qs_ot_get_p 110 10.4 0.001 0.001 9.636 9.755 multiply_cannon_metrocomm3 49320 15.4 0.080 0.086 6.911 8.219 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.114 7.621 apply_single 110 13.6 0.000 0.001 7.114 7.621 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 6.986 7.519 sum_up_and_integrate 110 10.3 0.037 0.044 7.116 7.132 integrate_v_rspace 110 11.3 0.002 0.003 7.079 7.104 init_scf_run 11 5.9 0.000 0.001 6.811 6.812 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.811 6.811 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.665 6.810 calculate_rho_elec 110 8.6 0.021 0.025 6.665 6.809 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.655 6.711 ot_diis_step 99 11.5 0.005 0.006 6.509 6.509 qs_ot_p2m_diag 48 11.0 0.012 0.018 6.379 6.406 mp_sum_l 6514 12.8 5.157 6.067 5.157 6.067 init_scf_loop 11 6.9 0.000 0.000 5.458 5.458 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.337 5.337 dbcsr_mm_accdrv_process 87628 16.1 2.074 2.173 4.908 5.196 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.879 4.881 cp_fm_redistribute_end 48 14.0 4.261 4.857 4.265 4.858 cp_fm_diag_elpa_base 48 14.0 0.587 4.715 0.591 4.739 rs_pw_transfer 902 11.9 0.011 0.013 3.653 4.224 wfi_extrapolate 11 7.9 0.001 0.001 4.041 4.041 density_rs2pw 110 9.6 0.004 0.005 3.450 3.976 make_m2s 4110 13.4 0.060 0.065 3.817 3.954 calculate_dm_sparse 110 9.5 0.000 0.001 3.796 3.886 make_images 4110 14.4 0.177 0.190 3.722 3.863 multiply_cannon_metrocomm1 49320 15.4 0.060 0.065 2.661 3.682 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.556 3.560 grid_integrate_task_list 110 12.3 3.268 3.427 3.268 3.427 pw_transfer 1331 11.6 0.055 0.070 3.277 3.341 prepare_preconditioner 11 7.9 0.000 0.000 3.293 3.313 make_preconditioner 11 8.9 0.000 0.000 3.293 3.313 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.226 3.280 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.211 3.263 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 3.189 3.256 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.078 3.119 fft_wrap_pw1pw2_140 451 13.1 0.172 0.193 2.716 2.786 fft3d_ps 1111 14.6 0.783 0.889 2.715 2.771 calculate_first_density_matrix 1 7.0 0.000 0.000 2.681 2.685 potential_pw2rs 110 12.3 0.006 0.007 2.639 2.658 jit_kernel_multiply 13 15.9 2.548 2.588 2.548 2.588 mp_alltoall_d11v 2046 13.8 2.056 2.484 2.056 2.484 mp_waitany 14300 13.8 1.781 2.419 1.781 2.419 grid_collocate_task_list 110 9.6 2.085 2.266 2.085 2.266 acc_transpose_blocks 49320 15.4 0.226 0.237 2.175 2.245 mp_sum_d 3879 11.9 1.493 2.020 1.493 2.020 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.927 1.943 make_images_data 4110 15.4 0.043 0.046 1.747 1.888 hybrid_alltoall_any 4261 16.3 0.082 0.481 1.517 1.828 cp_fm_cholesky_invert 11 10.9 1.794 1.798 1.794 1.798 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=82.931000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.909091, yerr=2.466302 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/530234f606c3418e1a725f5149ed409c2f956214_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 593.100800E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1972800 MPI messages size (bytes): total size 1.077520E+12 min size 0.000000E+00 max size 4.537280E+06 average size 546.188250E+03 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 222984 1826684928 8192 < size <= 32768 520356 13399818240 32768 < size <= 131072 372336 35386294272 131072 < size <= 4194304 787758 788321309808 4194304 < size <= 16777216 54450 238588003280 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65587. MP_Allreduce 9839 562. MP_Sync 100 MP_Alltoall 1717 1802496. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 MP_comm_split 48 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.030 69.647 69.648 qs_mol_dyn_low 1 2.0 0.003 0.005 69.421 69.431 qs_forces 11 3.9 0.002 0.003 69.351 69.352 qs_energies 11 4.9 0.001 0.002 65.992 65.996 scf_env_do_scf 11 5.9 0.000 0.001 57.354 57.357 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 49.568 49.568 dbcsr_multiply_generic 2055 12.4 0.113 0.118 37.763 37.920 velocity_verlet 10 3.0 0.001 0.002 36.701 36.702 qs_scf_new_mos 99 7.5 0.001 0.001 32.914 33.034 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.913 33.033 multiply_cannon 2055 13.4 0.223 0.245 31.087 32.356 ot_scf_mini 99 9.5 0.003 0.003 31.261 31.377 multiply_cannon_loop 2055 14.4 0.928 0.967 29.734 30.606 ot_mini 99 10.5 0.001 0.001 18.231 18.357 multiply_cannon_multrec 24660 15.4 7.633 9.669 13.811 15.619 rebuild_ks_matrix 110 8.3 0.000 0.001 14.034 14.116 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.015 14.034 14.116 qs_ot_get_derivative 99 11.5 0.001 0.001 12.423 12.541 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.373 12.448 mp_waitall_1 176588 16.5 7.663 10.359 7.663 10.359 multiply_cannon_sync_h2d 24660 15.4 7.089 8.148 7.089 8.148 multiply_cannon_metrocomm3 24660 15.4 0.070 0.072 5.156 7.768 init_scf_loop 11 6.9 0.000 0.000 7.749 7.750 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.516 7.262 apply_single 110 13.6 0.000 0.001 6.515 7.262 sum_up_and_integrate 110 10.3 0.052 0.059 6.748 6.758 integrate_v_rspace 110 11.3 0.002 0.002 6.696 6.710 qs_ot_get_p 110 10.4 0.001 0.001 6.350 6.498 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.309 6.318 calculate_rho_elec 110 8.6 0.040 0.047 6.309 6.317 dbcsr_mm_accdrv_process 52282 16.1 4.686 5.749 6.018 6.309 init_scf_run 11 5.9 0.000 0.001 6.265 6.266 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.265 6.266 ot_diis_step 99 11.5 0.010 0.010 5.761 5.761 prepare_preconditioner 11 7.9 0.000 0.000 5.667 5.692 make_preconditioner 11 8.9 0.000 0.000 5.667 5.692 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.228 5.386 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.670 5.386 make_m2s 4110 13.4 0.056 0.060 4.216 4.700 make_images 4110 14.4 0.401 0.445 4.108 4.589 qs_ot_p2m_diag 48 11.0 0.028 0.044 4.419 4.440 pw_transfer 1331 11.6 0.066 0.072 3.856 3.997 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.983 3.984 density_rs2pw 110 9.6 0.004 0.005 3.430 3.965 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.749 3.894 rs_pw_transfer 902 11.9 0.012 0.013 3.062 3.603 wfi_extrapolate 11 7.9 0.001 0.001 3.588 3.588 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.411 3.413 cp_fm_redistribute_end 48 14.0 2.549 3.382 2.551 3.382 grid_integrate_task_list 110 12.3 3.164 3.354 3.164 3.354 cp_fm_diag_elpa_base 48 14.0 0.796 3.239 0.827 3.320 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.311 3.313 fft_wrap_pw1pw2_140 451 13.1 0.202 0.219 3.131 3.280 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.211 3.272 fft3d_ps 1111 14.6 1.101 1.321 3.099 3.233 calculate_dm_sparse 110 9.5 0.001 0.001 2.954 2.987 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.954 2.985 hybrid_alltoall_any 4261 16.3 0.103 0.440 2.042 2.837 make_images_data 4110 15.4 0.047 0.051 2.347 2.829 cp_fm_cholesky_invert 11 10.9 2.643 2.650 2.643 2.650 calculate_first_density_matrix 1 7.0 0.000 0.000 2.588 2.591 potential_pw2rs 110 12.3 0.008 0.009 2.542 2.557 grid_collocate_task_list 110 9.6 2.057 2.482 2.057 2.482 jit_kernel_multiply 13 16.4 0.978 2.452 0.978 2.452 mp_sum_l 6514 12.8 1.840 2.354 1.840 2.354 mp_alltoall_d11v 2046 13.8 1.779 2.196 1.779 2.196 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.956 1.977 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.873 1.874 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.827 1.835 mp_waitany 10164 13.8 1.217 1.779 1.217 1.779 mp_allgather_i34 2055 14.4 0.682 1.718 0.682 1.718 acc_transpose_blocks 24660 15.4 0.111 0.114 1.545 1.577 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.531 1.544 rs_pw_transfer_RS2PW_140 121 11.5 0.206 0.214 0.959 1.497 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.368 1.476 dbcsr_complete_redistribute 325 12.2 0.239 0.293 1.181 1.449 multiply_cannon_metrocomm4 22605 15.4 0.072 0.076 0.764 1.445 mp_alltoall_z22v 1111 16.6 1.310 1.427 1.310 1.427 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=69.648000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=557.818182, yerr=8.726326 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/530234f606c3418e1a725f5149ed409c2f956214_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 659.058688E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65578. MP_Allreduce 9838 559. MP_Sync 100 MP_Alltoall 1496 4511006. MP_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 MP_comm_split 48 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.029 0.054 60.829 60.830 qs_mol_dyn_low 1 2.0 0.009 0.031 60.489 60.500 qs_forces 11 3.9 0.025 0.029 60.409 60.417 qs_energies 11 4.9 0.003 0.015 57.133 57.143 scf_env_do_scf 11 5.9 0.001 0.002 49.174 49.175 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 40.420 40.421 velocity_verlet 10 3.0 0.001 0.002 32.970 32.972 dbcsr_multiply_generic 2055 12.4 0.107 0.110 28.310 28.568 qs_scf_new_mos 99 7.5 0.001 0.001 25.143 25.235 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.142 25.235 ot_scf_mini 99 9.5 0.003 0.003 23.919 24.029 multiply_cannon 2055 13.4 0.212 0.226 22.023 23.168 multiply_cannon_loop 2055 14.4 0.616 0.634 20.816 21.917 ot_mini 99 10.5 0.001 0.001 13.567 13.679 rebuild_ks_matrix 110 8.3 0.000 0.001 12.612 12.743 qs_ks_build_kohn_sham_matrix 110 9.3 0.022 0.029 12.612 12.743 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.111 11.226 multiply_cannon_multrec 16440 15.4 3.952 5.094 9.711 10.753 mp_waitall_1 139946 16.5 6.881 10.055 6.881 10.055 qs_ot_get_derivative 99 11.5 0.001 0.001 9.104 9.216 