=== 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: ea454b3df0bf366fae64a409da442cd2660d5a8e ################# 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 7.4.3, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Author: Matthias Krack (03.02.2023) # # \ 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=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.1.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.1 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(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/ea454b3df0bf366fae64a409da442cd2660d5a8e_performance_tests/01 job id: 45315878 --- 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/ea454b3df0bf366fae64a409da442cd2660d5a8e_performance_tests/02 job id: 45315879 --- 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/ea454b3df0bf366fae64a409da442cd2660d5a8e_performance_tests/03 job id: 45315880 --- 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/ea454b3df0bf366fae64a409da442cd2660d5a8e_performance_tests/04 job id: 45315881 --- 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/ea454b3df0bf366fae64a409da442cd2660d5a8e_performance_tests/05 job id: 45315882 --- 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/ea454b3df0bf366fae64a409da442cd2660d5a8e_performance_tests/06 job id: 45315883 --- 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/ea454b3df0bf366fae64a409da442cd2660d5a8e_performance_tests/07 job id: 45315887 --- 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/ea454b3df0bf366fae64a409da442cd2660d5a8e_performance_tests/08 job id: 45315889 --- 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/ea454b3df0bf366fae64a409da442cd2660d5a8e_performance_tests/09 job id: 45315890 --- 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/ea454b3df0bf366fae64a409da442cd2660d5a8e_performance_tests/10 job id: 45315894 --- 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/ea454b3df0bf366fae64a409da442cd2660d5a8e_performance_tests/11 job id: 45315895 --- 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/ea454b3df0bf366fae64a409da442cd2660d5a8e_performance_tests/12 job id: 45315899 --- 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/ea454b3df0bf366fae64a409da442cd2660d5a8e_performance_tests/13 job id: 45315901 --- 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/ea454b3df0bf366fae64a409da442cd2660d5a8e_performance_tests/14 job id: 45315903 --- 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/ea454b3df0bf366fae64a409da442cd2660d5a8e_performance_tests/15 job id: 45315907 --- 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/ea454b3df0bf366fae64a409da442cd2660d5a8e_performance_tests/16 job id: 45315909 --- 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/ea454b3df0bf366fae64a409da442cd2660d5a8e_performance_tests/17 job id: 45315911 --- 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/ea454b3df0bf366fae64a409da442cd2660d5a8e_performance_tests/18 job id: 45315913 --- 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/ea454b3df0bf366fae64a409da442cd2660d5a8e_performance_tests/19 job id: 45315915 --- 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/ea454b3df0bf366fae64a409da442cd2660d5a8e_performance_tests/20 job id: 45315916 --- 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/ea454b3df0bf366fae64a409da442cd2660d5a8e_performance_tests/21 job id: 45315917 --- 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/ea454b3df0bf366fae64a409da442cd2660d5a8e_performance_tests/22 job id: 45315919 --- 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/ea454b3df0bf366fae64a409da442cd2660d5a8e_performance_tests/23 job id: 45315921 --- 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/ea454b3df0bf366fae64a409da442cd2660d5a8e_performance_tests/24 job id: 45315923 --- 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/ea454b3df0bf366fae64a409da442cd2660d5a8e_performance_tests/25 job id: 45315924 --- 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/ea454b3df0bf366fae64a409da442cd2660d5a8e_performance_tests/26 job id: 45315926 --- Point --- name: 510 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/1r/12t) --- Point --- name: 511 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: 512 H2O (4 NVE MD steps on 64 nodes) input file: benchmarks/QS/00512_H2O/H2O-512_md.inp required files: [] output file: result.log # nodes = 64 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ea454b3df0bf366fae64a409da442cd2660d5a8e_performance_tests/27 job id: 45315927 --- Point --- name: 601 plot: h2o_512_md regex: CP2K label: (64n/12r/1t) --- Point --- name: 602 plot: h2o_512_md_mem regex: Estimated peak process memory label: (64n/12r/1t) ~~~~~~~ END TEST ~~~~~~~ === END TESTS (description) === ===== PLOTS (description) ===== ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2_mem", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md_mem", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md_mem", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md_mem", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls_mem", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_512_md", title="512 H2O (4 NVE MD steps on 64 nodes)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_512_md_mem", title="512 H2O (4 NVE MD steps on 64 nodes)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" === END PLOTS (description) === ============ RESULTS ============ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ea454b3df0bf366fae64a409da442cd2660d5a8e_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 344 9. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.023 0.037 137.337 137.338 farming_run 1 2.0 136.601 136.601 137.299 137.303 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.461494E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 230 1103589. MP_Allreduce 491 2254389. MP_Sync 25 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 10 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.027 117.075 117.075 qs_energies 1 2.0 0.000 0.000 116.850 116.852 mp2_main 1 3.0 0.000 0.000 114.626 114.627 mp2_gpw_main 1 4.0 0.020 0.028 113.554 113.555 mp2_ri_gpw_compute_in 1 5.0 0.175 0.210 94.198 94.661 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 55.589 56.052 mp2_eri_3c_integrate_gpw 272 7.0 0.154 0.170 41.920 47.331 get_2c_integrals 1 6.0 0.001 0.021 37.711 38.435 integrate_v_rspace 273 8.0 0.439 0.450 25.223 30.459 pw_transfer 6555 10.6 0.372 0.397 27.598 28.308 fft_wrap_pw1pw2 5465 11.4 0.044 0.046 26.233 26.797 grid_integrate_task_list 273 9.0 21.019 26.773 21.019 26.773 fft_wrap_pw1pw2_100 2178 12.4 1.249 1.485 23.754 24.303 compute_2c_integrals 1 7.0 0.003 0.003 19.894 19.894 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 19.077 19.560 mp2_eri_2c_integrate_gpw 1 9.0 2.369 2.445 19.074 19.556 rpa_ri_compute_en 1 5.0 0.008 0.010 19.248 19.509 cp_fm_cholesky_decompose 12 8.2 17.951 18.666 17.951 18.666 cholesky_decomp 1 7.0 0.000 0.000 16.664 17.372 fft3d_s 5443 13.4 16.270 16.658 16.292 16.680 ao_to_mo_and_store_B_mult_1 272 7.0 10.851 15.592 10.851 15.592 calculate_wavefunction 272 8.0 5.476 5.620 12.616 13.225 rpa_num_int 1 6.0 0.000 0.001 10.762 10.762 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.732 10.758 calc_mat_Q 8 8.0 0.000 0.000 9.396 9.478 contract_S_to_Q 8 9.0 0.000 0.000 8.820 8.902 calc_potential_gpw 544 9.5 0.005 0.006 8.274 8.713 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.318 8.607 potential_pw2rs 545 10.0 0.107 0.108 7.789 8.514 parallel_gemm_fm 14 9.1 0.000 0.000 8.409 8.491 parallel_gemm_fm_cosma 14 10.1 8.409 8.491 8.409 8.491 create_integ_mat 1 6.0 0.006 0.008 7.920 7.920 collocate_single_gaussian 272 10.0 0.039 0.042 7.574 7.847 array2fm 1 7.0 0.000 0.000 6.823 7.235 pw_scatter_s 2720 13.7 4.408 4.650 4.408 4.650 pw_gather_s 2722 13.2 3.889 4.227 3.889 4.227 array2fm_buffer_send 1 8.0 3.021 3.179 3.021 3.179 pw_poisson_solve 545 10.5 1.128 1.190 2.204 2.398 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=113.555323, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2733.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ea454b3df0bf366fae64a409da442cd2660d5a8e_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 0.032 0.043 397.983 397.985 farming_run 1 2.0 396.936 396.946 397.931 397.937 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827128576 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788821 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.227190E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 743 386399. MP_Allreduce 1941 22272. MP_Sync 37 MP_Alltoall 77 8998963. 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 0.043 209.380 209.380 qs_energies 1 2.0 0.000 0.001 209.148 209.156 scf_env_do_scf 1 3.0 0.000 0.000 106.040 106.041 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.081 105.090 rebuild_ks_matrix 4 6.0 0.000 0.000 105.080 105.088 qs_ks_build_kohn_sham_matrix 4 7.0 0.057 0.065 105.080 105.088 hfx_ks_matrix 4 8.0 0.001 0.001 104.675 104.679 integrate_four_center 4 9.0 0.145 0.550 104.674 104.678 mp2_main 1 3.0 0.001 0.013 102.787 102.795 mp2_gpw_main 1 4.0 0.037 0.080 101.843 101.858 integrate_four_center_main 4 10.0 0.122 0.576 96.916 98.879 integrate_four_center_bin 264 11.0 96.795 98.858 96.795 98.858 init_scf_loop 1 4.0 0.000 0.000 91.938 91.939 mp2_ri_gpw_compute_in 1 5.0 0.067 0.103 74.932 75.975 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.492 55.536 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.163 42.189 47.255 integrate_v_rspace 95 8.0 0.398 0.558 28.576 33.479 pw_transfer 2240 10.6 0.144 0.161 29.967 30.391 ao_to_mo_and_store_B_mult_1 91 7.0 10.624 29.465 10.624 29.465 fft_wrap_pw1pw2 1868 11.4 0.017 0.020 28.963 29.416 grid_integrate_task_list 95 9.0 23.841 28.943 23.841 28.943 mp2_ri_gpw_compute_en 1 5.0 0.056 0.070 26.735 28.488 fft_wrap_pw1pw2_100 730 12.4 1.294 1.462 26.648 27.129 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.843 1.936 25.048 25.058 get_2c_integrals 1 6.0 0.000 0.000 20.322 20.372 compute_2c_integrals 1 7.0 0.003 0.003 19.297 19.304 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 18.914 19.156 mp2_eri_2c_integrate_gpw 1 9.0 1.739 1.881 18.912 19.154 fft3d_s 1823 13.4 18.428 18.704 18.442 18.717 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.101 14.101 calculate_wavefunction 91 8.0 2.025 2.052 9.748 9.973 potential_pw2rs 186 10.0 0.034 0.035 8.664 9.337 mp2_ri_gpw_compute_en_expansio 172 7.0 0.557 0.580 8.786 9.189 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.267 8.661 local_gemm 172 8.0 8.228 8.613 8.228 8.613 mp2_ri_gpw_compute_en_comm 22 7.0 0.494 0.515 8.036 8.466 calc_potential_gpw 182 9.5 0.002 0.002 7.932 8.172 collocate_single_gaussian 91 10.0 0.017 0.024 7.901 8.126 mp_sendrecv_dm3 2068 8.0 6.086 6.511 6.086 6.511 mp2_ri_gpw_compute_en_ener 172 7.0 6.338 6.415 6.338 6.415 pw_gather_s 912 13.2 4.905 5.515 4.905 5.515 mp_sync 37 10.5 2.351 4.749 2.351 4.749 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.848541, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1514.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ea454b3df0bf366fae64a409da442cd2660d5a8e_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 452.116480E+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 272. MP_Sync 530 MP_Alltoall 2083 578578. 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 0.029 0.072 58.382 58.383 qs_mol_dyn_low 1 2.0 0.003 0.004 58.010 58.017 qs_forces 11 3.9 0.004 0.014 57.945 57.947 qs_energies 11 4.9 0.002 0.011 56.348 56.375 scf_env_do_scf 11 5.9 0.000 0.001 49.489 49.489 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 47.066 47.067 qs_scf_new_mos 108 7.5 0.000 0.001 35.428 35.721 qs_scf_loop_do_ot 108 8.5 0.000 0.001 35.427 35.720 dbcsr_multiply_generic 2286 12.5 0.093 0.098 34.974 35.407 ot_scf_mini 108 9.5 0.002 0.002 33.676 33.869 velocity_verlet 10 3.0 0.001 0.001 28.578 28.579 multiply_cannon 2286 13.5 0.189 0.196 26.729 28.386 multiply_cannon_loop 2286 14.5 1.475 1.565 25.815 27.507 ot_mini 108 10.5 0.001 0.001 20.221 20.461 qs_ot_get_derivative 108 11.5 0.001 0.001 17.199 17.367 mp_waitall_1 245248 16.5 9.176 15.663 9.176 15.663 multiply_cannon_metrocomm3 54864 15.5 0.069 0.075 6.153 13.409 multiply_cannon_multrec 54864 15.5 4.206 6.434 7.925 11.436 qs_ot_get_p 119 10.4 0.001 0.003 8.713 9.033 rebuild_ks_matrix 119 8.3 0.000 0.000 8.755 8.905 qs_ks_build_kohn_sham_matrix 119 9.3 0.011 0.014 8.754 8.905 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.744 7.885 mp_sum_l 7207 12.9 5.647 7.485 5.647 7.485 multiply_cannon_sync_h2d 54864 15.5 5.849 7.084 5.849 7.084 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 6.001 6.468 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.653 5.775 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.696 5.748 init_scf_run 11 5.9 0.000 0.001 5.503 5.504 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.503 5.503 sum_up_and_integrate 119 10.3 0.012 0.014 5.189 5.198 qs_rho_update_rho_low 119 7.7 0.001 0.001 5.077 5.193 calculate_rho_elec 119 8.7 0.012 0.017 5.077 5.192 integrate_v_rspace 119 11.3 0.002 0.003 5.177 5.185 dbcsr_mm_accdrv_process 76910 16.1 1.156 1.852 3.640 5.043 cp_dbcsr_syevd 50 12.0 0.002 0.003 4.830 4.831 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.526 4.527 cp_fm_redistribute_end 50 14.0 2.311 4.490 2.324 4.495 