init_scf_loop 11 6.9 0.001 0.006 8.715 8.717 multiply_cannon_metrocomm3 16440 15.4 0.044 0.045 4.211 7.262 prepare_preconditioner 11 7.9 0.000 0.000 6.835 6.858 make_preconditioner 11 8.9 0.000 0.002 6.835 6.858 sum_up_and_integrate 110 10.3 0.060 0.061 6.683 6.703 integrate_v_rspace 110 11.3 0.002 0.003 6.623 6.644 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.163 6.529 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.103 6.110 calculate_rho_elec 110 8.6 0.058 0.059 6.102 6.110 dbcsr_mm_accdrv_process 34862 16.1 4.591 5.341 5.612 5.778 qs_ot_get_p 110 10.4 0.001 0.001 5.560 5.682 init_scf_run 11 5.9 0.000 0.001 5.494 5.494 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.494 5.494 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.968 5.410 apply_single 110 13.6 0.000 0.000 4.968 5.410 make_m2s 4110 13.4 0.050 0.051 4.189 4.564 make_images 4110 14.4 0.394 0.513 4.074 4.452 ot_diis_step 99 11.5 0.010 0.011 4.434 4.434 density_rs2pw 110 9.6 0.004 0.005 3.225 4.356 multiply_cannon_sync_h2d 16440 15.4 3.687 4.278 3.687 4.278 rs_pw_transfer 902 11.9 0.010 0.011 2.841 3.952 qs_ot_p2m_diag 48 11.0 0.041 0.044 3.931 3.935 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.145 3.786 pw_transfer 1331 11.6 0.066 0.072 3.760 3.773 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.652 3.667 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.587 3.587 grid_integrate_task_list 110 12.3 3.156 3.429 3.156 3.429 fft_wrap_pw1pw2_140 451 13.1 0.212 0.215 3.121 3.133 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.042 3.042 cp_fm_redistribute_end 48 14.0 1.895 3.010 1.898 3.011 wfi_extrapolate 11 7.9 0.001 0.001 3.006 3.006 fft3d_ps 1111 14.6 1.089 1.112 2.953 2.969 cp_fm_diag_elpa_base 48 14.0 1.049 2.861 1.109 2.966 make_images_data 4110 15.4 0.043 0.048 2.399 2.854 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.769 2.771 hybrid_alltoall_any 4261 16.3 0.105 0.373 2.074 2.739 cp_fm_cholesky_invert 11 10.9 2.706 2.712 2.706 2.712 calculate_dm_sparse 110 9.5 0.001 0.001 2.511 2.544 grid_collocate_task_list 110 9.6 2.082 2.487 2.082 2.487 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.436 2.478 multiply_cannon_metrocomm4 14385 15.4 0.045 0.048 0.864 2.427 potential_pw2rs 110 12.3 0.010 0.010 2.405 2.414 mp_waitany 17072 13.8 1.236 2.403 1.236 2.403 calculate_first_density_matrix 1 7.0 0.001 0.002 2.397 2.398 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.329 2.388 mp_irecv_dv 48980 15.7 0.794 2.303 0.794 2.303 mp_alltoall_d11v 2046 13.8 1.768 2.175 1.768 2.175 mp_sum_l 6514 12.8 1.504 2.125 1.504 2.125 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 2.026 2.034 rs_pw_transfer_RS2PW_140 121 11.5 0.175 0.179 0.913 2.021 dbcsr_complete_redistribute 325 12.2 0.327 0.372 1.491 1.962 cp_fm_upper_to_full 70 13.6 1.407 1.867 1.407 1.867 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.703 1.718 cp_fm_cholesky_decompose 22 10.9 1.608 1.624 1.608 1.624 mp_allgather_i34 2055 14.4 0.526 1.557 0.526 1.557 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.365 1.503 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.469 1.481 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.981 1.439 jit_kernel_multiply 7 16.3 0.627 1.338 0.627 1.338 rs_gather_matrices 110 12.3 0.234 0.267 0.976 1.313 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.255 1.261 acc_transpose_blocks 16440 15.4 0.072 0.073 1.220 1.243 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=60.830000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=623.909091, yerr=8.284208 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/530234f606c3418e1a725f5149ed409c2f956214_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 737.120256E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65576. MP_Allreduce 9838 600. MP_Sync 100 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 MP_comm_split 48 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.032 65.917 65.918 qs_mol_dyn_low 1 2.0 0.006 0.017 65.559 65.570 qs_forces 11 3.9 0.003 0.006 65.376 65.380 qs_energies 11 4.9 0.001 0.003 61.930 61.937 scf_env_do_scf 11 5.9 0.001 0.001 53.562 53.565 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 41.794 41.794 velocity_verlet 10 3.0 0.001 0.001 37.405 37.412 dbcsr_multiply_generic 2055 12.4 0.135 0.138 29.406 29.654 qs_scf_new_mos 99 7.5 0.001 0.001 26.514 26.611 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.513 26.611 ot_scf_mini 99 9.5 0.003 0.004 24.894 24.990 multiply_cannon 2055 13.4 0.244 0.260 22.500 23.689 multiply_cannon_loop 2055 14.4 0.887 0.906 21.069 21.697 ot_mini 99 10.5 0.001 0.001 14.217 14.335 multiply_cannon_multrec 24660 15.4 4.212 6.753 12.607 13.815 rebuild_ks_matrix 110 8.3 0.000 0.000 12.367 12.469 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 12.366 12.468 init_scf_loop 11 6.9 0.000 0.001 11.723 11.724 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.928 11.016 qs_ot_get_derivative 99 11.5 0.001 0.001 10.068 10.167 prepare_preconditioner 11 7.9 0.000 0.000 9.938 9.953 make_preconditioner 11 8.9 0.000 0.001 9.938 9.953 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.179 9.615 dbcsr_mm_accdrv_process 52304 16.0 6.859 8.407 8.249 9.193 sum_up_and_integrate 110 10.3 0.068 0.071 6.629 6.641 integrate_v_rspace 110 11.3 0.003 0.003 6.561 6.572 mp_waitall_1 121746 16.5 4.316 6.278 4.316 6.278 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.241 6.252 calculate_rho_elec 110 8.6 0.077 0.081 6.241 6.252 qs_ot_get_p 110 10.4 0.001 0.001 5.705 5.860 make_m2s 4110 13.4 0.059 0.060 5.292 5.661 init_scf_run 11 5.9 0.000 0.001 5.562 5.562 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.561 5.562 make_images 4110 14.4 0.577 0.699 5.152 5.517 cp_fm_upper_to_full 70 13.8 3.257 4.632 3.257 4.632 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.030 4.125 apply_single 110 13.6 0.000 0.000 4.029 4.125 ot_diis_step 99 11.5 0.011 0.011 4.114 4.114 pw_transfer 1331 11.6 0.065 0.075 3.947 3.980 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.929 3.943 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.841 3.878 dbcsr_complete_redistribute 325 12.2 0.470 0.507 2.722 3.845 density_rs2pw 110 9.6 0.004 0.004 3.166 3.677 multiply_cannon_metrocomm3 24660 15.4 0.036 0.036 1.513 3.542 grid_integrate_task_list 110 12.3 3.263 3.469 3.263 3.469 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.455 3.456 multiply_cannon_sync_h2d 24660 15.4 3.186 3.361 3.186 3.361 fft_wrap_pw1pw2_140 451 13.1 0.202 0.211 3.294 3.334 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.277 3.326 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.179 3.285 fft3d_ps 1111 14.6 1.090 1.131 3.133 3.161 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.990 3.044 rs_pw_transfer 902 11.9 0.010 0.011 2.502 3.039 hybrid_alltoall_any 4261 16.3 0.120 0.459 2.306 3.012 wfi_extrapolate 11 7.9 0.001 0.001 3.003 3.003 make_images_data 4110 15.4 0.047 0.051 2.611 2.991 calculate_dm_sparse 110 9.5 0.001 0.001 2.908 2.940 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.897 2.899 cp_fm_redistribute_end 48 14.0 1.449 2.866 1.450 2.867 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.849 2.852 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.750 2.841 cp_fm_diag_elpa_base 48 14.0 1.335 2.726 1.414 2.837 mp_alltoall_i22 605 13.7 1.625 2.786 1.625 2.786 cp_fm_cholesky_invert 11 10.9 2.741 2.750 2.741 2.750 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.429 2.466 calculate_first_density_matrix 1 7.0 0.000 0.000 2.458 2.460 grid_collocate_task_list 110 9.6 2.183 2.420 2.183 2.420 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 2.309 2.314 potential_pw2rs 110 12.3 0.013 0.013 2.276 2.286 mp_alltoall_d11v 2046 13.8 1.789 2.066 1.789 2.066 jit_kernel_multiply 10 15.9 1.056 1.817 1.056 1.817 cp_fm_cholesky_decompose 22 10.9 1.711 1.759 1.711 1.759 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.698 1.732 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.600 1.705 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.685 1.697 mp_waitany 13376 13.8 1.102 1.595 1.102 1.595 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.526 1.542 acc_transpose_blocks 24660 15.4 0.107 0.109 1.499 1.517 multiply_cannon_metrocomm4 20550 15.4 0.057 0.060 0.851 1.469 mp_allgather_i34 2055 14.4 0.532 1.461 0.532 1.461 mp_sum_l 6514 12.8 0.911 1.394 0.911 1.394 mp_irecv_dv 62702 16.1 0.751 1.389 0.751 1.389 rs_pw_transfer_RS2PW_140 121 11.5 0.168 0.176 0.847 1.385 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.250 1.367 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=65.918000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=697.909091, yerr=10.492028 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/530234f606c3418e1a725f5149ed409c2f956214_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 829.239296E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65574. MP_Allreduce 9838 640. MP_Sync 100 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 MP_comm_split 48 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.040 55.981 55.985 qs_mol_dyn_low 1 2.0 0.003 0.003 55.736 55.747 qs_forces 11 3.9 0.002 0.002 55.664 55.665 qs_energies 11 4.9 0.001 0.001 51.947 51.955 scf_env_do_scf 11 5.9 0.000 0.001 43.728 43.728 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 35.878 35.878 velocity_verlet 10 3.0 0.001 0.001 31.516 31.519 dbcsr_multiply_generic 2055 12.4 0.105 0.107 22.744 22.859 qs_scf_new_mos 99 7.5 0.001 0.001 20.524 20.568 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.523 20.567 ot_scf_mini 99 9.5 0.002 0.002 19.281 19.297 multiply_cannon 2055 13.4 0.248 0.269 17.246 18.623 multiply_cannon_loop 2055 14.4 0.321 0.333 15.891 16.170 rebuild_ks_matrix 110 8.3 0.000 0.000 12.137 12.155 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 12.137 12.154 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.782 10.802 ot_mini 99 10.5 0.001 0.001 10.342 10.351 multiply_cannon_multrec 8220 15.4 3.204 4.462 7.416 8.523 init_scf_loop 11 6.9 0.000 0.000 7.802 7.803 mp_waitall_1 103326 16.6 6.034 7.544 6.034 7.544 sum_up_and_integrate 110 10.3 0.079 0.080 6.631 6.645 qs_ot_get_derivative 99 11.5 0.001 0.001 6.564 6.579 integrate_v_rspace 110 11.3 0.003 0.003 6.552 6.565 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.431 6.448 calculate_rho_elec 110 8.6 0.116 0.116 6.430 6.448 prepare_preconditioner 11 7.9 0.000 0.000 6.114 6.119 make_preconditioner 11 8.9 0.000 0.000 6.113 6.119 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.685 5.767 init_scf_run 11 5.9 0.000 0.001 5.079 5.079 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.079 5.079 qs_ot_get_p 110 10.4 0.001 0.001 5.006 5.037 dbcsr_mm_accdrv_process 17442 15.9 2.720 3.611 4.083 4.974 make_m2s 4110 13.4 0.038 0.039 4.214 4.478 multiply_cannon_metrocomm3 8220 15.4 0.017 0.018 3.036 4.448 make_images 4110 14.4 0.640 0.699 4.085 4.350 pw_transfer 1331 11.6 0.066 0.072 4.218 4.234 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.110 4.130 ot_diis_step 99 11.5 0.012 0.012 3.757 3.757 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.682 3.711 apply_single 110 13.6 0.000 0.000 3.682 3.711 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.660 3.663 fft_wrap_pw1pw2_140 451 13.1 0.217 0.223 3.553 3.582 density_rs2pw 110 9.6 0.004 0.004 3.166 3.545 grid_integrate_task_list 110 12.3 3.355 3.503 3.355 3.503 fft3d_ps 1111 14.6 1.148 1.185 3.352 3.360 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.352 3.353 multiply_cannon_sync_h2d 8220 15.4 2.917 3.016 2.917 3.016 cp_fm_cholesky_invert 11 10.9 2.931 2.934 2.931 2.934 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.839 2.840 cp_fm_redistribute_end 48 14.0 0.722 2.811 0.727 2.812 make_images_data 4110 15.4 0.038 0.043 2.404 2.810 hybrid_alltoall_any 4261 16.3 0.200 0.866 2.348 2.791 cp_fm_diag_elpa_base 48 14.0 1.902 2.615 2.078 2.780 wfi_extrapolate 11 7.9 0.001 0.001 2.767 2.767 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.687 2.688 rs_pw_transfer 902 11.9 0.010 0.010 2.194 2.595 grid_collocate_task_list 110 9.6 2.277 2.544 2.277 2.544 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.508 2.510 calculate_dm_sparse 110 9.5 0.001 0.001 2.459 2.500 calculate_first_density_matrix 1 7.0 0.000 0.000 2.215 2.216 potential_pw2rs 110 12.3 0.015 0.016 2.165 2.172 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.153 2.169 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.772 1.990 mp_alltoall_d11v 2046 13.8 1.674 1.919 1.674 1.919 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.899 1.915 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.774 1.782 cp_fm_cholesky_decompose 22 10.9 1.687 1.698 1.687 1.698 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.693 1.697 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.524 1.651 mp_allgather_i34 2055 14.4 0.539 1.646 0.539 1.646 dbcsr_complete_redistribute 325 12.2 0.560 0.585 1.480 1.585 jit_kernel_multiply 9 15.9 1.049 1.580 1.049 1.580 mp_waitany 9240 13.8 1.048 1.474 1.048 1.474 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.420 1.432 qs_create_task_list 11 7.9 0.000 0.001 1.232 1.337 generate_qs_task_list 11 8.9 0.379 0.447 1.232 1.337 rs_pw_transfer_RS2PW_140 121 11.5 0.162 0.165 0.793 1.202 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.171 1.192 multiply_cannon_metrocomm4 6165 15.4 0.018 0.019 0.482 1.156 rs_gather_matrices 110 12.3 0.326 0.367 0.940 1.151 multiply_cannon_metrocomm1 8220 15.4 0.021 0.021 0.743 1.150 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=55.985000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=779.636364, yerr=12.374995 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/530234f606c3418e1a725f5149ed409c2f956214_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 1.335263E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67104. MP_Allreduce 9672 819. MP_Sync 52 MP_Alltoall 1474 16505187. MP_SendRecv 2310 360267. MP_ISendRecv 2310 360267. MP_Wait 5214 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.042 0.105 86.478 86.486 qs_mol_dyn_low 1 2.0 0.003 0.006 86.073 86.082 qs_forces 11 3.9 0.004 0.008 85.942 85.944 qs_energies 11 4.9 0.001 0.001 81.772 81.776 scf_env_do_scf 11 5.9 0.001 0.001 71.821 71.821 velocity_verlet 10 3.0 0.001 0.001 55.305 55.319 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 43.869 43.869 dbcsr_multiply_generic 2055 12.4 0.119 0.120 28.093 28.164 init_scf_loop 11 6.9 0.000 0.000 27.881 27.883 prepare_preconditioner 11 7.9 0.000 0.000 25.855 25.861 make_preconditioner 11 8.9 0.000 0.000 25.855 25.861 qs_scf_new_mos 99 7.5 0.001 0.001 25.535 25.570 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.534 25.570 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.188 25.330 ot_scf_mini 99 9.5 0.002 0.002 23.798 23.820 multiply_cannon 2055 13.4 0.341 0.366 21.232 22.113 multiply_cannon_loop 2055 14.4 0.341 0.345 19.479 19.864 cp_fm_upper_to_full 70 14.2 12.550 17.936 12.550 17.936 rebuild_ks_matrix 110 8.3 0.001 0.001 14.181 14.206 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 14.181 14.206 ot_mini 99 10.5 0.001 0.001 13.298 13.320 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.849 12.871 dbcsr_complete_redistribute 325 12.2 1.027 1.045 7.222 10.304 multiply_cannon_multrec 8220 15.4 4.365 4.599 9.538 9.637 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.203 9.279 qs_ot_get_derivative 99 11.5 0.001 0.001 8.848 8.870 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.651 8.696 mp_alltoall_i22 605 13.7 5.260 8.346 5.260 8.346 mp_waitall_1 84994 16.7 7.265 7.998 7.265 7.998 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.749 7.784 calculate_rho_elec 110 8.6 0.227 0.228 7.748 7.783 sum_up_and_integrate 110 10.3 0.150 0.151 7.357 7.374 integrate_v_rspace 110 11.3 0.004 0.004 7.207 7.223 init_scf_run 11 5.9 0.000 0.001 5.658 5.658 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.657 5.657 make_m2s 4110 13.4 0.043 0.043 5.158 5.651 make_images 4110 14.4 0.885 0.937 4.971 5.464 qs_ot_get_p 110 10.4 0.001 0.001 5.388 5.411 pw_transfer 1331 11.6 0.075 0.075 5.315 5.322 dbcsr_mm_accdrv_process 11614 15.7 3.249 3.591 5.031 5.294 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 5.198 5.205 cp_fm_cholesky_invert 11 10.9 5.148 5.153 5.148 5.153 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.583 5.058 apply_single 110 13.6 0.000 0.000 4.582 5.058 multiply_cannon_metrocomm3 8220 15.4 0.018 0.019 4.594 4.990 fft_wrap_pw1pw2_140 451 13.1 0.225 0.230 4.534 4.539 ot_diis_step 99 11.5 0.015 0.016 4.435 4.435 fft3d_ps 1111 14.6 1.305 1.313 4.389 4.399 multiply_cannon_sync_h2d 8220 15.4 3.950 3.954 3.950 3.954 qs_ot_p2m_diag 48 11.0 0.151 0.156 3.917 3.925 density_rs2pw 110 9.6 0.004 0.004 3.797 3.833 grid_integrate_task_list 110 12.3 3.657 3.717 3.657 3.717 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.704 3.708 hybrid_alltoall_any 4261 16.3 0.259 0.563 2.830 3.534 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.482 3.483 make_images_data 4110 15.4 0.042 0.045 2.799 3.395 wfi_extrapolate 11 7.9 0.001 0.001 3.274 3.275 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.798 3.248 calculate_dm_sparse 110 9.5 0.001 0.001 3.099 3.121 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.899 2.899 cp_fm_diag_elpa_base 48 14.0 2.368 2.564 2.897 2.897 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.839 2.841 grid_collocate_task_list 110 9.6 2.629 2.651 2.629 2.651 potential_pw2rs 110 12.3 0.021 0.022 2.495 2.497 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.356 2.364 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.323 2.335 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.233 2.290 rs_pw_transfer 902 11.9 0.011 0.011 2.249 2.266 calculate_first_density_matrix 1 7.0 0.000 0.000 2.260 2.260 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.083 2.184 mp_alltoall_d11v 2046 13.8 1.898 2.003 1.898 2.003 cp_fm_cholesky_decompose 22 10.9 1.958 1.984 1.958 1.984 qs_create_task_list 11 7.9 0.000 0.001 1.898 1.947 generate_qs_task_list 11 8.9 0.736 0.789 1.897 1.947 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.939 1.942 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.748 1.781 jit_kernel_multiply 10 15.3 1.579 1.749 1.579 1.749 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=86.486000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1194.636364, yerr=55.524397 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/530234f606c3418e1a725f5149ed409c2f956214_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 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 11444702699520 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 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 1.086553E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 11851392 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.8 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 627.777536E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10258080 MPI messages size (bytes): total size 4.456715E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.459031E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3542056 94711185408 32768 < size <= 131072 1282176 73356279808 131072 < size <= 4194304 5107038 3151762421624 4194304 < size <= 16777216 259842 1136842803272 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4089 56701. MP_Allreduce 11263 785. MP_Sync 168 MP_Alltoall 2210 2520169. MP_SendRecv 24130 18752. MP_ISendRecv 24130 18752. MP_Wait 42150 MP_comm_split 82 MP_ISend 15900 108037. MP_IRecv 15900 108037. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.034 203.301 203.303 qs_mol_dyn_low 1 2.0 0.003 0.004 202.928 202.952 qs_forces 11 3.9 0.003 0.004 202.838 202.839 qs_energies 11 4.9 0.001 0.002 197.123 197.141 scf_env_do_scf 11 5.9 0.001 0.001 180.642 180.647 scf_env_do_scf_inner_loop 116 6.6 0.002 0.008 160.256 160.258 dbcsr_multiply_generic 2485 12.5 0.175 0.179 123.574 124.304 velocity_verlet 10 3.0 0.001 0.001 122.143 122.144 qs_scf_new_mos 116 7.6 0.001 0.001 120.282 120.606 qs_scf_loop_do_ot 116 8.6 0.001 0.001 120.282 120.605 ot_scf_mini 116 9.6 0.003 0.003 113.759 114.060 multiply_cannon 2485 13.5 0.243 0.253 100.040 102.007 multiply_cannon_loop 2485 14.5 2.096 2.181 97.937 99.720 ot_mini 116 10.6 0.001 0.001 65.471 65.800 multiply_cannon_multrec 59640 15.5 32.809 34.448 41.129 42.881 qs_ot_get_derivative 116 11.6 0.001 0.001 40.771 41.086 rebuild_ks_matrix 127 8.3 0.001 0.001 33.635 34.160 qs_ks_build_kohn_sham_matrix 127 9.3 0.015 0.017 33.634 34.159 mp_waitall_1 264810 16.5 28.327 31.022 28.327 31.022 qs_ks_update_qs_env 127 7.6 0.001 0.001 30.188 30.662 multiply_cannon_sync_h2d 59640 15.5 26.953 29.282 26.953 29.282 qs_ot_get_p 127 10.4 0.001 0.001 26.563 26.882 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 24.103 24.720 apply_single 127 13.6 0.001 0.001 24.102 24.720 ot_diis_step 116 11.6 0.007 0.008 24.267 24.269 init_scf_loop 11 6.9 0.000 0.001 20.308 20.310 qs_ot_p2m_diag 82 11.4 0.076 0.090 19.855 19.965 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 18.970 19.246 multiply_cannon_metrocomm3 59640 15.5 0.118 0.125 15.793 17.747 cp_dbcsr_syevd 82 12.4 0.004 0.005 17.232 17.233 prepare_preconditioner 11 7.9 0.000 0.000 15.690 15.743 make_preconditioner 11 8.9 0.000 0.000 15.690 15.743 make_full_inverse_cholesky 11 9.9 0.000 0.000 14.916 15.098 cp_fm_diag_elpa 82 13.4 0.000 0.001 14.407 14.413 cp_fm_redistribute_end 82 14.4 11.395 14.342 11.405 14.345 sum_up_and_integrate 127 10.3 0.089 0.107 14.260 14.276 integrate_v_rspace 127 11.3 0.003 0.004 14.171 14.191 make_m2s 4970 13.5 0.103 0.111 13.740 14.127 cp_fm_diag_elpa_base 82 14.4 2.897 13.987 2.928 14.102 make_images 4970 14.5 0.397 0.417 13.560 13.950 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.765 13.913 calculate_rho_elec 127 8.7 0.045 0.064 13.765 13.913 init_scf_run 11 5.9 0.000 0.001 12.261 12.262 scf_env_initial_rho_setup 11 6.9 0.001 0.002 12.261 12.261 density_rs2pw 127 9.7 0.006 0.007 7.152 10.446 mp_sum_l 7804 13.0 8.542 10.016 8.542 10.016 wfi_extrapolate 11 7.9 0.001 0.001 9.046 9.046 rs_pw_transfer 1038 11.9 0.016 0.018 5.670 8.960 cp_fm_cholesky_invert 11 10.9 8.896 8.904 8.896 8.904 dbcsr_mm_accdrv_process 123452 16.2 3.272 3.426 7.892 8.481 calculate_dm_sparse 127 9.5 0.001 0.001 8.330 8.429 pw_transfer 1535 11.6 0.076 0.101 8.082 8.325 multiply_cannon_metrocomm1 59640 15.5 0.089 0.093 6.090 8.207 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.008 8.139 fft_wrap_pw1pw2 1281 12.7 0.010 0.012 7.878 8.126 qs_ot_get_orbitals 116 10.6 0.001 0.001 7.778 7.866 make_images_data 4970 15.5 0.067 0.073 6.738 7.642 grid_integrate_task_list 127 12.3 7.015 7.472 7.015 7.472 hybrid_alltoall_any 5155 16.4 0.288 2.232 5.937 7.426 fft_wrap_pw1pw2_140 519 13.2 0.441 0.507 6.843 7.138 fft3d_ps 1281 14.7 2.089 2.852 6.674 6.851 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.657 6.666 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.798 5.900 mp_waitany 15900 13.9 2.643 5.871 2.643 5.871 mp_alltoall_d11v 2401 14.1 4.245 5.814 4.245 5.814 grid_collocate_task_list 127 9.7 4.533 5.754 4.533 5.754 rs_pw_transfer_RS2PW_140 138 11.5 0.276 0.296 2.123 5.384 potential_pw2rs 127 12.3 0.009 0.011 4.692 4.718 cp_fm_cholesky_decompose 22 10.9 4.593 4.605 4.593 4.605 mp_sum_d 4444 12.2 3.745 4.572 3.745 4.572 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=203.303000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=594.909091, yerr=7.537235 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/530234f606c3418e1a725f5149ed409c2f956214_performance_tests/16/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 2.183246E+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 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 830.193664E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100943E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834845264 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4093 57094. MP_Allreduce 11272 945. MP_Sync 170 MP_Alltoall 1969 5229483. MP_SendRecv 12032 47072. MP_ISendRecv 12032 47072. MP_Wait 25916 MP_comm_split 83 MP_ISend 11748 212467. MP_IRecv 11748 212467. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.029 190.476 190.477 qs_mol_dyn_low 1 2.0 0.003 0.004 190.152 190.167 qs_forces 11 3.9 0.009 0.016 190.061 190.064 qs_energies 11 4.9 0.001 0.002 183.226 183.243 scf_env_do_scf 11 5.9 0.001 0.001 166.906 166.916 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 133.969 133.972 velocity_verlet 10 3.0 0.001 0.001 120.016 120.018 dbcsr_multiply_generic 2507 12.6 0.187 0.193 97.053 98.186 qs_scf_new_mos 117 7.6 0.001 0.001 93.710 94.201 qs_scf_loop_do_ot 117 8.6 0.001 0.001 93.709 94.201 ot_scf_mini 117 9.6 0.004 0.004 88.908 89.483 multiply_cannon 2507 13.6 0.479 0.532 77.228 82.092 multiply_cannon_loop 2507 14.6 1.254 1.301 73.997 76.427 ot_mini 117 10.6 0.001 0.001 49.547 50.080 mp_waitall_1 214728 16.6 24.272 37.808 24.272 37.808 multiply_cannon_multrec 30084 15.6 21.993 26.900 31.623 36.741 rebuild_ks_matrix 128 8.3 0.001 0.001 33.474 33.997 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.019 33.473 33.997 init_scf_loop 11 6.9 0.000 0.000 32.847 32.848 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.125 30.599 multiply_cannon_metrocomm3 30084 15.6 0.094 0.101 15.634 28.553 prepare_preconditioner 11 7.9 0.000 0.000 28.441 28.498 make_preconditioner 11 8.9 0.000 0.000 28.441 28.498 qs_ot_get_derivative 117 11.6 0.001 0.002 27.612 28.186 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.088 27.600 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 21.990 23.107 apply_single 128 13.6 0.001 0.001 21.990 23.107 qs_ot_get_p 128 10.4 0.001 0.001 21.157 21.814 ot_diis_step 117 11.6 0.014 0.015 21.766 21.768 multiply_cannon_sync_h2d 30084 15.6 19.286 21.199 19.286 21.199 cp_fm_cholesky_invert 11 10.9 16.567 16.579 16.567 16.579 qs_ot_p2m_diag 83 11.4 0.187 0.215 16.399 16.437 make_m2s 5014 13.6 0.089 0.094 14.129 15.631 make_images 5014 14.6 1.166 1.359 13.919 15.423 cp_dbcsr_syevd 83 12.4 0.005 0.006 15.253 15.254 sum_up_and_integrate 128 10.3 0.116 0.131 14.858 14.891 integrate_v_rspace 128 11.3 0.003 0.003 14.741 14.780 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.132 14.167 calculate_rho_elec 128 8.7 0.088 0.105 14.131 14.167 cp_fm_diag_elpa 83 13.4 0.000 0.001 12.153 12.162 cp_fm_redistribute_end 83 14.4 7.095 12.085 7.106 12.086 cp_fm_diag_elpa_base 83 14.4 4.747 11.585 4.963 11.948 init_scf_run 11 5.9 0.000 0.001 11.511 11.512 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.510 11.512 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 10.880 11.320 multiply_cannon_metrocomm4 27577 15.6 0.094 0.108 3.755 10.771 mp_irecv_dv 69486 16.3 3.562 10.379 3.562 10.379 make_images_data 5014 15.6 0.065 0.072 8.557 10.290 density_rs2pw 128 9.7 0.006 0.007 7.544 9.973 dbcsr_mm_accdrv_process 62242 16.2 4.546 5.211 9.084 9.671 hybrid_alltoall_any 5200 16.5 0.342 1.506 7.248 9.580 pw_transfer 1547 11.6 0.087 0.108 9.016 9.089 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.789 8.869 wfi_extrapolate 11 7.9 0.001 0.001 8.388 8.388 rs_pw_transfer 1046 11.9 0.014 0.016 5.744 8.191 fft_wrap_pw1pw2_140 523 13.2 0.471 0.526 7.753 7.865 grid_integrate_task_list 128 12.3 7.182 7.706 7.182 7.706 fft3d_ps 1291 14.7 2.751 2.915 7.154 7.198 cp_fm_cholesky_decompose 22 10.9 6.939 7.026 6.939 7.026 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.141 6.905 calculate_dm_sparse 128 9.5 0.001 0.001 6.472 6.594 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.147 6.154 mp_sum_l 7870 13.0 3.906 5.964 3.906 5.964 grid_collocate_task_list 128 9.7 4.702 5.863 4.702 5.863 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.278 5.431 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.337 5.398 potential_pw2rs 128 12.3 0.014 0.016 5.115 5.131 mp_allgather_i34 2507 14.6 1.801 4.897 1.801 4.897 mp_waitany 11748 13.9 2.448 4.823 2.448 4.823 mp_alltoall_d11v 2415 14.1 4.095 4.778 4.095 4.778 rs_pw_transfer_RS2PW_140 139 11.5 0.349 0.373 2.047 4.456 dbcsr_complete_redistribute 395 12.7 0.811 0.903 3.211 4.043 mp_sum_d 4464 12.2 2.627 3.940 2.627 3.940 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=190.477000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=790.272727, yerr=3.305393 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/530234f606c3418e1a725f5149ed409c2f956214_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430458527744 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613083000832 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228655E+12 0.0% 0.0% 100.0% flops max/rank 2.950886E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806314816 0.0% 0.0% 100.0% number of processed stacks 4015680 0.0% 0.0% 100.0% average stack size 0.0 0.0 1694.9 marketing flops 145.646636E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 932.950016E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1051232 MPI messages size (bytes): total size 2.737039E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.603648E+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 284160 37245419520 131072 < size <= 4194304 665408 1004401590272 4194304 < size <= 16777216 66080 937889582784 16777216 < size 28896 757491302400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4113 56816. MP_Allreduce 11321 983. MP_Sync 170 MP_Alltoall 1724 9394160. MP_SendRecv 7998 75008. MP_ISendRecv 7998 75008. MP_Wait 21986 MP_comm_split 83 MP_ISend 11836 275177. MP_IRecv 11836 275177. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.042 0.118 179.004 179.006 qs_mol_dyn_low 1 2.0 0.003 0.004 178.493 178.508 qs_forces 11 3.9 0.012 0.050 178.397 178.401 qs_energies 11 4.9 0.005 0.023 171.688 171.698 scf_env_do_scf 11 5.9 0.001 0.002 155.784 155.784 scf_env_do_scf_inner_loop 118 6.6 0.004 0.008 120.380 120.381 velocity_verlet 10 3.0 0.001 0.001 114.901 114.904 dbcsr_multiply_generic 2527 12.6 0.182 0.190 81.716 82.841 qs_scf_new_mos 118 7.6 0.001 0.001 81.504 81.832 qs_scf_loop_do_ot 118 8.6 0.001 0.002 81.503 81.831 ot_scf_mini 118 9.6 0.005 0.013 77.349 77.778 multiply_cannon 2527 13.6 0.507 0.523 62.006 65.377 multiply_cannon_loop 2527 14.6 0.861 0.891 58.762 61.549 ot_mini 118 10.6 0.001 0.001 42.654 43.099 init_scf_loop 11 6.9 0.017 0.073 35.304 35.305 mp_waitall_1 171878 16.6 24.776 34.111 24.776 34.111 rebuild_ks_matrix 129 8.3 0.001 0.001 31.237 31.750 qs_ks_build_kohn_sham_matrix 129 9.3 0.016 0.019 31.237 31.749 prepare_preconditioner 11 7.9 0.000 0.000 31.113 31.162 make_preconditioner 11 8.9 0.000 0.001 31.113 31.162 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.745 30.213 qs_ks_update_qs_env 129 7.6 0.001 0.001 28.165 28.635 multiply_cannon_multrec 20216 15.6 13.589 17.520 22.317 26.151 multiply_cannon_metrocomm3 20216 15.6 0.061 0.064 15.165 24.539 qs_ot_get_derivative 118 11.6 0.001 0.002 22.728 23.172 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 20.014 21.073 apply_single 129 13.6 0.001 0.001 20.014 21.073 qs_ot_get_p 129 10.4 0.001 0.001 20.057 20.527 ot_diis_step 118 11.6 0.018 0.018 19.818 19.818 multiply_cannon_sync_h2d 20216 15.6 14.341 16.660 14.341 16.660 qs_ot_p2m_diag 83 11.4 0.265 0.272 15.796 15.807 make_m2s 5054 13.6 0.081 0.085 14.588 15.542 make_images 5054 14.6 1.192 1.259 14.353 15.310 sum_up_and_integrate 129 10.3 0.135 0.147 14.971 15.001 integrate_v_rspace 129 11.3 0.004 0.004 14.835 14.866 cp_dbcsr_syevd 83 12.4 0.005 0.005 14.791 14.792 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.621 14.654 calculate_rho_elec 129 8.7 0.133 0.148 14.620 14.653 cp_fm_cholesky_invert 11 10.9 14.528 14.537 14.528 14.537 cp_fm_diag_elpa 83 13.4 0.000 0.001 11.623 11.624 cp_fm_redistribute_end 83 14.4 4.396 11.557 4.411 11.560 cp_fm_diag_elpa_base 83 14.4 6.715 10.953 7.128 11.433 init_scf_run 11 5.9 0.000 0.001 10.769 10.769 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.769 10.769 make_images_data 5054 15.6 0.062 0.068 8.962 10.312 density_rs2pw 129 9.7 0.006 0.006 7.409 9.675 hybrid_alltoall_any 5240 16.5 0.438 2.001 7.871 9.560 multiply_cannon_metrocomm4 17689 15.6 0.062 0.072 3.469 9.490 pw_transfer 1559 11.6 0.087 0.109 9.240 9.360 mp_irecv_dv 50625 16.2 3.347 9.241 3.347 9.241 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 8.863 9.172 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 9.013 9.139 dbcsr_mm_accdrv_process 41830 16.2 4.336 5.141 8.179 8.325 fft_wrap_pw1pw2_140 527 13.2 0.482 0.525 7.954 8.094 grid_integrate_task_list 129 12.3 7.372 7.898 7.372 7.898 rs_pw_transfer 1054 12.0 0.014 0.015 5.322 7.608 wfi_extrapolate 11 7.9 0.001 0.001 7.592 7.592 cp_fm_upper_to_full 105 14.5 5.841 7.537 5.841 7.537 cp_fm_cholesky_decompose 22 10.9 7.342 7.370 7.342 7.370 fft3d_ps 1301 14.7 2.687 2.927 7.261 7.347 dbcsr_complete_redistribute 395 12.7 1.166 1.190 4.677 6.480 calculate_dm_sparse 129 9.5 0.001 0.001 5.785 5.910 grid_collocate_task_list 129 9.7 4.939 5.814 4.939 5.814 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.486 5.497 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 4.648 5.375 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.484 5.279 potential_pw2rs 129 12.3 0.020 0.023 4.967 4.989 mp_alltoall_d11v 2423 14.1 4.391 4.958 4.391 4.958 mp_waitany 11836 13.9 2.393 4.742 2.393 4.742 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.549 4.696 mp_sum_l 7930 13.1 3.311 4.567 3.311 4.567 mp_allgather_i34 2527 14.6 1.752 4.516 1.752 4.516 rs_pw_transfer_RS2PW_140 140 11.5 0.332 0.353 1.955 4.242 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.072 4.111 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.349 4.092 mp_alltoall_i22 718 14.1 2.043 3.968 2.043 3.968 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.810 3.811 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.596 3.639 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=179.006000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=885.363636, yerr=9.412494 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/530234f606c3418e1a725f5149ed409c2f956214_performance_tests/18/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 4.320339E+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 5927808 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.2 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.174766E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1133160 MPI messages size (bytes): total size 2.008142E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.772161E+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 315952 35695099904 131072 < size <= 4194304 709496 778939400192 4194304 < size <= 16777216 69840 660837789680 16777216 < size 30480 532676608000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4085 57191. MP_Allreduce 11251 1068. MP_Sync 168 MP_Alltoall 1700 12496381. MP_SendRecv 5842 75008. MP_ISendRecv 5842 75008. MP_Wait 22272 MP_comm_split 82 MP_ISend 14840 244848. MP_IRecv 14840 244848. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.026 0.039 187.364 187.367 qs_mol_dyn_low 1 2.0 0.005 0.010 186.922 186.935 qs_forces 11 3.9 0.003 0.003 186.787 186.797 qs_energies 11 4.9 0.003 0.016 179.538 179.548 scf_env_do_scf 11 5.9 0.002 0.011 162.557 162.564 velocity_verlet 10 3.0 0.001 0.001 124.089 124.094 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 116.186 116.188 dbcsr_multiply_generic 2485 12.5 0.187 0.192 78.493 79.270 qs_scf_new_mos 116 7.6 0.001 0.001 78.588 78.859 qs_scf_loop_do_ot 116 8.6 0.001 0.002 78.587 78.859 ot_scf_mini 116 9.6 0.003 0.004 74.147 74.449 multiply_cannon 2485 13.5 0.552 0.590 54.177 56.996 multiply_cannon_loop 2485 14.5 1.173 1.202 50.599 52.319 init_scf_loop 11 6.9 0.001 0.003 46.240 46.241 prepare_preconditioner 11 7.9 0.000 0.000 42.101 42.135 make_preconditioner 11 8.9 0.000 0.001 42.101 42.135 ot_mini 116 10.6 0.001 0.001 41.845 42.124 make_full_inverse_cholesky 11 9.9 0.000 0.000 35.659 40.670 multiply_cannon_multrec 29820 15.5 13.957 18.899 25.774 30.379 rebuild_ks_matrix 127 8.3 0.001 0.001 29.964 30.243 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.019 29.963 30.242 mp_waitall_1 146592 16.7 17.241 27.885 17.241 27.885 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.965 27.215 qs_ot_get_derivative 116 11.6 0.001 0.002 22.293 22.599 make_m2s 4970 13.5 0.093 0.097 20.057 21.069 make_images 4970 14.5 1.944 2.274 19.754 20.766 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 18.984 19.556 apply_single 127 13.6 0.001 0.001 18.983 19.555 ot_diis_step 116 11.6 0.017 0.018 19.435 19.436 qs_ot_get_p 127 10.4 0.001 0.001 18.693 19.016 cp_fm_upper_to_full 104 14.7 11.021 16.311 11.021 16.311 cp_fm_cholesky_invert 11 10.9 16.049 16.059 16.049 16.059 multiply_cannon_metrocomm3 29820 15.5 0.046 0.050 6.489 15.529 sum_up_and_integrate 127 10.3 0.140 0.153 14.781 14.803 qs_ot_p2m_diag 82 11.4 0.338 0.385 14.673 14.725 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.686 14.717 calculate_rho_elec 127 8.7 0.174 0.189 14.685 14.717 integrate_v_rspace 127 11.3 0.003 0.004 14.640 14.668 cp_dbcsr_syevd 82 12.4 0.005 0.005 13.416 13.418 dbcsr_complete_redistribute 393 12.7 1.506 1.621 8.999 12.744 multiply_cannon_sync_h2d 29820 15.5 11.623 12.697 11.623 12.697 make_images_data 4970 15.5 0.064 0.067 10.797 12.322 dbcsr_mm_accdrv_process 61748 16.2 7.223 8.114 11.396 11.926 hybrid_alltoall_any 5155 16.4 0.523 2.192 9.801 11.713 copy_fm_to_dbcsr 208 11.6 0.002 0.002 7.640 11.369 init_scf_run 11 5.9 0.001 0.004 10.939 10.940 scf_env_initial_rho_setup 11 6.9 0.002 0.004 10.938 10.940 cp_fm_diag_elpa 82 13.4 0.000 0.001 10.364 10.366 cp_fm_redistribute_end 82 14.4 1.792 10.291 1.807 10.296 cp_fm_diag_elpa_base 82 14.4 7.893 9.690 8.458 10.164 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.419 10.078 pw_transfer 1535 11.6 0.088 0.106 9.499 9.580 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.166 9.368 fft_wrap_pw1pw2 1281 12.7 0.010 0.012 9.271 9.359 mp_alltoall_i22 712 14.1 5.575 9.267 5.575 9.267 density_rs2pw 127 9.7 0.006 0.006 7.129 8.500 fft_wrap_pw1pw2_140 519 13.2 0.479 0.490 8.231 8.346 grid_integrate_task_list 127 12.3 7.460 7.859 7.460 7.859 wfi_extrapolate 11 7.9 0.001 0.001 7.713 7.713 cp_fm_cholesky_decompose 22 10.9 7.470 7.571 7.470 7.571 fft3d_ps 1281 14.7 2.753 2.838 7.509 7.562 multiply_cannon_metrocomm4 24850 15.5 0.075 0.086 