cp_fm_diag_elpa_base 50 14.0 2.163 4.353 2.169 4.364 rs_pw_transfer 974 11.9 0.012 0.013 3.519 3.618 calculate_dm_sparse 119 9.5 0.000 0.001 3.266 3.411 jit_kernel_multiply 13 15.8 2.419 3.139 2.419 3.139 multiply_cannon_metrocomm1 54864 15.5 0.055 0.059 1.910 3.124 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.944 3.120 apply_single 119 13.6 0.000 0.000 2.944 3.120 density_rs2pw 119 9.7 0.004 0.005 2.965 3.101 calculate_first_density_matrix 1 7.0 0.001 0.006 3.090 3.096 ot_diis_step 108 11.5 0.006 0.006 2.748 2.749 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.525 2.633 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.567 2.569 pw_transfer 1439 11.6 0.053 0.059 2.444 2.545 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 2.367 2.472 potential_pw2rs 119 12.3 0.004 0.004 2.433 2.466 init_scf_loop 11 6.9 0.001 0.011 2.398 2.400 make_m2s 4572 13.5 0.054 0.057 2.287 2.367 wfi_extrapolate 11 7.9 0.001 0.001 2.322 2.323 acc_transpose_blocks 54864 15.5 0.231 0.260 1.781 2.293 make_images 4572 14.5 0.133 0.139 2.204 2.284 fft3d_ps 1201 14.6 0.373 0.481 2.139 2.238 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.156 2.196 grid_integrate_task_list 119 12.3 2.030 2.126 2.030 2.126 mp_sum_d 4135 12.0 1.424 2.016 1.424 2.016 fft_wrap_pw1pw2_140 487 13.2 0.081 0.096 1.811 1.918 mp_alltoall_d11v 2130 13.8 1.696 1.879 1.696 1.879 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.692 1.728 mp_waitany 12084 13.8 1.347 1.514 1.347 1.514 grid_collocate_task_list 119 9.7 1.290 1.367 1.290 1.367 dbcsr_dot_sd 1205 11.9 0.051 0.062 0.813 1.221 make_images_sizes 4572 15.5 0.004 0.005 0.959 1.216 mp_alltoall_z22v 1201 16.6 1.137 1.215 1.137 1.215 mp_alltoall_i44 4572 16.5 0.955 1.211 0.955 1.211 prepare_preconditioner 11 7.9 0.000 0.001 1.166 1.198 make_preconditioner 11 8.9 0.000 0.001 1.166 1.198 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=58.383000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=431.000000, yerr=1.414214 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ea454b3df0bf366fae64a409da442cd2660d5a8e_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 490.016768E+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 54 MP_Alltoall 2060 1030478. 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 0.053 0.101 40.890 40.892 qs_mol_dyn_low 1 2.0 0.003 0.004 40.464 40.475 qs_forces 11 3.9 0.009 0.027 40.397 40.398 qs_energies 11 4.9 0.002 0.011 38.546 38.552 scf_env_do_scf 11 5.9 0.000 0.001 32.671 32.671 scf_env_do_scf_inner_loop 108 6.5 0.002 0.008 29.976 29.978 dbcsr_multiply_generic 2286 12.5 0.100 0.103 22.229 22.648 qs_scf_new_mos 108 7.5 0.001 0.001 20.487 20.737 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.487 20.736 ot_scf_mini 108 9.5 0.002 0.003 19.591 19.761 velocity_verlet 10 3.0 0.003 0.014 18.811 18.811 multiply_cannon 2286 13.5 0.208 0.216 17.048 18.656 multiply_cannon_loop 2286 14.5 0.897 0.968 15.896 17.496 ot_mini 108 10.5 0.001 0.001 12.040 12.276 mp_waitall_1 200699 16.5 6.004 11.477 6.004 11.477 multiply_cannon_metrocomm3 27432 15.5 0.067 0.071 4.340 9.991 qs_ot_get_derivative 108 11.5 0.001 0.001 9.552 9.727 multiply_cannon_multrec 27432 15.5 1.966 4.263 6.328 9.193 rebuild_ks_matrix 119 8.3 0.000 0.000 7.520 7.654 qs_ks_build_kohn_sham_matrix 119 9.3 0.014 0.028 7.519 7.654 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.550 6.672 dbcsr_mm_accdrv_process 47894 16.0 3.307 5.342 4.293 6.213 qs_ot_get_p 119 10.4 0.002 0.012 4.678 4.915 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.732 4.609 init_scf_run 11 5.9 0.000 0.001 4.577 4.578 scf_env_initial_rho_setup 11 6.9 0.001 0.003 4.577 4.578 sum_up_and_integrate 119 10.3 0.025 0.027 4.462 4.483 integrate_v_rspace 119 11.3 0.002 0.003 4.437 4.458 mp_sum_l 7207 12.9 2.218 4.234 2.218 4.234 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.099 4.192 apply_single 119 13.6 0.000 0.000 3.099 4.192 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.964 4.002 calculate_rho_elec 119 8.7 0.021 0.024 3.964 4.001 rs_pw_transfer 974 11.9 0.010 0.011 2.841 3.310 qs_ot_p2m_diag 50 11.0 0.008 0.012 3.145 3.170 calculate_first_density_matrix 1 7.0 0.000 0.002 3.027 3.030 multiply_cannon_sync_h2d 27432 15.5 2.169 2.850 2.169 2.850 make_m2s 4572 13.5 0.053 0.054 2.577 2.826 density_rs2pw 119 9.7 0.004 0.004 2.286 2.768 make_images 4572 14.5 0.200 0.236 2.489 2.737 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.705 2.705 init_scf_loop 11 6.9 0.001 0.007 2.668 2.670 calculate_dm_sparse 119 9.5 0.000 0.001 2.362 2.444 ot_diis_step 108 11.5 0.011 0.011 2.428 2.429 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.347 2.348 cp_fm_redistribute_end 50 14.0 1.188 2.313 1.192 2.316 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.190 2.282 cp_fm_diag_elpa_base 50 14.0 1.090 2.209 1.119 2.257 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.233 2.236 pw_transfer 1439 11.6 0.065 0.072 2.161 2.217 potential_pw2rs 119 12.3 0.006 0.006 2.176 2.201 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.069 2.128 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.013 2.051 jit_kernel_multiply 8 16.1 0.933 1.969 0.933 1.969 grid_integrate_task_list 119 12.3 1.838 1.950 1.838 1.950 fft3d_ps 1201 14.6 0.516 0.570 1.770 1.825 prepare_preconditioner 11 7.9 0.000 0.000 1.664 1.692 make_preconditioner 11 8.9 0.000 0.001 1.664 1.692 fft_wrap_pw1pw2_140 487 13.2 0.079 0.086 1.617 1.673 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.558 1.615 make_images_data 4572 15.5 0.044 0.050 1.179 1.605 acc_transpose_blocks 27432 15.5 0.110 0.114 1.269 1.604 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.588 1.597 hybrid_alltoall_any 4725 16.4 0.051 0.112 1.047 1.539 wfi_extrapolate 11 7.9 0.001 0.001 1.494 1.494 mp_alltoall_d11v 2130 13.8 1.252 1.403 1.252 1.403 mp_allgather_i34 2286 14.5 0.593 1.389 0.593 1.389 grid_collocate_task_list 119 9.7 1.228 1.362 1.228 1.362 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.192 1.238 mp_sum_d 4135 12.0 0.624 1.088 0.624 1.088 rs_pw_transfer_RS2PW_140 130 11.5 0.139 0.148 0.582 1.061 mp_waitany 5720 13.7 0.558 1.053 0.558 1.053 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.030 1.043 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 0.988 0.992 acc_transpose_blocks_kernels 27432 16.5 0.182 0.270 0.726 0.979 mp_alltoall_z22v 1201 16.6 0.840 0.959 0.840 0.959 rs_pw_transfer_PW2RS_50 119 14.3 0.588 0.607 0.799 0.866 make_images_sizes 4572 15.5 0.005 0.005 0.577 0.824 mp_alltoall_i44 4572 16.5 0.572 0.819 0.572 0.819 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=40.892000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=464.727273, yerr=1.958242 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ea454b3df0bf366fae64a409da442cd2660d5a8e_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 522.117120E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63497. MP_Allreduce 10075 307. MP_Sync 54 MP_Alltoall 1821 1607811. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.032 0.077 42.202 42.203 qs_mol_dyn_low 1 2.0 0.003 0.004 41.855 41.863 qs_forces 11 3.9 0.003 0.010 41.792 41.792 qs_energies 11 4.9 0.004 0.025 39.945 39.949 scf_env_do_scf 11 5.9 0.001 0.001 33.804 33.805 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 29.904 29.905 velocity_verlet 10 3.0 0.001 0.001 20.624 20.625 dbcsr_multiply_generic 2286 12.5 0.098 0.101 19.033 19.088 qs_scf_new_mos 108 7.5 0.001 0.001 18.432 18.454 qs_scf_loop_do_ot 108 8.5 0.001 0.001 18.431 18.453 ot_scf_mini 108 9.5 0.002 0.003 17.624 17.639 multiply_cannon 2286 13.5 0.198 0.205 14.109 14.739 multiply_cannon_loop 2286 14.5 0.638 0.670 13.023 13.733 ot_mini 108 10.5 0.001 0.001 10.399 10.414 qs_ot_get_derivative 108 11.5 0.001 0.001 8.769 8.785 rebuild_ks_matrix 119 8.3 0.000 0.000 8.408 8.431 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.015 8.407 8.430 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.439 7.462 multiply_cannon_multrec 18288 15.5 1.921 2.870 7.025 7.388 mp_waitall_1 158411 16.6 4.962 6.273 4.962 6.273 dbcsr_mm_accdrv_process 38222 16.0 4.432 5.674 5.021 5.881 sum_up_and_integrate 119 10.3 0.030 0.031 5.726 5.732 integrate_v_rspace 119 11.3 0.003 0.004 5.696 5.705 qs_rho_update_rho_low 119 7.7 0.001 0.001 5.364 5.379 calculate_rho_elec 119 8.7 0.031 0.032 5.363 5.378 rs_pw_transfer 974 11.9 0.009 0.010 5.022 5.233 qs_ot_get_p 119 10.4 0.001 0.001 4.797 4.822 init_scf_run 11 5.9 0.000 0.001 4.575 4.576 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.575 4.575 density_rs2pw 119 9.7 0.004 0.004 3.793 4.016 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.244 3.865 init_scf_loop 11 6.9 0.001 0.007 3.859 3.861 potential_pw2rs 119 12.3 0.007 0.008 3.345 3.384 make_m2s 4572 13.5 0.045 0.046 3.119 3.320 make_images 4572 14.5 0.190 0.203 3.033 3.234 qs_ot_p2m_diag 50 11.0 0.012 0.013 3.219 3.224 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.896 2.897 pw_transfer 1439 11.6 0.066 0.070 2.809 2.854 calculate_first_density_matrix 1 7.0 0.001 0.004 2.798 2.799 prepare_preconditioner 11 7.9 0.000 0.000 2.785 2.787 make_preconditioner 11 8.9 0.000 0.001 2.785 2.787 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.715 2.758 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.580 2.690 multiply_cannon_metrocomm3 18288 15.5 0.044 0.045 1.627 2.514 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.474 2.485 cp_fm_diag_elpa_base 50 14.0 2.435 2.458 2.471 2.481 fft3d_ps 1201 14.6 0.533 0.551 2.398 2.445 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.132 2.442 apply_single 119 13.6 0.000 0.000 2.132 2.442 calculate_dm_sparse 119 9.5 0.000 0.001 2.227 2.238 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.059 2.069 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.977 1.979 grid_integrate_task_list 119 12.3 1.788 1.919 1.788 1.919 fft_wrap_pw1pw2_140 487 13.2 0.086 0.090 1.823 1.864 make_images_data 4572 15.5 0.044 0.048 1.426 1.811 mp_sum_l 7207 12.9 1.423 1.790 1.423 1.790 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.700 1.707 wfi_extrapolate 11 7.9 0.001 0.001 1.704 1.704 mp_sendrecv_dv 11067 12.7 1.550 1.604 1.550 1.604 ot_diis_step 108 11.5 0.011 0.011 1.586 1.587 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.535 1.541 hybrid_alltoall_any 4725 16.4 0.055 0.114 1.184 1.537 multiply_cannon_sync_h2d 18288 15.5 1.339 1.509 1.339 1.509 mp_alltoall_z22v 1201 16.6 1.390 1.508 1.390 1.508 make_images_sizes 4572 15.5 0.005 0.006 1.004 1.471 mp_alltoall_i44 4572 16.5 0.999 1.466 0.999 1.466 jit_kernel_multiply 7 16.2 0.535 1.453 0.535 1.453 mp_alltoall_d11v 2130 13.8 1.197 1.380 1.197 1.380 grid_collocate_task_list 119 9.7 1.214 1.357 1.214 1.357 acc_transpose_blocks 18288 15.5 0.077 0.079 1.311 1.332 cp_fm_cholesky_invert 11 10.9 1.295 1.301 1.295 1.301 rs_pw_transfer_PW2RS_50 119 14.3 0.401 0.414 1.108 1.222 qs_energies_init_hamiltonians 11 5.9 0.012 0.048 1.154 1.170 rs_pw_transfer_RS2PW_50 119 11.7 0.240 0.250 1.042 1.133 mp_allgather_i34 2286 14.5 0.544 1.059 0.544 1.059 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.045 1.047 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.013 1.035 dbcsr_complete_redistribute 329 12.2 0.097 0.137 0.873 1.003 mp_waitany 9880 13.7 0.726 0.986 0.726 0.986 rs_pw_transfer_RS2PW_140 130 11.5 0.124 0.129 0.700 0.938 yz_to_x 487 15.3 0.156 0.164 0.884 0.937 multiply_cannon_metrocomm1 18288 15.5 0.029 0.030 0.672 0.932 acc_transpose_blocks_kernels 18288 16.5 0.212 0.221 0.859 0.879 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=42.203000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.181818, yerr=1.266217 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ea454b3df0bf366fae64a409da442cd2660d5a8e_performance_tests/06/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 114.044384E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3805952 0.0% 0.0% 100.0% average stack size 0.0 0.0 38.6 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 557.191168E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63496. MP_Allreduce 10074 349. MP_Sync 54 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.020 0.137 45.888 45.968 qs_mol_dyn_low 1 2.0 0.003 0.003 45.640 45.645 qs_forces 11 3.9 0.003 0.004 45.558 45.561 qs_energies 11 4.9 0.001 0.001 43.581 43.591 scf_env_do_scf 11 5.9 0.000 0.001 37.756 37.758 scf_env_do_scf_inner_loop 108 6.5 0.003 0.008 32.807 32.807 velocity_verlet 10 3.0 0.001 0.001 23.712 23.713 dbcsr_multiply_generic 2286 12.5 0.114 0.121 20.980 21.107 qs_scf_new_mos 108 7.5 0.001 0.001 20.593 20.652 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.592 20.651 ot_scf_mini 108 9.5 0.002 0.003 19.512 19.562 multiply_cannon 2286 13.5 0.233 0.280 15.250 15.892 multiply_cannon_loop 2286 14.5 0.939 0.968 13.870 14.236 ot_mini 108 10.5 0.001 0.001 11.299 11.363 qs_ot_get_derivative 108 11.5 0.001 0.001 9.329 9.380 multiply_cannon_multrec 27432 15.5 2.369 3.069 8.665 9.060 rebuild_ks_matrix 119 8.3 0.000 0.000 8.805 8.859 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 8.804 8.858 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.866 7.914 dbcsr_mm_accdrv_process 47916 15.9 5.336 7.043 6.201 7.468 sum_up_and_integrate 119 10.3 0.035 0.037 5.674 5.680 qs_rho_update_rho_low 119 7.7 0.001 0.001 5.616 5.658 calculate_rho_elec 119 8.7 0.040 0.046 5.616 5.658 integrate_v_rspace 119 11.3 0.003 0.003 5.639 