2.761 7.280 mp_irecv_dv 75445 16.2 2.617 7.007 2.617 7.007 calculate_dm_sparse 127 9.5 0.001 0.001 6.098 6.177 rs_pw_transfer 1038 11.9 0.013 0.014 4.663 6.101 mp_alltoall_d11v 2401 14.1 4.932 5.740 4.932 5.740 grid_collocate_task_list 127 9.7 5.042 5.671 5.042 5.671 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.459 5.529 potential_pw2rs 127 12.3 0.022 0.022 4.643 4.653 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.389 4.511 qs_energies_init_hamiltonians 11 5.9 0.002 0.014 4.486 4.488 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.324 4.422 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.136 4.200 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=187.367000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1100.727273, yerr=30.088025 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/530234f606c3418e1a725f5149ed409c2f956214_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456039424 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613072052224 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176077312 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176077312 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.233020E+12 0.0% 0.0% 100.0% flops max/rank 5.910120E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383904 0.0% 0.0% 100.0% number of processed stacks 1976928 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.9 marketing flops 145.650931E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 1.495523E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 242784 MPI messages size (bytes): total size 1.341806E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.526748E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 115008 60297314304 4194304 < size <= 16777216 105840 554906419200 16777216 < size 20352 726592466352 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 9010 51. MP_Alltoall 9724 794507. MP_ISend 40420 2100460. MP_IRecv 40420 2099564. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4039 57686. MP_Allreduce 11094 1171. MP_Sync 88 MP_Alltoall 1724 18848034. MP_SendRecv 3870 122880. MP_ISendRecv 3870 122880. MP_Wait 16244 MP_ISend 10760 423501. MP_IRecv 10760 423501. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.024 0.067 174.866 174.868 qs_mol_dyn_low 1 2.0 0.003 0.003 174.450 174.463 qs_forces 11 3.9 0.003 0.003 174.338 174.346 qs_energies 11 4.9 0.003 0.003 166.762 166.769 scf_env_do_scf 11 5.9 0.001 0.001 149.521 149.528 velocity_verlet 10 3.0 0.001 0.001 114.549 114.553 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 114.070 114.071 qs_scf_new_mos 118 7.6 0.001 0.001 74.652 74.747 qs_scf_loop_do_ot 118 8.6 0.001 0.001 74.651 74.746 dbcsr_multiply_generic 2529 12.6 0.184 0.191 74.112 74.608 ot_scf_mini 118 9.6 0.004 0.004 70.226 70.303 multiply_cannon 2529 13.6 0.625 0.655 54.509 58.222 multiply_cannon_loop 2529 14.6 0.449 0.459 50.040 50.916 ot_mini 118 10.6 0.001 0.001 39.538 39.612 init_scf_loop 11 6.9 0.000 0.000 35.306 35.307 mp_waitall_1 126876 16.7 25.751 32.991 25.751 32.991 prepare_preconditioner 11 7.9 0.000 0.000 31.390 31.414 make_preconditioner 11 8.9 0.000 0.000 31.390 31.414 rebuild_ks_matrix 129 8.3 0.001 0.001 30.356 30.440 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.019 30.355 30.439 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.213 29.480 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.596 27.678 multiply_cannon_multrec 10116 15.6 10.493 14.509 17.849 20.924 ot_diis_step 118 11.6 0.020 0.021 19.818 19.818 qs_ot_get_derivative 118 11.6 0.001 0.002 19.652 19.727 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.316 19.621 apply_single 129 13.6 0.001 0.001 19.316 19.621 multiply_cannon_metrocomm3 10116 15.6 0.023 0.025 12.193 19.452 make_m2s 5058 13.6 0.067 0.071 15.999 17.906 qs_ot_get_p 129 10.4 0.001 0.001 17.796 17.887 cp_fm_cholesky_invert 11 10.9 17.829 17.835 17.829 17.835 make_images 5058 14.6 2.308 2.784 15.690 17.599 qs_rho_update_rho_low 129 7.7 0.001 0.001 15.720 15.757 calculate_rho_elec 129 8.7 0.261 0.270 15.720 15.757 sum_up_and_integrate 129 10.3 0.181 0.190 15.262 15.322 integrate_v_rspace 129 11.3 0.004 0.004 15.081 15.134 qs_ot_p2m_diag 84 11.4 0.501 0.507 14.039 14.053 cp_dbcsr_syevd 84 12.4 0.005 0.005 12.895 12.896 make_images_data 5058 15.6 0.053 0.060 9.697 12.296 multiply_cannon_sync_h2d 10116 15.6 11.736 12.087 11.736 12.087 hybrid_alltoall_any 5245 16.5 0.824 3.882 9.565 11.676 init_scf_run 11 5.9 0.000 0.001 10.370 10.371 scf_env_initial_rho_setup 11 6.9 0.027 0.028 10.370 10.370 pw_transfer 1559 11.6 0.086 0.095 10.051 10.105 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 9.825 9.885 cp_fm_diag_elpa 84 13.4 0.000 0.000 9.859 9.871 cp_fm_diag_elpa_base 84 14.4 9.610 9.691 9.849 9.860 fft_wrap_pw1pw2_140 527 13.2 0.500 0.523 8.673 8.736 grid_integrate_task_list 129 12.3 7.814 8.194 7.814 8.194 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 8.011 8.065 density_rs2pw 129 9.7 0.005 0.006 7.180 8.055 fft3d_ps 1301 14.7 2.729 2.829 7.971 7.996 cp_fm_cholesky_decompose 22 10.9 7.849 7.989 7.849 7.989 multiply_cannon_metrocomm1 10116 15.6 0.029 0.029 4.618 7.783 dbcsr_mm_accdrv_process 20934 16.1 2.676 3.478 6.983 7.682 wfi_extrapolate 11 7.9 0.001 0.001 7.379 7.379 mp_allgather_i34 2529 14.6 2.624 6.633 2.624 6.633 calculate_dm_sparse 129 9.5 0.001 0.001 6.076 6.169 grid_collocate_task_list 129 9.7 5.436 6.104 5.436 6.104 mp_alltoall_d11v 2429 14.1 5.072 5.577 5.072 5.577 dbcsr_complete_redistribute 397 12.7 2.131 2.211 5.173 5.558 qs_energies_init_hamiltonians 11 5.9 0.049 0.053 5.375 5.376 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.154 5.166 rs_pw_transfer 1054 12.0 0.012 0.013 4.191 5.087 potential_pw2rs 129 12.3 0.027 0.028 4.631 4.644 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.101 4.132 multiply_cannon_metrocomm4 7587 15.6 0.024 0.027 1.870 3.988 mp_irecv_dv 29102 15.9 1.831 3.915 1.831 3.915 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.553 3.859 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.680 3.705 copy_fm_to_dbcsr 210 11.7 0.002 0.002 3.402 3.701 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.675 3.685 copy_dbcsr_to_fm 187 11.8 0.004 0.004 3.513 3.612 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=174.868000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1391.454545, yerr=43.648104 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/530234f606c3418e1a725f5149ed409c2f956214_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456039424 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613072052224 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176077312 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176077312 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.233020E+12 0.0% 0.0% 100.0% flops max/rank 11.786061E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383904 0.0% 0.0% 100.0% number of processed stacks 1980288 0.0% 0.0% 100.0% average stack size 0.0 0.0 3437.1 marketing flops 145.650931E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 3.050013E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101160 MPI messages size (bytes): total size 1.144970E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.318403E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 45648 35433480192 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592466352 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4039 58620. MP_Allreduce 11094 1511. MP_Sync 88 MP_Alltoall 1724 36993632. MP_SendRecv 1806 218624. MP_ISendRecv 1806 218624. MP_Wait 9876 MP_ISend 6456 1080169. MP_IRecv 6456 1080169. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.030 0.062 297.631 297.637 qs_mol_dyn_low 1 2.0 0.003 0.004 297.103 297.115 qs_forces 11 3.9 0.003 0.003 296.947 296.951 qs_energies 11 4.9 0.001 0.002 287.649 287.661 scf_env_do_scf 11 5.9 0.001 0.001 265.483 265.498 velocity_verlet 10 3.0 0.001 0.001 214.758 214.774 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 141.255 141.258 init_scf_loop 11 6.9 0.000 0.000 123.972 123.975 prepare_preconditioner 11 7.9 0.000 0.000 118.794 118.825 make_preconditioner 11 8.9 0.000 0.000 118.794 118.825 make_full_inverse_cholesky 11 9.9 0.000 0.000 95.078 115.968 qs_scf_new_mos 118 7.6 0.001 0.001 89.965 90.090 qs_scf_loop_do_ot 118 8.6 0.001 0.001 89.964 90.089 ot_scf_mini 118 9.6 0.004 0.004 85.190 85.228 dbcsr_multiply_generic 2529 12.6 0.216 0.222 82.808 83.361 cp_fm_upper_to_full 106 14.8 53.236 75.955 53.236 75.955 multiply_cannon 2529 13.6 0.709 0.751 58.642 59.264 multiply_cannon_loop 2529 14.6 0.476 0.487 55.046 56.394 ot_mini 118 10.6 0.001 0.001 44.173 44.209 dbcsr_complete_redistribute 397 12.7 3.982 4.036 29.360 42.182 copy_fm_to_dbcsr 210 11.7 0.001 0.002 25.978 38.833 rebuild_ks_matrix 129 8.3 0.001 0.001 38.636 38.675 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.018 38.635 38.674 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 23.671 36.430 qs_ks_update_qs_env 129 7.6 0.001 0.001 35.667 35.709 mp_alltoall_i22 720 14.1 21.511 34.542 21.511 34.542 cp_fm_cholesky_invert 11 10.9 32.949 32.956 32.949 32.956 mp_waitall_1 104580 16.8 27.893 31.774 27.893 31.774 qs_ot_get_p 129 10.4 0.001 0.001 25.618 25.626 qs_ot_get_derivative 118 11.6 0.002 0.002 24.055 24.094 qs_ot_p2m_diag 84 11.4 0.890 0.895 21.628 21.657 qs_rho_update_rho_low 129 7.7 0.001 0.001 20.838 20.860 calculate_rho_elec 129 8.7 0.487 0.488 20.838 20.859 make_m2s 5058 13.6 0.076 0.079 19.649 20.562 ot_diis_step 118 11.6 0.022 0.022 20.093 20.093 make_images 5058 14.6 3.756 3.948 19.166 20.083 cp_dbcsr_syevd 84 12.4 0.006 0.006 19.907 19.908 multiply_cannon_metrocomm3 10116 15.6 0.024 0.025 18.722 19.886 sum_up_and_integrate 129 10.3 0.324 0.326 19.698 19.775 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.299 19.514 apply_single 129 13.6 0.001 0.001 19.299 19.514 integrate_v_rspace 129 11.3 0.004 0.004 19.374 19.452 multiply_cannon_multrec 10116 15.6 10.532 12.276 18.026 18.109 cp_fm_diag_elpa 84 13.4 0.000 0.000 16.719 16.719 cp_fm_diag_elpa_base 84 14.4 12.282 13.915 16.715 16.716 multiply_cannon_sync_h2d 10116 15.6 15.775 15.789 15.775 15.789 hybrid_alltoall_any 5245 16.5 1.312 3.055 10.697 12.501 pw_transfer 1559 11.6 0.093 0.095 12.468 12.480 fft_wrap_pw1pw2 1301 12.7 0.011 0.011 12.231 12.243 init_scf_run 11 5.9 0.000 0.001 12.120 12.121 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.120 12.121 make_images_data 5058 15.6 0.061 0.065 10.447 12.101 fft_wrap_pw1pw2_140 527 13.2 0.548 0.551 10.878 10.909 fft3d_ps 1301 14.7 2.769 2.792 10.252 10.264 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 9.466 9.504 dbcsr_mm_accdrv_process 20934 16.1 3.879 5.631 7.254 9.099 wfi_extrapolate 11 7.9 0.001 0.001 9.037 9.037 grid_integrate_task_list 129 12.3 8.629 8.816 8.629 8.816 mp_alltoall_d11v 2429 14.1 8.106 8.806 8.106 8.806 cp_fm_cholesky_decompose 22 10.9 8.712 8.794 8.712 8.794 density_rs2pw 129 9.7 0.005 0.005 8.372 8.501 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 8.024 8.025 calculate_dm_sparse 129 9.5 0.001 0.001 6.603 6.720 grid_collocate_task_list 129 9.7 6.407 6.453 6.407 6.453 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.234 6.328 rs_scatter_matrices 140 9.7 3.602 4.407 6.011 6.226 copy_dbcsr_to_fm 187 11.8 0.004 0.004 6.061 6.133 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=297.637000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2704.909091, yerr=176.858677 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/530234f606c3418e1a725f5149ed409c2f956214_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 1.262658E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255669. MP_Allreduce 3059 6274. MP_Sync 4 MP_Alltoall 54 6805335. 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 0.018 0.040 84.971 84.972 qs_energies 1 2.0 0.000 0.000 84.534 84.549 ls_scf 1 3.0 0.000 0.000 83.617 83.631 dbcsr_multiply_generic 111 6.7 0.017 0.022 72.544 72.695 multiply_cannon 111 7.7 0.020 0.025 55.908 57.239 multiply_cannon_loop 111 8.7 0.210 0.223 52.549 53.975 ls_scf_main 1 4.0 0.000 0.000 52.451 52.452 density_matrix_trs4 2 5.0 0.002 0.003 46.855 46.938 ls_scf_init_scf 1 4.0 0.000 0.000 28.115 28.117 ls_scf_init_matrix_S 1 5.0 0.000 0.001 26.990 27.044 mp_waitall_1 11031 10.9 22.326 25.882 22.326 25.882 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 24.855 24.883 multiply_cannon_multrec 2664 9.7 8.158 8.856 15.573 17.294 multiply_cannon_sync_h2d 2664 9.7 13.745 15.732 13.745 15.732 make_m2s 222 7.7 0.008 0.011 13.027 13.522 make_images 222 8.7 0.099 0.109 13.005 13.502 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 9.706 12.335 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.348 8.376 make_images_data 222 9.7 0.004 0.005 7.597 8.167 dbcsr_mm_accdrv_process 4760 10.4 0.509 0.616 7.034 8.036 hybrid_alltoall_any 227 10.6 0.215 1.827 6.523 7.632 dbcsr_mm_accdrv_process_sort 4760 11.4 6.327 7.259 6.327 7.259 calculate_norms 4752 9.8 5.529 6.187 5.529 6.187 apply_matrix_preconditioner 6 5.3 0.000 0.001 5.077 5.242 mp_sum_l 807 5.4 3.068 4.698 3.068 4.698 make_images_sizes 222 9.7 0.000 0.000 0.718 3.737 mp_alltoall_i44 222 10.7 0.718 3.737 0.718 3.737 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.252 3.454 arnoldi_extremal 4 6.8 0.000 0.000 3.331 3.361 arnoldi_normal_ev 4 7.8 0.001 0.003 3.331 3.361 build_subspace 16 8.4 0.009 0.012 3.231 3.233 multiply_cannon_metrocomm4 2442 9.7 0.012 0.014 2.027 3.175 mp_irecv_dv 6231 10.9 2.009 3.151 2.009 3.151 ls_scf_post 1 4.0 0.000 0.000 3.050 3.065 ls_scf_store_result 1 5.0 0.000 0.000 2.871 2.911 dbcsr_special_finalize 555 9.7 0.005 0.006 2.360 2.725 dbcsr_merge_single_wm 555 10.7 0.456 0.584 2.352 2.717 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.362 2.621 make_images_pack 222 9.7 2.206 2.615 2.207 2.617 dbcsr_sort_data 658 11.4 2.154 2.473 2.154 2.473 dbcsr_matrix_vector_mult_local 304 10.0 2.069 2.460 2.072 2.462 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.326 2.415 buffer_matrices_ensure_size 222 8.7 1.754 2.053 1.754 2.053 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.798 1.800 rebuild_ks_matrix 3 7.3 0.000 0.000 1.789 1.790 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.789 1.790 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=84.972000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1142.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/530234f606c3418e1a725f5149ed409c2f956214_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 2.101567E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 10339. MP_Sync 4 MP_Alltoall 47 15335933. MP_SendRecv 141 57600. MP_ISendRecv 141 57600. MP_Wait 687 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.026 0.042 90.726 90.805 qs_energies 1 2.0 0.000 0.000 90.259 90.336 ls_scf 1 3.0 0.000 0.000 88.941 89.017 dbcsr_multiply_generic 111 6.7 0.015 0.017 75.007 75.289 multiply_cannon 111 7.7 0.029 0.043 53.219 57.363 ls_scf_main 1 4.0 0.000 0.000 54.888 54.934 multiply_cannon_loop 111 8.7 0.115 0.122 49.932 53.355 density_matrix_trs4 2 5.0 0.002 0.003 49.113 49.294 ls_scf_init_scf 1 4.0 0.000 0.000 30.554 30.582 mp_waitall_1 9105 10.9 20.887 29.962 20.887 29.962 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.336 29.440 multiply_cannon_multrec 1332 9.7 13.067 16.681 22.394 27.169 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.945 26.971 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.664 20.688 make_m2s 222 7.7 0.006 0.007 15.076 15.748 make_images 222 8.7 1.578 1.972 15.046 15.719 dbcsr_mm_accdrv_process 4041 10.4 0.267 0.456 8.926 10.569 dbcsr_mm_accdrv_process_sort 4041 11.4 8.520 10.113 8.520 10.113 make_images_data 222 9.7 0.004 0.004 8.604 9.491 hybrid_alltoall_any 227 10.6 0.524 2.483 7.999 9.128 mp_sum_l 807 5.4 5.398 8.759 5.398 8.759 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.250 7.617 mp_irecv_dv 3311 11.0 3.231 7.545 3.231 7.545 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.124 6.975 calculate_norms 2376 9.8 6.015 6.693 6.015 6.693 multiply_cannon_sync_h2d 1332 9.7 4.772 6.200 4.772 6.200 apply_matrix_preconditioner 6 5.3 0.000 0.001 5.056 5.263 arnoldi_extremal 4 6.8 0.000 0.000 4.618 4.633 arnoldi_normal_ev 4 7.8 0.001 0.005 4.618 4.633 build_subspace 16 8.4 0.014 0.021 4.365 4.371 ls_scf_post 1 4.0 0.000 0.000 3.498 3.502 dbcsr_matrix_vector_mult 304 9.0 0.009 0.021 3.138 3.365 ls_scf_store_result 1 5.0 0.000 0.000 3.197 3.324 dbcsr_matrix_vector_mult_local 304 10.0 2.736 3.223 2.738 3.225 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.258 2.997 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.613 2.700 make_images_pack 222 9.7 2.020 2.415 2.022 2.417 mp_allgather_i34 111 8.7 0.983 2.298 0.983 2.298 dbcsr_sort_data 436 11.2 1.810 2.042 1.810 2.042 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.899 1.902 rebuild_ks_matrix 3 7.3 0.000 0.000 1.886 1.889 qs_ks_build_kohn_sham_matrix 3 8.3 0.005 0.014 1.886 1.889 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=90.805000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1726.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/530234f606c3418e1a725f5149ed409c2f956214_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 2.717004E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265470. MP_Allreduce 3058 11181. MP_Sync 4 MP_Alltoall 47 23526250. MP_SendRecv 93 57600. MP_ISendRecv 93 57600. MP_Wait 639 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.035 0.053 94.218 94.219 qs_energies 1 2.0 0.000 0.000 93.640 93.645 ls_scf 1 3.0 0.004 0.031 92.183 92.189 dbcsr_multiply_generic 111 6.7 0.015 0.018 76.869 77.170 ls_scf_main 1 4.0 0.000 0.000 57.583 57.593 multiply_cannon 111 7.7 0.048 0.101 53.016 57.125 multiply_cannon_loop 111 8.7 0.100 0.106 49.415 53.449 density_matrix_trs4 2 5.0 0.002 0.003 51.457 51.638 mp_waitall_1 7281 11.0 23.884 34.230 23.884 34.230 ls_scf_init_scf 1 4.0 0.000 0.001 30.952 30.955 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.673 29.746 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.258 27.274 multiply_cannon_multrec 888 9.7 12.521 14.948 21.161 24.167 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.186 23.509 make_m2s 222 7.7 0.006 0.007 16.926 18.157 make_images 222 8.7 1.974 2.289 16.888 18.118 make_images_data 222 9.7 0.003 0.004 9.627 10.659 hybrid_alltoall_any 227 10.6 0.620 2.868 9.316 10.591 dbcsr_mm_accdrv_process 3754 10.4 0.252 0.414 8.156 9.378 dbcsr_mm_accdrv_process_sort 3754 11.4 7.787 8.963 7.787 8.963 mp_sum_l 807 5.4 5.383 8.819 5.383 8.819 multiply_cannon_sync_h2d 888 9.7 6.045 7.286 6.045 7.286 multiply_cannon_metrocomm1 888 9.7 0.002 0.003 3.694 6.958 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.456 6.940 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.136 6.922 mp_irecv_dv 2335 11.1 2.440 6.879 2.440 6.879 apply_matrix_preconditioner 6 5.3 0.001 0.001 5.045 5.217 arnoldi_extremal 4 6.8 0.000 0.000 5.035 5.053 arnoldi_normal_ev 4 7.8 0.001 0.005 5.035 5.053 build_subspace 16 8.4 0.014 0.020 4.731 4.737 calculate_norms 1584 9.8 4.255 4.615 4.255 4.615 mp_allgather_i34 111 8.7 1.441 3.798 1.441 3.798 dbcsr_matrix_vector_mult 304 9.0 0.009 0.020 3.420 3.734 ls_scf_post 1 4.0 0.001 0.008 3.644 3.652 dbcsr_matrix_vector_mult_local 304 10.0 3.011 3.573 3.013 3.575 ls_scf_store_result 1 5.0 0.000 0.000 3.369 3.470 ls_scf_dm_to_ks 2 5.0 0.001 0.024 2.881 2.983 dbcsr_sort_data 325 11.1 1.890 2.142 1.890 2.142 make_images_pack 222 9.7 1.815 2.117 1.818 2.120 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.014 2.016 rebuild_ks_matrix 3 7.3 0.000 0.000 1.996 1.998 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.996 1.998 make_images_sizes 222 9.7 0.000 0.000 0.967 1.977 mp_alltoall_i44 222 10.7 0.967 1.976 0.967 1.976 dbcsr_data_release 9322 10.9 1.314 1.974 1.314 1.974 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=94.219000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2211.