5.645 qs_ot_get_p 119 10.4 0.001 0.001 5.227 5.314 mp_waitall_1 137007 16.6 4.331 5.304 4.331 5.304 rs_pw_transfer 974 11.9 0.009 0.009 4.885 5.095 init_scf_loop 11 6.9 0.000 0.000 4.912 4.912 init_scf_run 11 5.9 0.000 0.001 4.325 4.325 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.325 4.325 density_rs2pw 119 9.7 0.004 0.004 3.933 4.123 make_m2s 4572 13.5 0.055 0.057 3.853 4.096 make_images 4572 14.5 0.269 0.330 3.742 3.991 prepare_preconditioner 11 7.9 0.000 0.000 3.799 3.808 make_preconditioner 11 8.9 0.000 0.000 3.799 3.808 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.372 3.771 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.342 3.698 qs_ot_p2m_diag 50 11.0 0.015 0.023 3.319 3.337 potential_pw2rs 119 12.3 0.009 0.010 3.281 3.302 pw_transfer 1439 11.6 0.066 0.071 3.001 3.057 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.907 2.967 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.897 2.898 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.655 2.684 fft3d_ps 1201 14.6 0.572 0.618 2.584 2.639 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.408 2.417 cp_fm_diag_elpa_base 50 14.0 2.360 2.381 2.405 2.414 calculate_first_density_matrix 1 7.0 0.000 0.000 2.354 2.356 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.171 2.304 apply_single 119 13.6 0.000 0.000 2.171 2.304 calculate_dm_sparse 119 9.5 0.000 0.001 2.193 2.252 fft_wrap_pw1pw2_140 487 13.2 0.085 0.096 2.118 2.181 make_images_data 4572 15.5 0.044 0.048 1.707 2.139 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.074 2.076 ot_diis_step 108 11.5 0.012 0.012 1.921 1.921 wfi_extrapolate 11 7.9 0.001 0.001 1.908 1.909 grid_integrate_task_list 119 12.3 1.814 1.903 1.814 1.903 jit_kernel_multiply 10 15.8 0.802 1.835 0.802 1.835 mp_sum_l 7207 12.9 1.463 1.825 1.463 1.825 hybrid_alltoall_any 4725 16.4 0.062 0.152 1.396 1.747 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.722 1.740 multiply_cannon_metrocomm3 27432 15.5 0.039 0.040 0.877 1.670 mp_sendrecv_dv 8211 12.7 1.634 1.663 1.634 1.663 mp_alltoall_z22v 1201 16.6 1.598 1.651 1.598 1.651 make_images_sizes 4572 15.5 0.005 0.005 1.120 1.605 mp_alltoall_i44 4572 16.5 1.115 1.600 1.115 1.600 dbcsr_complete_redistribute 329 12.2 0.128 0.163 1.226 1.521 acc_transpose_blocks 27432 15.5 0.114 0.117 1.481 1.502 mp_alltoall_d11v 2130 13.8 1.305 1.495 1.305 1.495 cp_fm_cholesky_invert 11 10.9 1.380 1.385 1.380 1.385 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.330 1.351 grid_collocate_task_list 119 9.7 1.222 1.333 1.222 1.333 cp_fm_upper_to_full 72 14.2 0.892 1.281 0.892 1.281 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.265 1.276 rs_pw_transfer_RS2PW_50 119 11.7 0.228 0.233 1.053 1.144 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.126 1.129 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.834 1.126 rs_pw_transfer_PW2RS_50 119 14.3 0.311 0.325 1.014 1.110 yz_to_x 606 15.1 0.188 0.222 1.027 1.052 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.037 1.042 multiply_cannon_sync_h2d 27432 15.5 0.935 0.992 0.935 0.992 x_to_yz 595 16.2 0.214 0.246 0.974 0.981 mp_allgather_i34 2286 14.5 0.670 0.965 0.670 0.965 cp_fm_cholesky_decompose 22 10.9 0.933 0.939 0.933 0.939 multiply_cannon_metrocomm1 27432 15.5 0.034 0.035 0.449 0.937 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=45.968000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=528.272727, yerr=3.670296 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ea454b3df0bf366fae64a409da442cd2660d5a8e_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 600.248320E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63495. MP_Allreduce 10074 348. MP_Sync 54 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.047 30.065 30.076 qs_mol_dyn_low 1 2.0 0.003 0.003 29.831 29.836 qs_forces 11 3.9 0.002 0.003 29.740 29.741 qs_energies 11 4.9 0.001 0.001 27.966 27.969 scf_env_do_scf 11 5.9 0.000 0.001 22.917 22.918 scf_env_do_scf_inner_loop 108 6.5 0.003 0.008 20.200 20.201 velocity_verlet 10 3.0 0.001 0.001 15.526 15.530 dbcsr_multiply_generic 2286 12.5 0.091 0.096 12.645 12.727 qs_scf_new_mos 108 7.5 0.001 0.001 11.590 11.620 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.590 11.620 ot_scf_mini 108 9.5 0.002 0.002 10.893 10.922 multiply_cannon 2286 13.5 0.231 0.241 9.800 10.347 multiply_cannon_loop 2286 14.5 0.329 0.339 8.848 9.072 rebuild_ks_matrix 119 8.3 0.000 0.000 6.367 6.393 qs_ks_build_kohn_sham_matrix 119 9.3 0.014 0.015 6.367 6.393 multiply_cannon_multrec 9144 15.5 1.580 1.806 5.807 6.045 ot_mini 108 10.5 0.001 0.001 5.976 6.011 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.659 5.682 qs_ot_get_derivative 108 11.5 0.001 0.001 4.654 4.683 dbcsr_mm_accdrv_process 12550 15.8 3.177 4.185 4.125 4.225 sum_up_and_integrate 119 10.3 0.038 0.042 3.965 3.970 integrate_v_rspace 119 11.3 0.003 0.003 3.927 3.932 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.912 3.924 calculate_rho_elec 119 8.7 0.060 0.061 3.912 3.923 init_scf_run 11 5.9 0.000 0.001 3.569 3.569 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.569 3.569 qs_ot_get_p 119 10.4 0.001 0.001 3.081 3.121 init_scf_loop 11 6.9 0.000 0.000 2.684 2.685 mp_waitall_1 115863 16.7 1.861 2.464 1.861 2.464 pw_transfer 1439 11.6 0.066 0.071 2.391 2.401 density_rs2pw 119 9.7 0.004 0.004 2.200 2.342 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.296 2.308 calculate_first_density_matrix 1 7.0 0.000 0.000 2.272 2.272 rs_pw_transfer 974 11.9 0.008 0.008 2.012 2.145 make_m2s 4572 13.5 0.035 0.036 1.963 2.132 make_images 4572 14.5 0.267 0.302 1.873 2.042 qs_ot_p2m_diag 50 11.0 0.022 0.023 2.021 2.024 fft3d_ps 1201 14.6 0.569 0.582 1.956 1.966 grid_integrate_task_list 119 12.3 1.849 1.917 1.849 1.917 prepare_preconditioner 11 7.9 0.000 0.000 1.887 1.891 make_preconditioner 11 8.9 0.000 0.000 1.887 1.891 fft_wrap_pw1pw2_140 487 13.2 0.087 0.090 1.831 1.848 jit_kernel_multiply 10 15.9 0.910 1.842 0.910 1.842 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.818 1.819 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.764 1.796 calculate_dm_sparse 119 9.5 0.000 0.000 1.739 1.758 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.756 1.757 potential_pw2rs 119 12.3 0.010 0.011 1.663 1.672 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.611 1.628 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.523 1.532 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.496 1.504 cp_fm_diag_elpa_base 50 14.0 1.467 1.487 1.494 1.502 grid_collocate_task_list 119 9.7 1.273 1.391 1.273 1.391 ot_diis_step 108 11.5 0.013 0.013 1.305 1.306 wfi_extrapolate 11 7.9 0.001 0.001 1.245 1.245 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.219 1.233 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.228 1.229 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.172 1.197 apply_single 119 13.6 0.000 0.000 1.171 1.196 hybrid_alltoall_any 4725 16.4 0.062 0.174 0.878 1.180 make_images_data 4572 15.5 0.038 0.042 0.923 1.177 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.150 1.158 mp_alltoall_d11v 2130 13.8 0.907 1.006 0.907 1.006 cp_fm_cholesky_invert 11 10.9 0.963 0.966 0.963 0.966 acc_transpose_blocks 9144 15.5 0.038 0.040 0.933 0.953 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.867 0.913 mp_alltoall_z22v 1201 16.6 0.835 0.886 0.835 0.886 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.832 0.835 multiply_cannon_sync_h2d 9144 15.5 0.703 0.777 0.703 0.777 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.762 0.771 mp_allgather_i34 2286 14.5 0.282 0.737 0.282 0.737 multiply_cannon_metrocomm1 9144 15.5 0.022 0.023 0.441 0.733 qs_env_update_s_mstruct 11 6.9 0.001 0.003 0.684 0.732 yz_to_x 606 15.1 0.264 0.281 0.708 0.723 acc_transpose_blocks_kernels 9144 16.5 0.118 0.120 0.688 0.705 x_to_yz 595 16.2 0.276 0.288 0.667 0.676 dbcsr_complete_redistribute 329 12.2 0.180 0.216 0.647 0.675 mp_sum_l 7207 12.9 0.451 0.623 0.451 0.623 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=30.076000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=570.545455, yerr=4.185769 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ea454b3df0bf366fae64a409da442cd2660d5a8e_performance_tests/08/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 235.585836E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 1388964 0.0% 0.0% 100.0% average stack size 0.0 0.0 105.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 754.204672E+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 0.019 0.042 51.415 51.416 qs_mol_dyn_low 1 2.0 0.003 0.003 51.114 51.121 qs_forces 11 3.9 0.002 0.003 51.050 51.052 qs_energies 11 4.9 0.001 0.001 48.804 48.807 scf_env_do_scf 11 5.9 0.001 0.001 42.265 42.265 scf_env_do_scf_inner_loop 108 6.5 0.003 0.008 32.556 32.558 velocity_verlet 10 3.0 0.001 0.001 28.867 28.872 dbcsr_multiply_generic 2286 12.5 0.101 0.104 20.296 20.522 qs_scf_new_mos 108 7.5 0.001 0.001 19.979 20.089 qs_scf_loop_do_ot 108 8.5 0.001 0.001 19.978 20.088 ot_scf_mini 108 9.5 0.002 0.002 18.783 18.888 multiply_cannon 2286 13.5 0.311 0.318 14.504 15.719 multiply_cannon_loop 2286 14.5 0.344 0.349 13.011 13.862 ot_mini 108 10.5 0.001 0.001 10.655 10.786 init_scf_loop 11 6.9 0.000 0.000 9.663 9.665 rebuild_ks_matrix 119 8.3 0.000 0.000 9.115 9.263 qs_ks_build_kohn_sham_matrix 119 9.3 0.024 0.024 9.115 9.263 multiply_cannon_multrec 9144 15.5 3.575 5.143 8.797 8.964 qs_ot_get_derivative 108 11.5 0.001 0.001 8.433 8.539 prepare_preconditioner 11 7.9 0.000 0.000 8.470 8.483 make_preconditioner 11 8.9 0.000 0.000 8.470 8.483 qs_ks_update_qs_env 119 7.6 0.001 0.001 8.204 8.338 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.877 8.325 dbcsr_mm_accdrv_process 12550 15.8 4.131 5.724 5.098 6.555 qs_rho_update_rho_low 119 7.7 0.001 0.001 5.752 5.796 calculate_rho_elec 119 8.7 0.118 0.122 5.751 5.796 mp_waitall_1 94719 16.7 4.340 5.490 4.340 5.490 sum_up_and_integrate 119 10.3 0.066 0.067 5.346 5.352 integrate_v_rspace 119 11.3 0.003 0.003 5.280 5.288 qs_ot_get_p 119 10.4 0.001 0.001 5.016 5.166 cp_fm_upper_to_full 72 14.2 3.331 4.850 3.331 4.850 init_scf_run 11 5.9 0.000 0.001 4.332 4.332 scf_env_initial_rho_setup 11 6.9 0.001 0.002 4.331 4.332 make_m2s 4572 13.5 0.039 0.039 3.858 4.163 make_images 4572 14.5 0.352 0.380 3.737 4.042 pw_transfer 1439 11.6 0.069 0.069 3.898 3.903 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 3.799 3.804 density_rs2pw 119 9.7 0.004 0.004 3.427 3.441 fft3d_ps 1201 14.6 0.605 0.614 3.423 3.425 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.968 3.404 dbcsr_complete_redistribute 329 12.2 0.277 0.283 2.412 3.308 qs_ot_p2m_diag 50 11.0 0.043 0.044 3.194 3.196 fft_wrap_pw1pw2_140 487 13.2 0.095 0.097 3.016 3.024 rs_pw_transfer 974 11.9 0.009 0.010 2.826 2.849 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.839 2.840 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.933 2.816 make_images_data 4572 15.5 0.043 0.045 2.107 2.734 hybrid_alltoall_any 4725 16.4 0.087 0.148 1.965 2.638 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.364 2.578 apply_single 119 13.6 0.000 0.000 2.364 2.578 mp_alltoall_i22 627 13.8 1.578 2.508 1.578 2.508 calculate_dm_sparse 119 9.5 0.000 0.000 2.442 2.488 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.588 2.463 potential_pw2rs 119 12.3 0.014 0.014 2.449 2.458 calculate_first_density_matrix 1 7.0 0.000 0.000 2.364 2.407 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 1.524 2.381 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.368 2.369 cp_fm_diag_elpa_base 50 14.0 2.191 2.263 2.364 2.365 cp_fm_cholesky_invert 11 10.9 2.320 2.323 2.320 2.323 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.193 2.250 ot_diis_step 108 11.5 0.015 0.015 2.179 2.180 grid_integrate_task_list 119 12.3 2.159 2.177 2.159 2.177 mp_sum_l 7207 12.9 1.403 2.123 1.403 2.123 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.073 2.075 mp_alltoall_z22v 1201 16.6 1.868 1.908 1.868 1.908 wfi_extrapolate 11 7.9 0.001 0.001 1.844 1.844 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.819 1.820 mp_alltoall_d11v 2130 13.8 1.735 1.804 1.735 1.804 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.716 1.767 grid_collocate_task_list 119 9.7 1.559 1.575 1.559 1.575 jit_kernel_multiply 8 15.4 0.940 1.570 0.940 1.570 yz_to_x 606 15.1 0.457 0.469 1.479 1.490 make_images_sizes 4572 15.5 0.005 0.005 0.805 1.465 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.423 1.463 mp_alltoall_i44 4572 16.5 0.800 1.460 0.800 1.460 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.341 1.368 x_to_yz 595 16.2 0.479 0.492 1.326 1.328 qs_env_update_s_mstruct 11 6.9 0.003 0.004 1.153 1.170 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.129 1.141 cp_fm_cholesky_decompose 22 10.9 1.084 1.100 1.084 1.100 multiply_cannon_sync_h2d 9144 15.5 1.043 1.047 1.043 1.047 mp_allgather_i34 2286 14.5 0.464 1.030 0.464 1.030 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=51.416000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=708.090909, yerr=16.076265 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ea454b3df0bf366fae64a409da442cd2660d5a8e_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 503.083008E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66218. MP_Allreduce 9696 492. MP_Sync 52 MP_Alltoall 1938 1379060. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.031 103.846 103.847 qs_mol_dyn_low 1 2.0 0.003 0.003 103.458 103.508 qs_forces 11 3.9 0.003 0.004 103.369 103.370 qs_energies 11 4.9 0.001 0.001 99.968 99.982 scf_env_do_scf 11 5.9 0.000 0.001 89.465 89.467 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 80.392 80.394 qs_scf_new_mos 99 7.5 0.000 0.001 57.202 57.333 qs_scf_loop_do_ot 99 8.5 0.001 0.001 57.202 57.332 dbcsr_multiply_generic 2055 12.4 0.116 0.126 55.297 55.655 ot_scf_mini 99 9.5 0.002 0.002 54.703 54.806 velocity_verlet 10 3.0 0.001 0.001 52.503 52.506 multiply_cannon 2055 13.4 0.187 0.192 43.411 44.473 multiply_cannon_loop 2055 14.4 1.505 1.536 41.740 42.904 ot_mini 99 10.5 0.001 0.001 29.934 30.010 qs_ot_get_derivative 99 11.5 0.001 0.001 23.069 23.164 rebuild_ks_matrix 110 8.3 0.000 0.000 18.086 18.313 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 18.085 18.313 multiply_cannon_multrec 49320 15.4 12.128 12.817 16.991 17.790 qs_ks_update_qs_env 110 7.6 0.001 0.001 15.885 16.091 qs_ot_get_p 110 10.4 0.001 0.001 15.798 15.929 mp_waitall_1 220248 16.4 14.070 15.634 14.070 15.634 qs_ot_p2m_diag 48 11.0 0.012 0.019 11.386 11.440 multiply_cannon_sync_h2d 49320 15.4 9.902 10.647 9.902 10.647 cp_dbcsr_syevd 48 12.0 0.002 0.003 10.202 10.202 qs_rho_update_rho_low 110 7.6 0.001 0.001 10.067 10.188 calculate_rho_elec 110 8.6 0.021 0.026 10.067 10.188 sum_up_and_integrate 110 10.3 0.036 0.042 9.778 9.793 integrate_v_rspace 110 11.3 0.003 0.004 9.742 9.766 cp_fm_diag_elpa 48 13.0 0.000 0.000 9.071 9.094 cp_fm_diag_elpa_base 48 14.0 9.033 9.066 9.067 9.090 init_scf_loop 11 6.9 0.000 0.000 9.008 9.010 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.920 8.438 init_scf_run 11 5.9 0.000 0.001 8.025 8.025 scf_env_initial_rho_setup 11 6.9 0.001 0.002 8.025 8.025 multiply_cannon_metrocomm3 49320 15.4 0.077 0.082 6.690 7.953 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.270 7.822 apply_single 110 13.6 0.000 0.000 7.269 7.822 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 7.544 7.591 rs_pw_transfer 902 11.9 0.012 0.013 6.845 7.512 density_rs2pw 110 9.6 0.004 0.005 6.754 7.401 pw_transfer 1331 11.6 0.056 0.066 6.556 6.740 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 6.467 6.654 ot_diis_step 99 11.5 0.006 0.006 6.644 6.645 mp_sum_l 6514 12.8 5.215 6.426 5.215 6.426 prepare_preconditioner 11 7.9 0.000 0.000 6.364 6.380 make_preconditioner 11 8.9 0.000 0.000 6.364 6.380 make_m2s 4110 13.4 0.062 0.065 6.210 6.342 make_images 4110 14.4 0.179 0.192 6.112 6.246 fft3d_ps 1111 14.6 0.779 0.882 6.005 6.179 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.074 6.144 fft_wrap_pw1pw2_140 451 13.1 0.163 0.185 5.062 5.250 potential_pw2rs 110 12.3 0.006 0.007 5.116 5.218 wfi_extrapolate 11 7.9 0.001 0.001 5.072 5.073 dbcsr_mm_accdrv_process 87628 16.1 1.930 2.050 4.741 5.064 multiply_cannon_metrocomm1 49320 15.4 0.063 0.066 3.357 4.860 mp_alltoall_z22v 1111 16.6 4.192 4.557 4.192 4.557 calculate_dm_sparse 110 9.5 0.000 0.001 3.901 3.993 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.907 3.912 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.594 3.644 cp_fm_cholesky_invert 11 10.9 3.566 3.573 3.566 3.573 grid_integrate_task_list 110 12.3 3.214 3.473 3.214 3.473 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.385 3.446 mp_alltoall_d11v 2046 13.8 2.917 3.326 2.917 3.326 make_images_data 4110 15.4 0.042 0.045 2.618 2.931 calculate_first_density_matrix 1 7.0 0.000 0.000 2.825 2.828 mp_waitany 14300 13.8 2.188 2.824 2.188 2.824 make_images_sizes 4110 15.4 0.004 0.005 2.102 2.806 mp_alltoall_i44 4110 16.4 2.098 2.802 2.098 2.802 yz_to_x 341 15.4 0.210 0.237 2.325 2.573 hybrid_alltoall_any 4261 16.3 0.081 0.463 2.199 2.563 jit_kernel_multiply 13 15.9 2.531 2.546 2.531 2.546 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.349 2.370 x_to_yz 330 15.4 0.260 0.312 2.329 2.344 grid_collocate_task_list 110 9.6 2.081 2.326 2.081 2.326 acc_transpose_blocks 49320 15.4 0.213 0.222 2.184 2.269 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=103.847000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=477.454545, yerr=2.571297 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ea454b3df0bf366fae64a409da442cd2660d5a8e_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 587.866112E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1972800 MPI messages size (bytes): total size 1.077520E+12 min size 0.000000E+00 max size 4.537280E+06 average size 546.188250E+03 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 222984 1826684928 8192 < size <= 32768 520356 13399818240 32768 < size <= 131072 372336 35386294272 131072 < size <= 4194304 787758 788321309808 4194304 < size <= 16777216 54450 238588003280 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66437. MP_Allreduce 9695 570. MP_Sync 52 MP_Alltoall 1717 3475951. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.085 89.632 89.644 qs_mol_dyn_low 1 2.0 0.003 0.003 89.245 89.254 qs_forces 11 3.9 0.003 0.003 89.153 89.155 qs_energies 11 4.9 0.001 0.001 85.525 85.537 scf_env_do_scf 11 5.9 0.000 0.001 75.791 75.795 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 64.042 64.043 velocity_verlet 10 3.0 0.001 0.001 46.726 46.728 dbcsr_multiply_generic 2055 12.4 0.120 0.124 43.232 44.327 qs_scf_new_mos 99 7.5 0.001 0.001 43.683 43.832 qs_scf_loop_do_ot 99 8.5 0.001 0.001 43.682 43.831 ot_scf_mini 99 9.5 0.003 0.003 41.713 41.866 multiply_cannon 2055 13.4 0.223 0.246 32.911 36.610 multiply_cannon_loop 2055 14.4 0.926 0.957 30.923 34.432 ot_mini 99 10.5 0.001 0.001 22.895 23.013 mp_waitall_1 176588 16.5 11.095 17.402 11.095 17.402 qs_ot_get_derivative 99 11.5 0.001 0.001 16.573 16.746 rebuild_ks_matrix 110 8.3 0.000 0.000 16.026 16.169 qs_ks_build_kohn_sham_matrix 110 9.3 0.021 0.025 16.026 16.169 multiply_cannon_multrec 24660 15.4 7.292 9.506 13.554 15.697 qs_ks_update_qs_env 110 7.6 0.001 0.001 14.161 14.287 multiply_cannon_metrocomm3 24660 15.4 0.068 0.070 6.620 13.380 init_scf_loop 11 6.9 0.000 0.000 11.687 11.690 qs_ot_get_p 110 10.4 0.001 0.001 11.399 11.514 prepare_preconditioner 11 7.9 0.000 0.000 9.379 9.398 make_preconditioner 11 8.9 0.000 0.000 9.379 9.398 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.857 9.066 sum_up_and_integrate 110 10.3 0.053 0.059 8.553 8.568 integrate_v_rspace 110 11.3 0.003 0.003 8.500 8.517 qs_rho_update_rho_low 110 7.6 0.001 0.001 8.483 8.501 calculate_rho_elec 110 8.6 0.040 0.048 8.482 8.501 qs_ot_p2m_diag 48 11.0 0.029 0.044 8.409 8.449 multiply_cannon_sync_h2d 24660 15.4 6.751 7.946 6.751 7.946 cp_dbcsr_syevd 48 12.0 0.003 0.003 7.759 7.760 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.698 7.475 apply_single 110 13.6 0.000 0.001 6.698 7.475 init_scf_run 11 5.9 0.000 0.001 7.132 7.133 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.132 7.132 cp_fm_diag_elpa 48 13.0 0.000 0.000 6.708 6.720 cp_fm_diag_elpa_base 48 14.0 6.638 6.667 6.704 6.717 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 5.984 6.678 make_m2s 4110 13.4 0.057 0.060 6.037 6.552 make_images 4110 14.4 0.396 0.439 5.927 6.441 dbcsr_mm_accdrv_process 52282 16.1 4.756 5.700 6.111 6.397 rs_pw_transfer 902 11.9 0.012 0.013 5.511 6.291 density_rs2pw 110 9.6 0.004 0.005 5.496 6.151 ot_diis_step 99 11.5 0.010 0.010 6.087 6.088 pw_transfer 1331 11.6 0.066 0.077 5.259 5.484 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 5.152 5.380 cp_fm_cholesky_invert 11 10.9 4.886 4.904 4.886 4.904 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 4.683 4.775 mp_sum_l 6514 12.8 3.697 4.733 3.697 4.733 fft3d_ps 1111 14.6 1.110 1.335 4.511 4.727 wfi_extrapolate 11 7.9 0.001 0.001 4.340 4.341 potential_pw2rs 110 12.3 0.008 0.008 4.210 4.271 fft_wrap_pw1pw2_140 451 13.1 0.194 0.217 3.592 3.810 make_images_data 4110 15.4 0.046 0.050 3.165 3.798 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.571 3.575 hybrid_alltoall_any 4261 16.3 0.102 0.442 2.697 3.398 calculate_dm_sparse 110 9.5 0.001 0.001 3.305 3.344 grid_integrate_task_list 110 12.3 3.116 3.326 3.116 3.326 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.083 3.108 mp_alltoall_d11v 2046 13.8 2.500 3.028 2.500 3.028 mp_alltoall_z22v 1111 16.6 2.581 2.774 2.581 2.774 qs_ot_get_orbitals 99 10.5 0.001 0.001 2.585 2.693 cp_fm_cholesky_decompose 22 10.9 2.682 2.689 2.682 2.689 calculate_first_density_matrix 1 7.0 0.000 0.000 2.672 2.674 mp_allgather_i34 2055 14.4 1.315 2.560 1.315 2.560 grid_collocate_task_list 110 9.6 2.053 2.477 2.053 2.477 multiply_cannon_metrocomm4 22605 15.4 0.077 0.081 0.832 2.254 mp_waitany 10164 13.8 1.577 2.220 1.577 2.220 make_images_sizes 4110 15.4 0.005 0.005 1.522 2.100 mp_alltoall_i44 4110 16.4 1.517 2.095 1.517 2.095 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.039 2.057 dbcsr_complete_redistribute 325 12.2 0.233 0.282 1.722 2.039 mp_irecv_dv 57340 16.2 0.708 2.032 0.708 2.032 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.961 1.963 mp_sum_d 3889 11.9 1.480 1.905 1.480 1.905 yz_to_x 451 15.2 0.268 0.311 1.701 1.832 jit_kernel_multiply 9 16.3 1.002 1.808 1.002 1.808 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=89.644000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=555.090909, yerr=7.102555 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ea454b3df0bf366fae64a409da442cd2660d5a8e_performance_tests/11/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 404.681598E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 3346752 0.0% 0.0% 100.0% average stack size 0.0 0.0 294.1 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 661.561344E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66428. MP_Allreduce 9694 567. MP_Sync 52 MP_Alltoall 1496 4511006. MP_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.036 0.083 66.947 66.949 qs_mol_dyn_low 1 2.0 0.003 0.005 66.297 66.306 qs_forces 11 3.9 0.008 0.009 66.229 66.230 qs_energies 11 4.9 0.006 0.041 62.931 62.935 scf_env_do_scf 11 5.9 0.001 0.002 54.100 54.100 scf_env_do_scf_inner_loop 99 6.5 0.003 0.006 44.299 44.300 velocity_verlet 10 3.0 0.001 0.001 35.617 35.619 dbcsr_multiply_generic 2055 12.4 0.109 0.113 30.118 30.540 qs_scf_new_mos 99 7.5 0.001 0.001 27.998 28.103 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.997 28.103 ot_scf_mini 99 9.5 0.002 0.004 26.725 26.847 multiply_cannon 2055 13.4 0.212 0.227 22.915 24.486 multiply_cannon_loop 2055 14.4 0.617 0.630 21.566 22.746 ot_mini 99 10.5 0.001 0.001 14.885 15.005 rebuild_ks_matrix 110 8.3 0.000 0.000 13.246 13.424 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.016 13.246 13.424 mp_waitall_1 139946 16.5 8.334 11.943 8.334 11.943 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.672 11.839 multiply_cannon_multrec 16440 15.4 3.722 4.699 9.545 10.439 qs_ot_get_derivative 99 11.5 0.001 0.001 10.215 10.336 init_scf_loop 11 6.9 0.001 0.009 9.755 9.756 multiply_cannon_metrocomm3 16440 15.4 0.042 0.045 5.052 8.330 prepare_preconditioner 11 7.9 0.000 0.001 7.843 7.867 make_preconditioner 11 8.9 0.000 0.002 7.843 7.867 make_full_inverse_cholesky 11 9.9 0.000 0.001 7.145 7.514 sum_up_and_integrate 110 10.3 0.061 0.061 7.220 7.236 integrate_v_rspace 110 11.3 0.003 0.003 7.160 7.176 qs_ot_get_p 110 10.4 0.001 0.001 6.837 6.989 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.630 6.641 calculate_rho_elec 110 8.6 0.059 0.059 6.630 6.641 init_scf_run 11 5.9 0.000 0.001 6.031 6.031 scf_env_initial_rho_setup 11 6.9 0.002 0.004 6.031 6.031 dbcsr_mm_accdrv_process 34862 16.1 4.714 5.228 5.680 5.961 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.163 5.619 apply_single 110 13.6 0.000 0.000 5.163 5.618 make_m2s 4110 13.4 0.050 0.052 4.640 5.064 make_images 4110 14.4 0.390 0.509 4.525 4.948 density_rs2pw 110 9.6 0.004 0.005 3.660 4.926 qs_ot_p2m_diag 48 11.0 0.042 0.044 4.852 4.859 rs_pw_transfer 902 11.9 0.011 0.012 3.407 4.653 ot_diis_step 99 11.5 0.011 0.011 4.618 4.619 cp_dbcsr_syevd 48 12.0 0.003 0.005 4.483 4.484 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.566 4.226 pw_transfer 1331 11.6 0.066 0.073 4.197 4.216 multiply_cannon_sync_h2d 16440 15.4 3.588 4.166 3.588 4.166 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.090 4.110 cp_fm_diag_elpa 48 13.0 0.000 0.001 3.821 3.831 cp_fm_diag_elpa_base 48 14.0 3.745 3.781 3.817 3.827 grid_integrate_task_list 110 12.3 3.152 3.447 3.152 3.447 fft_wrap_pw1pw2_140 451 13.1 0.212 0.218 3.409 3.432 fft3d_ps 1111 14.6 1.095 1.107 3.390 3.411 cp_fm_cholesky_invert 11 10.9 3.397 3.406 3.397 3.406 wfi_extrapolate 11 7.9 0.001 0.001 3.244 3.244 make_images_data 4110 15.4 0.043 0.048 2.655 3.154 hybrid_alltoall_any 4261 16.3 0.105 0.373 2.343 2.986 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.921 2.924 potential_pw2rs 110 12.3 0.011 0.011 2.859 2.885 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.668 2.732 mp_sum_l 6514 12.8 1.976 2.715 1.976 2.715 calculate_first_density_matrix 1 7.0 0.006 0.046 2.676 2.678 mp_alltoall_d11v 2046 13.8 2.105 2.662 2.105 2.662 mp_waitany 17072 13.8 1.311 2.648 1.311 2.648 calculate_dm_sparse 110 9.5 0.001 0.001 2.615 2.643 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.517 2.573 multiply_cannon_metrocomm4 14385 15.4 0.046 0.049 0.910 2.554 grid_collocate_task_list 110 9.6 2.083 2.468 2.083 2.468 mp_irecv_dv 48980 15.7 0.839 2.425 0.839 2.425 rs_pw_transfer_RS2PW_140 121 11.5 0.182 0.197 0.958 2.220 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 2.084 2.085 dbcsr_complete_redistribute 325 12.2 0.317 0.358 1.603 2.069 cp_fm_cholesky_decompose 22 10.9 1.888 1.907 1.888 1.907 mp_allgather_i34 2055 14.4 0.673 1.830 0.673 1.830 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.784 1.799 cp_fm_upper_to_full 70 14.2 1.406 1.795 1.406 1.795 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.536 1.555 mp_alltoall_z22v 1111 16.6 1.525 1.552 1.525 1.552 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.056 1.519 rs_gather_matrices 110 12.3 0.232 0.259 1.062 1.505 jit_kernel_multiply 8 16.3 0.578 1.484 0.578 1.484 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.351 1.438 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.384 1.399 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=66.949000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=625.454545, yerr=9.018805 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ea454b3df0bf366fae64a409da442cd2660d5a8e_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 735.993856E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66426. MP_Allreduce 9694 608. MP_Sync 52 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.028 66.815 66.816 qs_mol_dyn_low 1 2.0 0.003 0.003 66.555 66.563 qs_forces 11 3.9 0.003 0.003 66.233 66.234 qs_energies 11 4.9 0.001 0.001 62.724 62.728 scf_env_do_scf 11 5.9 0.000 0.001 54.328 54.331 scf_env_do_scf_inner_loop 99 6.5 0.003 0.006 42.491 42.492 velocity_verlet 10 3.0 0.001 0.001 37.912 37.924 dbcsr_multiply_generic 2055 12.4 0.114 0.120 29.919 30.165 qs_scf_new_mos 99 7.5 0.001 0.001 27.028 27.143 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.027 27.143 ot_scf_mini 99 9.5 0.003 0.003 25.391 25.503 multiply_cannon 2055 13.4 0.245 0.261 22.845 23.978 multiply_cannon_loop 2055 14.4 0.884 0.907 21.372 21.946 ot_mini 99 10.5 0.001 0.001 14.496 14.621 multiply_cannon_multrec 24660 15.4 4.192 6.840 12.660 13.887 rebuild_ks_matrix 110 8.3 0.000 0.000 12.525 12.625 qs_ks_build_kohn_sham_matrix 110 9.3 0.037 0.042 12.524 12.625 init_scf_loop 11 6.9 0.000 0.000 11.795 11.795 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.041 11.130 qs_ot_get_derivative 99 11.5 0.001 0.001 10.254 10.371 prepare_preconditioner 11 7.9 0.000 0.000 10.023 10.039 make_preconditioner 11 8.9 0.000 0.000 10.023 10.039 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.238 9.713 dbcsr_mm_accdrv_process 52304 16.0 6.991 8.215 8.319 9.259 sum_up_and_integrate 110 10.3 0.067 0.071 6.728 6.743 integrate_v_rspace 110 11.3 0.003 0.003 6.661 6.675 mp_waitall_1 121746 16.5 4.672 6.607 4.672 6.607 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.299 6.308 calculate_rho_elec 110 8.6 0.078 0.081 6.298 6.308 qs_ot_get_p 110 10.4 0.001 0.001 5.863 6.016 make_m2s 4110 13.4 0.061 0.063 5.467 5.779 init_scf_run 11 5.9 0.000 0.001 5.667 5.668 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.667 5.668 make_images 4110 14.4 0.573 0.694 5.324 5.633 cp_fm_upper_to_full 70 14.2 3.365 4.882 3.365 4.882 ot_diis_step 99 11.5 0.011 0.011 4.197 4.197 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.096 4.193 apply_single 110 13.6 0.000 0.000 4.096 4.193 pw_transfer 1331 11.6 0.065 0.074 4.053 4.083 qs_ot_p2m_diag 48 11.0 0.055 0.064 4.049 4.064 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.947 3.981 dbcsr_complete_redistribute 325 12.2 0.426 0.483 2.694 3.828 density_rs2pw 110 9.6 0.004 0.004 3.230 3.738 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.561 3.562 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.386 3.439 multiply_cannon_metrocomm3 24660 15.4 0.036 0.037 1.615 3.436 grid_integrate_task_list 110 12.3 3.293 3.423 3.293 3.423 fft_wrap_pw1pw2_140 451 13.1 0.203 0.212 3.361 3.396 multiply_cannon_sync_h2d 24660 15.4 3.186 3.346 3.186 3.346 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.171 3.297 fft3d_ps 1111 14.6 1.096 1.139 3.239 3.263 hybrid_alltoall_any 4261 16.3 0.120 0.455 2.370 3.117 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.055 3.116 make_images_data 4110 15.4 0.046 0.050 2.783 3.101 calculate_dm_sparse 110 9.5 0.001 0.001 3.036 3.068 rs_pw_transfer 902 11.9 0.010 0.011 2.508 3.030 wfi_extrapolate 11 7.9 0.001 0.001 3.011 3.011 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.973 2.982 cp_fm_diag_elpa_base 48 14.0 2.821 2.877 2.971 2.981 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.776 2.875 mp_alltoall_i22 605 13.7 1.662 2.849 1.662 2.849 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.836 2.837 cp_fm_cholesky_invert 11 10.9 2.792 2.801 2.792 2.801 calculate_first_density_matrix 1 7.0 0.000 0.000 2.562 2.563 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.451 2.484 grid_collocate_task_list 110 9.6 2.180 2.437 2.180 2.437 potential_pw2rs 110 12.3 0.012 0.013 2.337 2.349 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.241 2.243 mp_alltoall_d11v 2046 13.8 1.784 1.958 1.784 1.958 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.724 1.757 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.728 1.739 cp_fm_cholesky_decompose 22 10.9 1.685 1.727 1.685 1.727 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.597 1.703 jit_kernel_multiply 9 15.8 0.994 1.678 0.994 1.678 mp_allgather_i34 2055 14.4 0.571 1.655 0.571 1.655 acc_transpose_blocks 24660 15.4 0.110 0.111 1.584 1.655 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.639 1.654 mp_sum_l 6514 12.8 0.922 1.609 0.922 1.609 mp_waitany 13376 13.8 1.075 1.589 1.075 1.589 multiply_cannon_metrocomm4 20550 15.4 0.062 0.066 0.846 1.497 mp_irecv_dv 62702 16.1 0.742 1.419 0.742 1.419 mp_alltoall_z22v 1111 16.6 1.280 1.383 1.280 1.383 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=66.816000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=690.272727, yerr=8.697867 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ea454b3df0bf366fae64a409da442cd2660d5a8e_performance_tests/13/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 807.299199E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 1438408 0.0% 0.0% 100.0% average stack size 0.0 0.0 684.2 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 832.237568E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66424. MP_Allreduce 9694 649. MP_Sync 52 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.036 69.202 69.203 qs_mol_dyn_low 1 2.0 0.003 0.003 68.809 68.819 qs_forces 11 3.9 0.004 0.005 68.222 68.222 qs_energies 11 4.9 0.001 0.001 64.302 64.311 scf_env_do_scf 11 5.9 0.001 0.001 54.932 54.932 scf_env_do_scf_inner_loop 99 6.5 0.004 0.009 44.043 44.045 velocity_verlet 10 3.0 0.001 0.001 38.747 38.783 dbcsr_multiply_generic 2055 12.4 0.107 0.110 26.661 26.820 qs_scf_new_mos 99 7.5 0.001 0.001 26.437 26.494 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.436 26.494 ot_scf_mini 99 9.5 0.002 0.002 24.968 25.013 multiply_cannon 2055 13.4 0.248 0.259 19.043 20.853 multiply_cannon_loop 2055 14.4 0.322 0.335 17.420 17.847 rebuild_ks_matrix 110 8.3 0.000 0.000 13.514 13.540 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 13.514 13.540 ot_mini 99 10.5 0.001 0.001 12.998 13.027 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.019 12.042 init_scf_loop 11 6.9 0.000 0.000 10.821 10.824 mp_waitall_1 103326 16.6 8.481 10.211 8.481 10.211 qs_ot_get_derivative 99 11.5 0.001 0.001 9.087 9.122 prepare_preconditioner 11 7.9 0.000 0.000 8.953 8.959 make_preconditioner 11 8.9 0.000 0.000 8.953 8.959 multiply_cannon_multrec 8220 15.4 3.334 4.689 7.847 8.942 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.445 8.583 sum_up_and_integrate 110 10.3 0.080 0.081 7.720 7.731 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.690 7.717 calculate_rho_elec 110 8.6 0.116 0.117 7.690 7.716 integrate_v_rspace 110 11.3 0.003 0.003 7.640 7.652 qs_ot_get_p 110 10.4 0.001 0.001 7.567 7.604 make_m2s 4110 13.4 0.040 0.041 5.750 6.176 make_images 4110 14.4 0.634 0.689 5.617 6.043 init_scf_run 11 5.9 0.000 0.001 5.966 5.966 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.966 5.966 qs_ot_p2m_diag 48 11.0 0.081 0.084 5.531 5.539 dbcsr_mm_accdrv_process 17442 15.9 2.965 4.321 4.382 5.453 cp_dbcsr_syevd 48 12.0 0.003 0.003 5.096 5.098 pw_transfer 1331 11.6 0.066 0.072 5.047 5.060 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.939 4.955 density_rs2pw 110 9.6 0.004 0.004 4.215 4.586 cp_fm_cholesky_invert 11 10.9 4.580 4.585 4.580 4.585 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.307 4.314 cp_fm_diag_elpa_base 48 14.0 4.226 4.271 4.304 4.311 fft3d_ps 1111 14.6 1.155 1.177 4.184 4.203 fft_wrap_pw1pw2_140 451 13.1 0.213 0.219 4.073 4.094 make_images_data 4110 15.4 0.038 0.043 3.279 4.084 multiply_cannon_metrocomm3 8220 15.4 0.017 0.018 2.820 4.073 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.822 3.865 apply_single 110 13.6 0.000 0.000 3.822 3.864 ot_diis_step 99 11.5 0.012 0.012 3.853 3.854 hybrid_alltoall_any 4261 16.3 0.200 0.861 3.093 3.845 rs_pw_transfer 902 11.9 0.010 0.011 3.417 3.809 grid_integrate_task_list 110 12.3 3.341 3.539 3.341 3.539 wfi_extrapolate 11 7.9 0.001 0.001 3.417 3.417 multiply_cannon_sync_h2d 8220 15.4 2.879 3.052 2.879 3.052 potential_pw2rs 110 12.3 0.015 0.015 3.033 3.044 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.802 2.858 calculate_dm_sparse 110 9.5 0.001 0.001 2.800 2.840 mp_alltoall_d11v 2046 13.8 2.421 2.796 2.421 2.796 cp_fm_cholesky_decompose 22 10.9 2.762 2.776 2.762 2.776 qs_energies_init_hamiltonians 11 5.9 0.045 0.056 2.757 2.759 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.720 2.722 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.633 2.653 multiply_cannon_metrocomm1 8220 15.4 0.021 0.022 2.023 2.576 grid_collocate_task_list 110 9.6 2.283 2.541 2.283 2.541 calculate_first_density_matrix 1 7.0 0.000 0.000 2.428 2.430 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.254 2.268 mp_allgather_i34 2055 14.4 0.804 2.188 0.804 2.188 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.770 1.990 make_images_sizes 4110 15.4 0.005 0.005 1.056 1.986 mp_alltoall_i44 4110 16.4 1.052 1.981 1.052 1.981 dbcsr_complete_redistribute 325 12.2 0.542 0.570 1.836 1.979 mp_alltoall_z22v 1111 16.6 1.921 1.956 1.921 1.956 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.833 1.839 mp_waitany 9240 13.8 1.304 1.764 1.304 1.764 qs_env_update_s_mstruct 11 6.9 0.001 0.002 1.546 1.665 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.572 1.589 yz_to_x 561 15.1 0.523 0.534 1.512 1.525 x_to_yz 550 16.1 0.571 0.581 1.503 1.514 rs_gather_matrices 110 12.3 0.323 0.361 1.154 1.498 mp_sum_l 6514 12.8 1.174 1.487 1.174 1.487 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.413 1.432 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=69.203000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=779.909091, yerr=11.492721 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ea454b3df0bf366fae64a409da442cd2660d5a8e_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 1.315406E+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 0.022 0.040 102.806 102.808 qs_mol_dyn_low 1 2.0 0.003 0.003 102.244 102.254 qs_forces 11 3.9 0.016 0.022 102.172 102.173 qs_energies 11 4.9 0.001 0.001 97.655 97.658 scf_env_do_scf 11 5.9 0.001 0.001 86.509 86.510 velocity_verlet 10 3.0 0.001 0.001 64.467 64.473 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 53.390 53.391 init_scf_loop 11 6.9 0.000 0.000 33.027 33.030 qs_scf_new_mos 99 7.5 0.001 0.001 32.099 32.125 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.098 32.124 dbcsr_multiply_generic 2055 12.4 0.120 0.124 31.824 31.891 prepare_preconditioner 11 7.9 0.000 0.000 30.718 30.723 make_preconditioner 11 8.9 0.000 0.000 30.718 30.723 ot_scf_mini 99 9.5 0.002 0.002 30.200 30.221 make_full_inverse_cholesky 11 9.9 0.000 0.000 24.441 30.146 multiply_cannon 2055 13.4 0.358 0.384 22.521 23.783 multiply_cannon_loop 2055 14.4 0.342 0.344 20.347 20.779 cp_fm_upper_to_full 70 14.2 13.148 19.145 13.148 19.145 rebuild_ks_matrix 110 8.3 0.001 0.001 16.121 16.158 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.015 16.120 16.158 ot_mini 99 10.5 0.001 0.001 15.735 15.752 qs_ks_update_qs_env 110 7.6 0.001 0.001 14.507 14.543 dbcsr_complete_redistribute 325 12.2 1.014 1.042 8.194 11.695 mp_waitall_1 84994 16.7 9.880 11.023 9.880 11.023 qs_ot_get_derivative 99 11.5 0.001 0.001 10.876 10.892 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.969 10.461 multiply_cannon_multrec 8220 15.4 4.386 4.566 9.774 9.898 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.261 9.715 qs_rho_update_rho_low 110 7.6 0.001 0.001 9.391 9.423 calculate_rho_elec 110 8.6 0.227 0.228 9.390 9.423 mp_alltoall_i22 605 13.7 5.906 9.409 5.906 9.409 qs_ot_get_p 110 10.4 0.001 0.001 8.785 8.808 sum_up_and_integrate 110 10.3 0.150 0.152 8.774 8.793 integrate_v_rspace 110 11.3 0.003 0.003 8.624 8.642 make_m2s 4110 13.4 0.044 0.045 7.189 7.861 make_images 4110 14.4 0.881 0.936 6.997 7.669 cp_fm_cholesky_invert 11 10.9 7.614 7.620 7.614 7.620 pw_transfer 1331 11.6 0.075 0.076 6.704 6.715 init_scf_run 11 5.9 0.000 0.001 6.648 6.648 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.648 6.648 qs_ot_p2m_diag 48 11.0 0.151 0.156 6.637 6.645 fft_wrap_pw1pw2 1111 12.6 0.009 0.010 6.587 6.597 cp_dbcsr_syevd 48 12.0 0.003 0.003 6.110 6.111 fft3d_ps 1111 14.6 1.302 1.313 5.780 5.792 fft_wrap_pw1pw2_140 451 13.1 0.222 0.223 5.522 5.534 multiply_cannon_metrocomm3 8220 15.4 0.018 0.019 5.081 5.524 dbcsr_mm_accdrv_process 11614 15.7 3.383 3.688 5.244 5.484 