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/530234f606c3418e1a725f5149ed409c2f956214_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 3.366384E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 13371. MP_Sync 4 MP_Alltoall 47 30278988. MP_SendRecv 69 86400. MP_ISendRecv 69 86400. MP_Wait 531 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.065 0.098 97.018 97.019 qs_energies 1 2.0 0.000 0.000 96.337 96.346 ls_scf 1 3.0 0.000 0.001 94.698 94.707 dbcsr_multiply_generic 111 6.7 0.016 0.018 78.448 78.678 ls_scf_main 1 4.0 0.000 0.000 58.945 58.947 multiply_cannon 111 7.7 0.053 0.138 51.590 55.791 density_matrix_trs4 2 5.0 0.002 0.003 52.770 52.902 multiply_cannon_loop 111 8.7 0.114 0.126 46.559 49.864 ls_scf_init_scf 1 4.0 0.000 0.002 32.519 32.520 ls_scf_init_matrix_S 1 5.0 0.000 0.002 31.279 31.344 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.791 28.804 mp_waitall_1 6369 11.0 22.692 28.751 22.692 28.751 multiply_cannon_multrec 1332 9.7 14.267 17.287 22.217 26.138 make_m2s 222 7.7 0.006 0.007 21.167 22.544 make_images 222 8.7 3.137 3.594 21.117 22.496 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.154 16.405 make_images_data 222 9.7 0.004 0.004 11.830 13.420 hybrid_alltoall_any 227 10.6 0.797 3.833 11.163 12.639 dbcsr_mm_accdrv_process 3641 10.4 0.220 0.381 7.596 9.154 dbcsr_mm_accdrv_process_sort 3641 11.4 7.222 8.740 7.222 8.740 mp_sum_l 807 5.4 4.114 7.871 4.114 7.871 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.087 6.092 multiply_cannon_sync_h2d 1332 9.7 5.483 6.079 5.483 6.079 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.045 5.831 mp_irecv_dv 3229 10.9 2.022 5.758 2.022 5.758 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.540 5.230 arnoldi_extremal 4 6.8 0.000 0.000 5.202 5.219 arnoldi_normal_ev 4 7.8 0.001 0.005 5.202 5.219 build_subspace 16 8.4 0.014 0.021 4.871 4.879 apply_matrix_preconditioner 6 5.3 0.001 0.001 4.577 4.730 calculate_norms 2376 9.8 4.182 4.525 4.182 4.525 mp_allgather_i34 111 8.7 2.156 4.352 2.156 4.352 dbcsr_matrix_vector_mult 304 9.0 0.010 0.020 3.572 3.887 dbcsr_matrix_vector_mult_local 304 10.0 3.165 3.661 3.167 3.663 dbcsr_sort_data 658 11.4 3.086 3.538 3.086 3.538 ls_scf_post 1 4.0 0.000 0.001 3.234 3.242 dbcsr_special_finalize 555 9.7 0.006 0.007 2.833 3.227 dbcsr_merge_single_wm 555 10.7 0.534 0.663 2.825 3.219 ls_scf_dm_to_ks 2 5.0 0.000 0.001 3.045 3.094 ls_scf_store_result 1 5.0 0.000 0.000 2.985 3.039 dbcsr_data_release 10477 10.7 1.578 2.456 1.578 2.456 dbcsr_finalize 304 7.8 0.049 0.061 1.803 2.063 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.981 1.982 rebuild_ks_matrix 3 7.3 0.000 0.000 1.958 1.959 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.003 1.958 1.959 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.019000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2680.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/530234f606c3418e1a725f5149ed409c2f956214_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 4.631015E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265558. MP_Allreduce 3049 15663. MP_Sync 4 MP_Alltoall 47 46208988. MP_SendRecv 45 115200. MP_ISendRecv 45 115200. MP_Wait 528 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.039 0.057 99.499 99.499 qs_energies 1 2.0 0.000 0.000 98.739 98.745 ls_scf 1 3.0 0.000 0.000 96.776 96.787 dbcsr_multiply_generic 111 6.7 0.017 0.017 78.123 78.280 ls_scf_main 1 4.0 0.000 0.000 62.280 62.280 multiply_cannon 111 7.7 0.090 0.179 55.462 61.011 density_matrix_trs4 2 5.0 0.002 0.003 55.010 55.101 multiply_cannon_loop 111 8.7 0.069 0.075 51.066 52.626 mp_waitall_1 5436 11.0 26.795 32.146 26.795 32.146 ls_scf_init_scf 1 4.0 0.000 0.000 30.888 30.893 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.500 29.561 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.338 27.346 multiply_cannon_multrec 444 9.7 13.981 16.243 21.111 23.218 make_m2s 222 7.7 0.004 0.005 17.868 20.254 make_images 222 8.7 3.723 4.399 17.806 20.194 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.281 17.066 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.251 14.407 make_images_data 222 9.7 0.003 0.004 10.062 12.448 hybrid_alltoall_any 227 10.6 0.786 3.757 9.915 12.365 dbcsr_mm_accdrv_process 3003 10.4 0.186 0.343 6.836 7.987 dbcsr_mm_accdrv_process_sort 3003 11.4 6.521 7.641 6.521 7.641 multiply_cannon_sync_h2d 444 9.7 6.559 7.578 6.559 7.578 mp_allgather_i34 111 8.7 2.632 7.005 2.632 7.005 arnoldi_extremal 4 6.8 0.000 0.000 5.761 5.776 arnoldi_normal_ev 4 7.8 0.002 0.004 5.761 5.775 build_subspace 16 8.4 0.015 0.020 5.378 5.387 apply_matrix_preconditioner 6 5.3 0.001 0.001 4.545 4.661 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.659 4.593 mp_irecv_dv 1241 11.2 1.641 4.578 1.641 4.578 mp_sum_l 807 5.4 2.833 4.555 2.833 4.555 dbcsr_matrix_vector_mult 304 9.0 0.011 0.020 4.144 4.351 dbcsr_matrix_vector_mult_local 304 10.0 3.666 4.151 3.668 4.153 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.685 3.763 calculate_norms 792 9.8 3.560 3.673 3.560 3.673 ls_scf_post 1 4.0 0.000 0.000 3.609 3.614 make_images_sizes 222 9.7 0.000 0.000 1.153 3.542 mp_alltoall_i44 222 10.7 1.153 3.541 1.153 3.541 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.929 3.443 ls_scf_store_result 1 5.0 0.000 0.000 3.396 3.433 dbcsr_finalize 304 7.8 0.062 0.077 2.198 2.243 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.214 2.215 rebuild_ks_matrix 3 7.3 0.000 0.000 2.181 2.182 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.181 2.182 dbcsr_merge_all 275 8.9 0.471 0.517 2.049 2.088 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=99.499000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3633.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/530234f606c3418e1a725f5149ed409c2f956214_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 8.751895E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284111. MP_Allreduce 3043 21950. MP_Sync 4 MP_Alltoall 47 88727262. MP_SendRecv 42 732600. MP_ISendRecv 42 732600. MP_Wait 267 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.092 0.127 108.183 108.184 qs_energies 1 2.0 0.000 0.000 106.745 106.763 ls_scf 1 3.0 0.000 0.000 103.803 103.819 dbcsr_multiply_generic 111 6.7 0.023 0.027 77.424 77.556 ls_scf_main 1 4.0 0.000 0.000 65.960 65.961 density_matrix_trs4 2 5.0 0.002 0.003 56.651 56.707 multiply_cannon 111 7.7 0.150 0.246 49.836 51.778 multiply_cannon_loop 111 8.7 0.068 0.069 46.308 47.353 ls_scf_init_scf 1 4.0 0.000 0.000 34.183 34.184 ls_scf_init_matrix_S 1 5.0 0.000 0.001 32.521 32.535 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.696 29.704 mp_waitall_1 4527 11.1 22.037 26.038 22.037 26.038 make_m2s 222 7.7 0.005 0.005 23.894 24.924 make_images 222 8.7 4.584 5.009 23.788 24.816 multiply_cannon_multrec 444 9.7 17.884 18.560 22.548 23.156 make_images_data 222 9.7 0.003 0.003 13.139 15.607 hybrid_alltoall_any 227 10.6 1.660 3.620 12.942 15.566 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.320 10.728 multiply_cannon_sync_h2d 444 9.7 8.846 8.898 8.846 8.898 arnoldi_extremal 4 6.8 0.000 0.000 7.208 7.219 arnoldi_normal_ev 4 7.8 0.003 0.009 7.208 7.219 build_subspace 16 8.4 0.026 0.036 6.658 6.669 dbcsr_matrix_vector_mult 304 9.0 0.016 0.032 5.336 5.483 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.265 5.352 apply_matrix_preconditioner 6 5.3 0.001 0.001 5.012 5.267 dbcsr_matrix_vector_mult_local 304 10.0 4.934 5.225 4.936 5.228 dbcsr_mm_accdrv_process 1814 10.4 0.252 0.318 4.493 4.631 dbcsr_mm_accdrv_process_sort 1814 11.4 4.194 4.337 4.194 4.337 ls_scf_post 1 4.0 0.000 0.000 3.659 3.676 mp_allgather_i34 111 8.7 1.165 3.657 1.165 3.657 make_images_sizes 222 9.7 0.000 0.000 1.424 3.480 mp_alltoall_i44 222 10.7 1.424 3.480 1.424 3.480 ls_scf_store_result 1 5.0 0.000 0.000 3.403 3.415 calculate_norms 792 9.8 3.243 3.279 3.243 3.279 dbcsr_finalize 304 7.8 0.082 0.089 3.091 3.200 dbcsr_merge_all 275 8.9 0.893 0.924 2.877 2.978 qs_energies_init_hamiltonians 1 3.0 0.002 0.002 2.912 2.912 dbcsr_complete_redistribute 5 7.6 1.459 1.486 2.792 2.899 dbcsr_data_release 12724 10.6 2.336 2.837 2.336 2.837 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.586 2.587 matrix_ls_to_qs 2 6.0 0.000 0.000 2.425 2.543 dbcsr_sort_data 325 11.1 2.443 2.522 2.443 2.522 rebuild_ks_matrix 3 7.3 0.000 0.000 2.518 2.520 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.518 2.520 dbcsr_new_transposed 4 7.5 0.245 0.256 2.259 2.278 dbcsr_frobenius_norm 74 6.6 2.054 2.127 2.193 2.238 dbcsr_add_d 103 6.2 0.000 0.000 2.135 2.211 dbcsr_add_anytype 103 7.2 0.860 0.892 2.134 2.211 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=108.184000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6875.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 530234f606c3418e1a725f5149ed409c2f956214 Summary: empty Status: OK