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.946 5.389 apply_single 110 13.6 0.000 0.000 4.946 5.389 make_images_data 4110 15.4 0.041 0.044 4.205 5.371 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.281 5.281 cp_fm_diag_elpa_base 48 14.0 4.670 4.923 5.278 5.278 density_rs2pw 110 9.6 0.004 0.004 5.204 5.241 hybrid_alltoall_any 4261 16.3 0.257 0.555 3.957 5.150 ot_diis_step 99 11.5 0.016 0.016 4.815 4.816 wfi_extrapolate 11 7.9 0.001 0.001 3.948 3.948 multiply_cannon_sync_h2d 8220 15.4 3.943 3.947 3.943 3.947 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.412 3.857 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.779 3.781 grid_integrate_task_list 110 12.3 3.694 3.763 3.694 3.763 rs_pw_transfer 902 11.9 0.011 0.011 3.693 3.755 potential_pw2rs 110 12.3 0.021 0.021 3.630 3.637 calculate_dm_sparse 110 9.5 0.001 0.001 3.375 3.401 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.221 3.223 cp_fm_cholesky_decompose 22 10.9 3.179 3.202 3.179 3.202 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.056 3.071 mp_alltoall_d11v 2046 13.8 2.750 2.843 2.750 2.843 grid_collocate_task_list 110 9.6 2.656 2.679 2.656 2.679 mp_alltoall_z22v 1111 16.6 2.596 2.612 2.596 2.612 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.583 2.603 calculate_first_density_matrix 1 7.0 0.000 0.000 2.561 2.562 qs_env_update_s_mstruct 11 6.9 0.001 0.001 2.300 2.368 yz_to_x 561 15.1 0.935 0.939 2.321 2.329 mp_allgather_i34 2055 14.4 0.943 2.310 0.943 2.310 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.223 2.225 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.100 2.182 x_to_yz 550 16.1 0.933 0.939 2.143 2.151 copy_dbcsr_to_fm 151 11.3 0.003 0.003 2.051 2.090 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=102.808000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1192.545455, yerr=46.560544 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ea454b3df0bf366fae64a409da442cd2660d5a8e_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 628.187136E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175955383376 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 57767. MP_Allreduce 11004 802. MP_Sync 87 MP_Alltoall 2226 2523677. MP_SendRecv 24320 18752. MP_ISendRecv 24320 18752. MP_Wait 42476 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.024 0.059 213.005 213.006 qs_mol_dyn_low 1 2.0 0.005 0.021 212.517 212.529 qs_forces 11 3.9 0.005 0.005 212.351 212.353 qs_energies 11 4.9 0.011 0.084 206.607 206.628 scf_env_do_scf 11 5.9 0.001 0.002 189.575 189.579 scf_env_do_scf_inner_loop 117 6.6 0.005 0.017 168.178 168.180 velocity_verlet 10 3.0 0.006 0.043 127.217 127.226 qs_scf_new_mos 117 7.6 0.001 0.001 126.524 126.890 qs_scf_loop_do_ot 117 8.6 0.001 0.001 126.524 126.889 dbcsr_multiply_generic 2507 12.6 0.179 0.183 125.648 126.464 ot_scf_mini 117 9.6 0.003 0.004 119.910 120.280 multiply_cannon 2507 13.6 0.240 0.249 101.486 103.940 multiply_cannon_loop 2507 14.6 2.086 2.135 99.259 101.534 ot_mini 117 10.6 0.001 0.001 66.606 66.959 multiply_cannon_multrec 60168 15.6 33.360 35.380 41.583 43.659 qs_ot_get_derivative 117 11.6 0.001 0.001 41.629 41.987 rebuild_ks_matrix 128 8.3 0.001 0.001 34.702 35.127 qs_ks_build_kohn_sham_matrix 128 9.3 0.021 0.033 34.701 35.126 mp_waitall_1 267128 16.5 29.321 32.903 29.321 32.903 qs_ot_get_p 128 10.4 0.001 0.001 31.256 31.633 qs_ks_update_qs_env 128 7.6 0.001 0.001 31.172 31.592 multiply_cannon_sync_h2d 60168 15.6 27.456 29.545 27.456 29.545 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.507 25.410 apply_single 128 13.6 0.001 0.001 24.507 25.410 ot_diis_step 117 11.6 0.007 0.008 24.722 24.723 qs_ot_p2m_diag 83 11.4 0.079 0.091 24.417 24.480 cp_dbcsr_syevd 83 12.4 0.005 0.005 21.759 21.761 init_scf_loop 11 6.9 0.001 0.004 21.316 21.317 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 20.034 20.287 cp_fm_diag_elpa 83 13.4 0.000 0.000 18.583 18.613 cp_fm_diag_elpa_base 83 14.4 18.494 18.541 18.579 18.610 multiply_cannon_metrocomm3 60168 15.6 0.113 0.117 15.937 17.811 prepare_preconditioner 11 7.9 0.000 0.001 16.589 16.638 make_preconditioner 11 8.9 0.000 0.001 16.589 16.638 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.790 15.966 sum_up_and_integrate 128 10.3 0.089 0.108 14.962 14.977 integrate_v_rspace 128 11.3 0.004 0.005 14.872 14.890 make_m2s 5014 13.6 0.104 0.113 14.403 14.819 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.624 14.742 calculate_rho_elec 128 8.7 0.046 0.065 14.623 14.742 make_images 5014 14.6 0.401 0.421 14.221 14.646 init_scf_run 11 5.9 0.000 0.001 12.692 12.692 scf_env_initial_rho_setup 11 6.9 0.001 0.004 12.692 12.692 density_rs2pw 128 9.7 0.006 0.008 7.942 11.165 mp_sum_l 7870 13.0 8.494 10.200 8.494 10.200 rs_pw_transfer 1046 11.9 0.017 0.020 6.506 9.778 cp_fm_cholesky_invert 11 10.9 9.465 9.473 9.465 9.473 wfi_extrapolate 11 7.9 0.001 0.001 9.202 9.202 pw_transfer 1547 11.6 0.077 0.096 8.631 8.903 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 8.424 8.704 calculate_dm_sparse 128 9.5 0.001 0.001 8.410 8.554 dbcsr_mm_accdrv_process 124484 16.2 3.197 3.307 7.788 8.243 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.046 8.235 make_images_data 5014 15.6 0.065 0.071 7.065 8.019 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.863 7.962 multiply_cannon_metrocomm1 60168 15.6 0.091 0.095 6.223 7.754 grid_integrate_task_list 128 12.3 7.044 7.480 7.044 7.480 fft_wrap_pw1pw2_140 523 13.2 0.442 0.513 7.099 7.427 fft3d_ps 1291 14.7 2.105 2.881 7.211 7.416 hybrid_alltoall_any 5200 16.5 0.291 2.250 6.149 7.384 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.898 6.906 mp_alltoall_d11v 2415 14.1 4.531 6.215 4.531 6.215 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.991 6.077 mp_waitany 16020 13.9 2.656 5.862 2.656 5.862 grid_collocate_task_list 128 9.7 4.552 5.815 4.552 5.815 rs_pw_transfer_RS2PW_140 139 11.5 0.279 0.302 2.115 5.339 potential_pw2rs 128 12.3 0.009 0.011 5.244 5.284 cp_fm_cholesky_decompose 22 10.9 4.869 4.882 4.869 4.882 mp_sum_d 4470 12.1 3.573 4.280 3.573 4.280 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=213.006000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=595.272727, yerr=6.634495 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ea454b3df0bf366fae64a409da442cd2660d5a8e_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 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 11528891191296 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 93.514751E+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 6755938624 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 833.855488E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100942E+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 1463834332048 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58363. MP_Allreduce 10978 967. MP_Sync 87 MP_Alltoall 1969 5618844. MP_SendRecv 12032 47072. MP_ISendRecv 12032 47072. MP_Wait 25916 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 0.055 0.147 219.211 219.212 qs_mol_dyn_low 1 2.0 0.004 0.006 218.438 218.446 qs_forces 11 3.9 0.006 0.011 218.118 218.121 qs_energies 11 4.9 0.003 0.008 210.997 211.012 scf_env_do_scf 11 5.9 0.001 0.003 192.935 192.944 scf_env_do_scf_inner_loop 117 6.6 0.011 0.028 152.581 152.585 velocity_verlet 10 3.0 0.001 0.002 136.117 136.123 qs_scf_new_mos 117 7.6 0.001 0.001 108.101 108.491 qs_scf_loop_do_ot 117 8.6 0.001 0.001 108.100 108.490 ot_scf_mini 117 9.6 0.004 0.005 103.181 103.624 dbcsr_multiply_generic 2507 12.6 0.190 0.196 101.993 103.215 multiply_cannon 2507 13.6 0.478 0.525 79.035 83.976 multiply_cannon_loop 2507 14.6 1.254 1.300 75.121 78.262 ot_mini 117 10.6 0.001 0.001 52.988 53.408 mp_waitall_1 214728 16.6 27.639 42.455 27.639 42.455 init_scf_loop 11 6.9 0.001 0.005 40.246 40.251 multiply_cannon_multrec 30084 15.6 22.086 27.017 32.079 37.499 rebuild_ks_matrix 128 8.3 0.001 0.001 36.076 36.633 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.022 36.076 36.632 prepare_preconditioner 11 7.9 0.000 0.001 35.347 35.407 make_preconditioner 11 8.9 0.002 0.014 35.347 35.407 make_full_inverse_cholesky 11 9.9 0.000 0.000 33.820 34.475 qs_ks_update_qs_env 128 7.6 0.001 0.001 32.508 33.017 qs_ot_get_p 128 10.4 0.003 0.015 31.539 31.955 qs_ot_get_derivative 117 11.6 0.001 0.001 30.874 31.297 multiply_cannon_metrocomm3 30084 15.6 0.091 0.097 16.015 29.657 qs_ot_p2m_diag 83 11.4 0.188 0.216 25.754 25.803 cp_dbcsr_syevd 83 12.4 0.005 0.006 24.312 24.315 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.124 23.196 apply_single 128 13.6 0.001 0.001 22.124 23.195 ot_diis_step 117 11.6 0.014 0.015 21.932 21.935 multiply_cannon_sync_h2d 30084 15.6 19.258 21.751 19.258 21.751 cp_fm_cholesky_invert 11 10.9 20.560 20.577 20.560 20.577 cp_fm_diag_elpa 83 13.4 0.000 0.001 20.360 20.386 cp_fm_diag_elpa_base 83 14.4 20.032 20.199 20.353 20.382 make_m2s 5014 13.6 0.089 0.093 16.634 18.470 make_images 5014 14.6 1.150 1.336 16.428 18.264 sum_up_and_integrate 128 10.3 0.117 0.134 16.934 16.968 integrate_v_rspace 128 11.3 0.004 0.006 16.817 16.854 qs_rho_update_rho_low 128 7.7 0.001 0.001 16.402 16.432 calculate_rho_elec 128 8.7 0.088 0.106 16.402 16.431 init_scf_run 11 5.9 0.000 0.001 12.983 12.985 scf_env_initial_rho_setup 11 6.9 0.001 0.002 12.983 12.985 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 12.444 12.720 make_images_data 5014 15.6 0.063 0.071 9.940 12.256 density_rs2pw 128 9.7 0.006 0.008 9.674 12.088 rs_pw_transfer 1046 11.9 0.015 0.017 8.845 11.306 hybrid_alltoall_any 5200 16.5 0.343 1.492 8.351 11.237 multiply_cannon_metrocomm4 27577 15.6 0.101 0.114 3.743 10.502 pw_transfer 1547 11.6 0.088 0.111 10.195 10.292 mp_irecv_dv 69486 16.3 3.548 10.121 3.548 10.121 dbcsr_mm_accdrv_process 62242 16.2 4.732 5.469 9.460 10.090 fft_wrap_pw1pw2 1291 12.7 0.011 0.012 9.969 10.072 cp_fm_cholesky_decompose 22 10.9 9.610 9.696 9.610 9.696 wfi_extrapolate 11 7.9 0.001 0.001 9.244 9.244 fft3d_ps 1291 14.7 2.757 2.936 8.337 8.393 fft_wrap_pw1pw2_140 523 13.2 0.469 0.530 8.265 8.393 grid_integrate_task_list 128 12.3 7.094 7.775 7.094 7.775 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.638 7.448 potential_pw2rs 128 12.3 0.016 0.019 7.045 7.099 calculate_dm_sparse 128 9.5 0.001 0.001 6.722 6.844 cp_dbcsr_sm_fm_multiply 37 9.5 0.003 0.007 6.701 6.711 mp_sum_l 7870 13.0 4.563 6.416 4.563 6.416 mp_alltoall_d11v 2415 14.1 4.962 6.393 4.962 6.393 mp_allgather_i34 2507 14.6 2.489 5.868 2.489 5.868 grid_collocate_task_list 128 9.7 4.702 5.865 4.702 5.865 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 5.666 5.840 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.621 5.682 mp_waitany 11748 13.9 2.822 5.314 2.822 5.314 dbcsr_complete_redistribute 395 12.7 0.773 0.878 3.885 4.851 rs_pw_transfer_RS2PW_140 139 11.5 0.353 0.375 2.232 4.667 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=219.212000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=793.545455, yerr=5.549179 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ea454b3df0bf366fae64a409da442cd2660d5a8e_performance_tests/17/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 2.906045E+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 3951168 0.0% 0.0% 100.0% average stack size 0.0 0.0 1697.1 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 937.160704E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1033760 MPI messages size (bytes): total size 2.695213E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.607194E+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 279168 36591108096 131072 < size <= 4194304 654272 987691483136 4194304 < size <= 16777216 65184 925172769472 16777216 < size 28448 745747251200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4001 58227. MP_Allreduce 11000 1007. MP_Sync 86 MP_Alltoall 1700 9383497. MP_SendRecv 7874 75008. MP_ISendRecv 7874 75008. MP_Wait 21654 MP_ISend 11660 275234. MP_IRecv 11660 275234. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.034 210.430 210.431 qs_mol_dyn_low 1 2.0 0.003 0.003 209.751 209.764 qs_forces 11 3.9 0.006 0.007 209.597 209.602 qs_energies 11 4.9 0.001 0.005 202.678 202.697 scf_env_do_scf 11 5.9 0.001 0.001 185.718 185.719 scf_env_do_scf_inner_loop 116 6.6 0.003 0.009 142.538 142.539 velocity_verlet 10 3.0 0.001 0.001 131.252 131.255 qs_scf_new_mos 116 7.6 0.001 0.001 100.286 100.600 qs_scf_loop_do_ot 116 8.6 0.001 0.001 100.285 100.599 ot_scf_mini 116 9.6 0.003 0.004 95.973 96.370 dbcsr_multiply_generic 2485 12.5 0.180 0.184 86.768 88.591 multiply_cannon 2485 13.5 0.494 0.517 63.540 70.763 multiply_cannon_loop 2485 14.5 0.848 0.877 59.872 64.809 ot_mini 116 10.6 0.001 0.001 46.950 47.387 init_scf_loop 11 6.9 0.000 0.000 43.044 43.045 prepare_preconditioner 11 7.9 0.000 0.000 38.735 38.779 make_preconditioner 11 8.9 0.000 0.000 38.735 38.779 mp_waitall_1 169034 16.6 28.596 38.363 28.596 38.363 make_full_inverse_cholesky 11 9.9 0.000 0.000 36.109 37.666 qs_ot_get_p 127 10.4 0.001 0.001 33.896 34.331 rebuild_ks_matrix 127 8.3 0.001 0.001 32.848 33.483 qs_ks_build_kohn_sham_matrix 127 9.3 0.024 0.030 32.847 33.483 qs_ks_update_qs_env 127 7.6 0.001 0.001 29.620 30.205 multiply_cannon_metrocomm3 19880 15.5 0.058 0.063 16.363 29.047 qs_ot_p2m_diag 82 11.4 0.263 0.269 28.504 28.577 qs_ot_get_derivative 116 11.6 0.001 0.002 26.864 27.253 cp_dbcsr_syevd 82 12.4 0.005 0.005 27.134 27.137 multiply_cannon_multrec 19880 15.5 13.192 16.148 22.116 25.096 cp_fm_diag_elpa 82 13.4 0.000 0.000 22.528 22.545 cp_fm_diag_elpa_base 82 14.4 22.043 22.218 22.523 22.539 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.960 21.352 apply_single 127 13.6 0.001 0.001 19.960 21.352 ot_diis_step 116 11.6 0.017 0.018 19.898 19.899 cp_fm_cholesky_invert 11 10.9 19.304 19.322 19.304 19.322 make_m2s 4970 13.5 0.080 0.083 16.486 17.283 make_images 4970 14.5 1.163 1.269 16.258 17.051 qs_rho_update_rho_low 127 7.7 0.001 0.001 16.747 16.785 calculate_rho_elec 127 8.7 0.131 0.146 16.746 16.784 sum_up_and_integrate 127 10.3 0.130 0.143 16.534 16.562 integrate_v_rspace 127 11.3 0.004 0.004 16.404 16.437 multiply_cannon_sync_h2d 19880 15.5 13.912 15.650 13.912 15.650 init_scf_run 11 5.9 0.000 0.001 11.763 11.763 scf_env_initial_rho_setup 11 6.9 0.001 0.002 11.763 11.763 density_rs2pw 127 9.7 0.006 0.006 9.482 11.514 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 11.078 11.355 make_images_data 4970 15.5 0.059 0.067 9.992 11.297 pw_transfer 1535 11.6 0.087 0.107 10.309 10.450 fft_wrap_pw1pw2 1281 12.7 0.010 0.012 10.085 10.234 hybrid_alltoall_any 5155 16.4 0.432 1.981 8.464 10.135 rs_pw_transfer 1038 11.9 0.014 0.015 7.968 10.048 cp_fm_cholesky_decompose 22 10.9 9.944 10.007 9.944 10.007 multiply_cannon_metrocomm4 17395 15.5 0.064 0.075 3.459 9.496 mp_irecv_dv 49801 16.2 3.338 9.244 3.338 9.244 dbcsr_mm_accdrv_process 41158 16.2 4.749 5.576 8.418 8.604 fft_wrap_pw1pw2_140 519 13.2 0.468 0.523 8.359 8.513 fft3d_ps 1281 14.7 2.658 2.878 8.366 8.456 wfi_extrapolate 11 7.9 0.001 0.001 8.246 8.246 grid_integrate_task_list 127 12.3 7.161 7.795 7.161 7.795 cp_fm_upper_to_full 104 14.8 5.919 7.658 5.919 7.658 dbcsr_complete_redistribute 393 12.7 1.162 1.197 5.407 7.297 mp_sum_l 7804 13.0 4.955 7.012 4.955 7.012 potential_pw2rs 127 12.3 0.020 0.022 6.534 6.589 mp_alltoall_d11v 2401 14.1 5.275 6.446 5.275 6.446 calculate_dm_sparse 127 9.5 0.001 0.001 6.101 6.222 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 5.238 5.973 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.821 5.828 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.928 5.813 grid_collocate_task_list 127 9.7 4.873 5.685 4.873 5.685 mp_allgather_i34 2485 14.5 2.207 5.303 2.207 5.303 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.768 4.925 mp_waitany 11660 13.9 2.711 4.723 2.711 4.723 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.506 4.545 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.606 4.455 mp_alltoall_i22 712 14.1 2.275 4.331 2.275 4.331 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=210.431000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=890.181818, yerr=8.099587 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ea454b3df0bf366fae64a409da442cd2660d5a8e_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430457200640 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986252263424 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992001093632 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613077360640 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239162695680 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239162695680 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.232994E+12 0.0% 0.0% 100.0% flops max/rank 4.387242E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806382528 0.0% 0.0% 100.0% number of processed stacks 6026880 0.0% 0.0% 100.0% average stack size 0.0 0.0 1129.3 marketing flops 145.651870E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 1.138762E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1153224 MPI messages size (bytes): total size 2.039489E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.768511E+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 322096 36390305792 131072 < size <= 4194304 721976 792118951936 4194304 < size <= 16777216 70800 669921808880 16777216 < size 30960 541065216000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 57632. MP_Allreduce 11104 1087. MP_Sync 88 MP_Alltoall 1724 12509632. MP_SendRecv 5934 75008. MP_ISendRecv 5934 75008. MP_Wait 22612 MP_ISend 15064 244788. MP_IRecv 15064 244788. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.025 0.053 197.117 197.129 qs_mol_dyn_low 1 2.0 0.004 0.012 196.620 196.633 qs_forces 11 3.9 0.023 0.104 196.469 196.476 qs_energies 11 4.9 0.009 0.036 189.022 189.031 scf_env_do_scf 11 5.9 0.003 0.018 171.859 171.865 velocity_verlet 10 3.0 0.001 0.001 129.911 129.915 scf_env_do_scf_inner_loop 118 6.6 0.007 0.026 123.408 123.414 qs_scf_new_mos 118 7.6 0.001 0.001 84.242 84.551 qs_scf_loop_do_ot 118 8.6 0.001 0.001 84.241 84.551 dbcsr_multiply_generic 2529 12.6 0.191 0.198 81.010 81.678 ot_scf_mini 118 9.6 0.004 0.005 79.670 79.977 multiply_cannon 2529 13.6 0.557 0.593 55.506 58.166 multiply_cannon_loop 2529 14.6 1.191 1.227 51.699 53.593 init_scf_loop 11 6.9 0.009 0.045 48.321 48.333 prepare_preconditioner 11 7.9 0.000 0.000 43.935 43.966 make_preconditioner 11 8.9 0.000 0.001 43.935 43.966 ot_mini 118 10.6 0.001 0.001 43.393 43.707 make_full_inverse_cholesky 11 9.9 0.000 0.000 37.490 42.568 rebuild_ks_matrix 129 8.3 0.001 0.001 31.032 31.332 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.021 31.031 31.332 multiply_cannon_multrec 30348 15.6 14.112 19.785 26.175 31.267 mp_waitall_1 149172 16.7 18.490 28.884 18.490 28.884 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.968 28.225 qs_ot_get_derivative 118 11.6 0.002 0.002 23.523 23.822 qs_ot_get_p 129 10.4 0.003 0.017 22.356 22.646 make_m2s 5058 13.6 0.098 0.104 20.999 22.093 make_images 5058 14.6 1.936 2.256 20.691 21.786 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 19.222 19.794 apply_single 129 13.6 0.001 0.001 19.222 19.794 ot_diis_step 118 11.6 0.018 0.018 19.727 19.730 qs_ot_p2m_diag 84 11.4 0.347 0.395 17.958 18.016 cp_fm_cholesky_invert 11 10.9 17.156 17.165 17.156 17.165 cp_fm_upper_to_full 106 14.8 11.650 17.140 11.650 17.140 cp_dbcsr_syevd 84 12.4 0.005 0.006 16.618 16.621 multiply_cannon_metrocomm3 30348 15.6 0.048 0.050 6.783 15.836 sum_up_and_integrate 129 10.3 0.141 0.153 15.406 15.442 qs_rho_update_rho_low 129 7.7 0.001 0.001 15.332 15.366 calculate_rho_elec 129 8.7 0.177 0.192 15.332 15.366 integrate_v_rspace 129 11.3 0.004 0.005 15.264 15.304 make_images_data 5058 15.6 0.063 0.069 11.491 13.278 cp_fm_diag_elpa 84 13.4 0.000 0.000 13.257 13.271 cp_fm_diag_elpa_base 84 14.4 12.235 12.581 13.247 13.260 multiply_cannon_sync_h2d 30348 15.6 11.817 13.068 11.817 13.068 dbcsr_complete_redistribute 397 12.7 1.503 1.625 9.183 12.982 hybrid_alltoall_any 5245 16.5 0.528 2.223 10.147 12.251 dbcsr_mm_accdrv_process 62780 16.2 7.553 8.561 11.630 12.156 copy_fm_to_dbcsr 210 11.7 0.002 0.003 7.723 11.523 init_scf_run 11 5.9 0.001 0.002 11.131 11.132 scf_env_initial_rho_setup 11 6.9 0.001 0.002 11.131 11.132 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.423 10.146 pw_transfer 1559 11.6 0.089 0.107 10.061 10.125 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.869 10.083 fft_wrap_pw1pw2 1301 12.7 0.010 0.012 9.831 9.900 mp_alltoall_i22 720 14.1 5.667 9.535 5.667 9.535 density_rs2pw 129 9.7 0.006 0.006 7.622 9.104 fft_wrap_pw1pw2_140 527 13.2 0.487 0.496 8.629 8.716 fft3d_ps 1301 14.7 2.806 2.879 8.040 8.126 cp_fm_cholesky_decompose 22 10.9 8.004 8.091 8.004 8.091 grid_integrate_task_list 129 12.3 7.602 8.052 7.602 8.052 wfi_extrapolate 11 7.9 0.001 0.001 7.875 7.876 multiply_cannon_metrocomm4 25290 15.6 0.085 0.098 2.804 7.380 mp_irecv_dv 76751 16.2 2.650 7.104 2.650 7.104 rs_pw_transfer 1054 12.0 0.014 0.015 5.099 6.648 calculate_dm_sparse 129 9.5 0.001 0.001 6.259 6.355 mp_alltoall_d11v 2429 14.1 5.169 5.918 5.169 5.918 grid_collocate_task_list 129 9.7 5.115 5.833 5.115 5.833 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.588 5.633 potential_pw2rs 129 12.3 0.023 0.023 5.046 5.071 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.519 4.614 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.461 4.584 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 4.504 4.506 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.288 4.345 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=197.129000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1073.727273, yerr=21.562134 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ea454b3df0bf366fae64a409da442cd2660d5a8e_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430454546432 0.0% 0.0% 100.0% flops 32 x 32 x 32 1975684956160 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992006770688 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613065416704 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239182565376 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239182565376 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.245913E+12 0.0% 0.0% 100.0% flops max/rank 5.910926E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806580192 0.0% 0.0% 100.0% number of processed stacks 1979136 0.0% 0.0% 100.0% average stack size 0.0 0.0 3439.2 marketing flops 145.663816E+12 ------------------------------------------------------------------------------- # multiplications 2535 max memory usage/rank 1.482174E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 243360 MPI messages size (bytes): total size 1.342108E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.514910E+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 115584 60599304192 4194304 < size <= 16777216 105840 554906419200 16777216 < size 20352 726592540656 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 9028 51. MP_Alltoall 9736 793528. MP_ISend 40516 2095952. MP_IRecv 40516 2095058. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4057 57435. MP_Allreduce 11139 1170. MP_Sync 88 MP_Alltoall 1724 18848081. 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 0.038 0.043 213.120 213.124 qs_mol_dyn_low 1 2.0 0.003 0.003 212.456 212.471 qs_forces 11 3.9 0.004 0.005 212.328 212.337 qs_energies 11 4.9 0.001 0.001 204.498 204.507 scf_env_do_scf 11 5.9 0.001 0.001 185.398 185.403 scf_env_do_scf_inner_loop 118 6.6 0.005 0.025 141.156 141.158 velocity_verlet 10 3.0 0.001 0.001 135.082 135.086 qs_scf_new_mos 118 7.6 0.001 0.001 97.928 98.049 qs_scf_loop_do_ot 118 8.6 0.001 0.001 97.927 98.049 ot_scf_mini 118 9.6 0.003 0.004 93.120 93.212 dbcsr_multiply_generic 2535 12.6 0.184 0.195 81.338 82.385 multiply_cannon 2535 13.6 0.590 0.627 58.039 62.574 multiply_cannon_loop 2535 14.6 0.448 0.455 52.646 54.994 ot_mini 118 10.6 0.001 0.001 44.757 44.872 init_scf_loop 11 6.9 0.000 0.000 44.048 44.050 prepare_preconditioner 11 7.9 0.000 0.000 39.846 39.874 make_preconditioner 11 8.9 0.000 0.000 39.846 39.874 mp_waitall_1 127164 16.7 30.326 37.920 30.326 37.920 make_full_inverse_cholesky 11 9.9 0.000 0.000 37.467 37.840 qs_ot_get_p 129 10.4 0.001 0.001 34.656 34.807 rebuild_ks_matrix 129 8.3 0.001 0.001 32.605 32.821 qs_ks_build_kohn_sham_matrix 129 9.3 0.028 0.029 32.604 32.821 qs_ks_update_qs_env 129 7.6 0.001 0.001 29.706 29.905 qs_ot_p2m_diag 84 11.4 0.502 0.507 29.663 29.682 cp_dbcsr_syevd 84 12.4 0.005 0.006 28.028 28.032 qs_ot_get_derivative 118 11.6 0.001 0.002 24.110 24.211 cp_fm_diag_elpa 84 13.4 0.000 0.000 23.612 23.619 cp_fm_diag_elpa_base 84 14.4 23.253 23.432 23.601 23.608 cp_fm_cholesky_invert 11 10.9 23.440 23.452 23.440 23.452 multiply_cannon_multrec 10140 15.6 10.473 14.459 18.360 22.348 multiply_cannon_metrocomm3 10140 15.6 0.023 0.025 12.608 21.569 make_m2s 5070 13.6 0.067 0.071 18.560 20.740 ot_diis_step 118 11.6 0.020 0.021 20.505 20.506 make_images 5070 14.6 2.283 2.648 18.255 20.438 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.813 20.420 apply_single 129 13.6 0.001 0.001 19.812 20.419 qs_rho_update_rho_low 129 7.7 0.001 0.001 17.737 17.812 calculate_rho_elec 129 8.7 0.261 0.271 17.737 17.812 sum_up_and_integrate 129 10.3 0.183 0.193 17.115 17.163 integrate_v_rspace 129 11.3 0.004 0.006 16.931 16.990 make_images_data 5070 15.6 0.051 0.060 11.356 13.909 hybrid_alltoall_any 5257 16.5 0.851 3.853 10.847 13.453 multiply_cannon_sync_h2d 10140 15.6 11.593 12.214 11.593 12.214 init_scf_run 11 5.9 0.000 0.001 12.040 12.041 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.040 12.041 pw_transfer 1559 11.6 0.087 0.097 11.601 11.652 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 11.374 11.431 cp_fm_cholesky_decompose 22 10.9 10.378 10.481 10.378 10.481 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 10.204 10.264 density_rs2pw 129 9.7 0.005 0.005 8.936 9.807 fft_wrap_pw1pw2_140 527 13.2 0.495 0.519 9.670 9.730 multiply_cannon_metrocomm1 10140 15.6 0.029 0.029 6.340 9.598 fft3d_ps 1301 14.7 2.739 2.820 9.522 9.551 mp_allgather_i34 2535 14.6 3.586 8.798 3.586 8.798 wfi_extrapolate 11 7.9 0.001 0.001 8.317 8.318 grid_integrate_task_list 129 12.3 7.790 8.294 7.790 8.294 dbcsr_mm_accdrv_process 20958 16.1 2.777 3.751 7.524 8.253 mp_alltoall_d11v 2429 14.1 6.495 7.928 6.495 7.928 calculate_dm_sparse 129 9.5 0.001 0.001 7.045 7.171 rs_pw_transfer 1054 12.0 0.013 0.013 6.077 7.003 dbcsr_complete_redistribute 397 12.7 2.106 2.145 6.033 6.538 potential_pw2rs 129 12.3 0.026 0.027 6.069 6.102 grid_collocate_task_list 129 9.7 5.437 6.098 5.437 6.098 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.461 5.467 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.344 5.346 multiply_cannon_metrocomm4 7605 15.6 0.025 0.027 1.903 5.210 mp_irecv_dv 29150 15.9 1.868 5.138 1.868 5.138 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.262 4.378 copy_fm_to_dbcsr 210 11.7 0.001 0.002 3.905 4.343 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.256 4.304 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=213.124000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1393.818182, yerr=36.702287 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ea454b3df0bf366fae64a409da442cd2660d5a8e_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430454546432 0.0% 0.0% 100.0% flops 32 x 32 x 32 1975684956160 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992006770688 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613065416704 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239182565376 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239182565376 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.245913E+12 0.0% 0.0% 100.0% flops max/rank 11.787674E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806580192 0.0% 0.0% 100.0% number of processed stacks 1982496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3433.3 marketing flops 145.663816E+12 ------------------------------------------------------------------------------- # multiplications 2535 max memory usage/rank 3.100123E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101400 MPI messages size (bytes): total size 1.145171E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.293599E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 45888 35634806784 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592540656 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4041 58592. MP_Allreduce 11099 1511. MP_Sync 88 MP_Alltoall 1724 36993632. MP_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 0.035 0.140 310.644 310.692 qs_mol_dyn_low 1 2.0 0.003 0.003 310.005 310.013 qs_forces 11 3.9 0.004 0.005 309.898 309.902 qs_energies 11 4.9 0.001 0.001 300.378 300.385 scf_env_do_scf 11 5.9 0.001 0.001 277.300 277.311 velocity_verlet 10 3.0 0.001 0.001 221.533 221.541 scf_env_do_scf_inner_loop 118 6.6 0.015 0.038 149.708 149.709 init_scf_loop 11 6.9 0.000 0.000 127.332 127.334 prepare_preconditioner 11 7.9 0.000 0.000 122.023 122.054 make_preconditioner 11 8.9 0.000 0.000 122.023 122.054 make_full_inverse_cholesky 11 9.9 0.000 0.000 97.702 119.194 qs_scf_new_mos 118 7.6 0.001 0.001 96.824 96.960 qs_scf_loop_do_ot 118 8.6 0.001 0.001 96.823 96.959 ot_scf_mini 118 9.6 0.004 0.004 91.904 91.985 dbcsr_multiply_generic 2535 12.6 0.214 0.227 86.861 87.490 cp_fm_upper_to_full 106 14.8 53.618 77.040 53.618 77.040 multiply_cannon 2535 13.6 0.722 0.771 60.729 61.542 multiply_cannon_loop 2535 14.6 0.475 0.482 56.708 58.532 ot_mini 118 10.6 0.001 0.001 46.834 46.909 dbcsr_complete_redistribute 397 12.7 3.977 4.013 30.122 43.378 rebuild_ks_matrix 129 8.3 0.001 0.001 39.842 39.954 qs_ks_build_kohn_sham_matrix 129 9.3 0.019 0.019 39.841 39.953 copy_fm_to_dbcsr 210 11.7 0.002 0.002 26.630 39.861 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 24.277 37.307 qs_ks_update_qs_env 129 7.6 0.001 0.001 36.742 36.851 mp_alltoall_i22 720 14.1 22.082 35.090 22.082 35.090 cp_fm_cholesky_invert 11 10.9 34.635 34.641 34.635 34.641 mp_waitall_1 104820 16.8 30.294 34.246 30.294 34.246 qs_ot_get_p 129 10.4 0.001 0.001 29.104 29.161 qs_ot_get_derivative 118 11.6 0.002 0.002 25.581 25.661 qs_ot_p2m_diag 84 11.4 0.889 0.894 24.715 24.750 cp_dbcsr_syevd 84 12.4 0.006 0.006 22.881 22.883 make_m2s 5070 13.6 0.078 0.080 21.114 22.374 make_images 5070 14.6 3.785 3.910 20.632 21.895 qs_rho_update_rho_low 129 7.7 0.001 0.001 21.530 21.550 calculate_rho_elec 129 8.7 0.486 0.487 21.530 21.550 multiply_cannon_metrocomm3 10140 15.6 0.023 0.025 19.830 21.498 ot_diis_step 118 11.6 0.023 0.023 21.209 21.210 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 20.371 20.681 apply_single 129 13.6 0.001 0.001 20.370 20.681 sum_up_and_integrate 129 10.3 0.325 0.328 20.474 20.563 integrate_v_rspace 129 11.3 0.004 0.004 20.148 20.237 cp_fm_diag_elpa 84 13.4 0.000 0.000 19.441 19.442 cp_fm_diag_elpa_base 84 14.4 14.884 16.530 19.436 19.437 multiply_cannon_multrec 10140 15.6 10.590 12.461 18.324 18.418 multiply_cannon_sync_h2d 10140 15.6 15.773 15.787 15.773 15.787 hybrid_alltoall_any 5257 16.5 1.311 3.059 11.695 13.824 make_images_data 5070 15.6 0.060 0.064 11.613 13.797 pw_transfer 1559 11.6 0.094 0.094 12.971 13.003 init_scf_run 11 5.9 0.000 0.001 12.918 12.919 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.917 12.918 fft_wrap_pw1pw2 1301 12.7 0.011 0.011 12.733 12.767 fft_wrap_pw1pw2_140 527 13.2 0.546 0.548 11.235 11.268 fft3d_ps 1301 14.7 2.767 2.775 10.754 10.786 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 10.142 10.212 mp_alltoall_d11v 2429 14.1 8.865 9.815 8.865 9.815 wfi_extrapolate 11 7.9 0.001 0.001 9.426 9.426 cp_fm_cholesky_decompose 22 10.9 9.341 9.369 9.341 9.369 dbcsr_mm_accdrv_process 20958 16.1 3.987 5.868 7.494 9.359 density_rs2pw 129 9.7 0.005 0.005 8.830 8.958 grid_integrate_task_list 129 12.3 8.689 8.838 8.689 8.838 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 8.095 8.096 calculate_dm_sparse 129 9.5 0.001 0.001 6.965 7.043 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.645 6.753 grid_collocate_task_list 129 9.7 6.439 6.465 6.439 6.465 rs_scatter_matrices 140 9.7 3.648 4.578 6.234 6.464 copy_dbcsr_to_fm 187 11.8 0.004 0.004 6.245 6.320 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=310.692000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2734.636364, yerr=195.939356 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ea454b3df0bf366fae64a409da442cd2660d5a8e_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 1.262219E+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 0.054 0.167 85.233 85.245 qs_energies 1 2.0 0.000 0.001 84.486 84.495 ls_scf 1 3.0 0.000 0.000 83.484 83.493 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.489 72.654 multiply_cannon 111 7.7 0.017 0.020 56.021 57.188 multiply_cannon_loop 111 8.7 0.208 0.224 52.646 53.989 ls_scf_main 1 4.0 0.000 0.000 52.118 52.129 density_matrix_trs4 2 5.0 0.002 0.003 46.577 46.651 ls_scf_init_scf 1 4.0 0.000 0.000 28.307 28.309 ls_scf_init_matrix_S 1 5.0 0.000 0.001 27.183 27.239 mp_waitall_1 11031 10.9 22.411 25.788 22.411 25.788 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.061 25.075 multiply_cannon_multrec 2664 9.7 8.159 9.086 15.464 17.392 multiply_cannon_sync_h2d 2664 9.7 13.722 15.736 13.722 15.736 make_m2s 222 7.7 0.008 0.011 12.955 13.362 make_images 222 8.7 0.099 0.110 12.933 13.340 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 9.835 12.333 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.372 8.188 make_images_data 222 9.7 0.004 0.005 7.531 8.012 dbcsr_mm_accdrv_process 4760 10.4 0.513 0.641 6.923 7.949 hybrid_alltoall_any 227 10.6 0.215 1.848 6.524 7.679 dbcsr_mm_accdrv_process_sort 4760 11.4 6.205 7.132 6.205 7.132 calculate_norms 4752 9.8 5.529 6.275 5.529 6.275 apply_matrix_preconditioner 6 5.3 0.001 0.001 5.046 5.222 mp_sum_l 807 5.4 3.043 4.443 3.043 4.443 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.041 3.693 mp_irecv_dv 6231 10.9 2.023 3.661 2.023 3.661 make_images_sizes 222 9.7 0.000 0.000 0.673 3.564 mp_alltoall_i44 222 10.7 0.672 3.564 0.672 3.564 arnoldi_extremal 4 6.8 0.000 0.000 3.251 3.280 arnoldi_normal_ev 4 7.8 0.001 0.002 3.251 3.280 build_subspace 16 8.4 0.009 0.012 3.150 3.153 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.189 3.068 ls_scf_post 1 4.0 0.000 0.000 3.059 3.067 ls_scf_store_result 1 5.0 0.000 0.000 2.859 2.900 dbcsr_special_finalize 555 9.7 0.005 0.006 2.398 2.728 dbcsr_merge_single_wm 555 10.7 0.461 0.617 2.390 2.720 make_images_pack 222 9.7 2.206 2.617 2.208 2.619 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.327 2.589 dbcsr_sort_data 658 11.4 2.187 2.499 2.187 2.499 dbcsr_matrix_vector_mult_local 304 10.0 2.068 2.481 2.070 2.483 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.304 2.387 buffer_matrices_ensure_size 222 8.7 1.749 2.093 1.749 2.093 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.798 1.799 rebuild_ks_matrix 3 7.3 0.000 0.000 1.788 1.790 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.788 1.790 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.245000, 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/ea454b3df0bf366fae64a409da442cd2660d5a8e_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 2.105037E+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 0.045 0.071 91.939 91.942 qs_energies 1 2.0 0.000 0.000 91.184 91.190 ls_scf 1 3.0 0.000 0.001 89.772 89.779 dbcsr_multiply_generic 111 6.7 0.019 0.042 75.257 75.564 multiply_cannon 111 7.7 0.029 0.039 53.339 57.440 ls_scf_main 1 4.0 0.000 0.000 55.090 55.097 multiply_cannon_loop 111 8.7 0.116 0.122 50.022 53.347 density_matrix_trs4 2 5.0 0.002 0.003 49.285 49.524 ls_scf_init_scf 1 4.0 0.000 0.002 31.153 31.154 mp_waitall_1 9105 10.9 21.132 30.681 21.132 30.681 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.843 29.954 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 27.442 27.463 multiply_cannon_multrec 1332 9.7 13.151 16.893 22.360 27.200 multiply_cannon_metrocomm3 1332 9.7 0.006 0.007 11.710 21.369 make_m2s 222 7.7 0.007 0.009 15.320 16.035 make_images 222 8.7 1.579 1.963 15.290 16.006 dbcsr_mm_accdrv_process 4041 10.4 0.288 0.456 8.806 10.402 dbcsr_mm_accdrv_process_sort 4041 11.4 8.399 9.947 8.399 9.947 make_images_data 222 9.7 0.004 0.004 8.802 9.789 hybrid_alltoall_any 227 10.6 0.524 2.476 8.203 9.285 mp_sum_l 807 5.4 5.306 8.587 5.306 8.587 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.217 7.743 mp_irecv_dv 3311 11.0 3.198 7.678 3.198 7.678 calculate_norms 2376 9.8 6.013 6.730 6.013 6.730 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.048 6.667 multiply_cannon_sync_h2d 1332 9.7 4.810 6.102 4.810 6.102 apply_matrix_preconditioner 6 5.3 0.001 0.001 5.080 5.336 arnoldi_extremal 4 6.8 0.000 0.000 4.949 4.979 arnoldi_normal_ev 4 7.8 0.001 0.005 4.949 4.978 build_subspace 16 8.4 0.014 0.020 4.607 4.612 ls_scf_post 1 4.0 0.000 0.000 3.529 3.535 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.240 3.453 ls_scf_store_result 1 5.0 0.000 0.000 3.239 3.349 dbcsr_matrix_vector_mult_local 304 10.0 2.733 3.215 2.735 3.217 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.238 2.785 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.663 2.760 mp_allgather_i34 111 8.7 1.003 2.437 1.003 2.437 make_images_pack 222 9.7 2.025 2.399 2.027 2.401 dbcsr_sort_data 436 11.2 1.794 2.025 1.794 2.025 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.997 1.999 rebuild_ks_matrix 3 7.3 0.000 0.000 1.984 1.986 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.003 1.984 1.986 dbcsr_data_new 4174 10.1 1.610 1.857 1.610 1.857 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.942000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1714.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ea454b3df0bf366fae64a409da442cd2660d5a8e_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 2.685579E+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 0.065 0.127 95.086 95.089 qs_energies 1 2.0 0.000 0.000 94.072 94.076 ls_scf 1 3.0 0.002 0.014 92.644 92.647 dbcsr_multiply_generic 111 6.7 0.016 0.016 76.653 76.942 ls_scf_main 1 4.0 0.001 0.009 57.667 57.673 multiply_cannon 111 7.7 0.044 0.089 53.031 56.971 multiply_cannon_loop 111 8.7 0.099 0.106 49.455 53.293 density_matrix_trs4 2 5.0 0.003 0.016 51.634 51.833 mp_waitall_1 7281 11.0 23.849 33.923 23.849 33.923 ls_scf_init_scf 1 4.0 0.000 0.003 31.386 31.388 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.934 30.021 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.502 27.513 multiply_cannon_multrec 888 9.7 12.553 15.119 21.075 24.195 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.204 23.324 make_m2s 222 7.7 0.006 0.007 16.933 18.164 make_images 222 8.7 1.965 2.257 16.895 18.124 hybrid_alltoall_any 227 10.6 0.623 2.856 9.254 10.710 make_images_data 222 9.7 0.003 0.004 9.563 10.590 dbcsr_mm_accdrv_process 3754 10.4 0.251 0.458 8.043 9.272 dbcsr_mm_accdrv_process_sort 3754 11.4 7.661 8.814 7.661 8.814 mp_sum_l 807 5.4 5.154 8.541 5.154 8.541 multiply_cannon_sync_h2d 888 9.7 6.077 7.707 6.077 7.707 multiply_cannon_metrocomm1 888 9.7 0.002 0.003 3.701 7.129 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.439 6.880 mp_irecv_dv 2335 11.1 2.425 6.838 2.425 6.838 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.935 6.699 arnoldi_extremal 4 6.8 0.000 0.000 5.402 5.415 arnoldi_normal_ev 4 7.8 0.001 0.004 5.401 5.415 apply_matrix_preconditioner 6 5.3 0.002 0.009 4.960 5.207 build_subspace 16 8.4 0.014 0.020 5.090 5.096 calculate_norms 1584 9.8 4.267 4.590 4.267 4.590 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.579 3.873 mp_allgather_i34 111 8.7 1.406 3.809 1.406 3.809 ls_scf_post 1 4.0 0.000 0.001 3.589 3.595 dbcsr_matrix_vector_mult_local 304 10.0 3.005 3.586 3.007 3.588 ls_scf_store_result 1 5.0 0.000 0.000 3.303 3.417 ls_scf_dm_to_ks 2 5.0 0.002 0.024 2.890 2.990 dbcsr_sort_data 325 11.1 1.877 2.138 1.877 2.138 make_images_pack 222 9.7 1.813 2.117 1.816 2.119 make_images_sizes 222 9.7 0.000 0.000 1.079 2.076 mp_alltoall_i44 222 10.7 1.078 2.076 1.078 2.076 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.023 2.025 rebuild_ks_matrix 3 7.3 0.000 0.000 2.005 2.007 qs_ks_build_kohn_sham_matrix 3 8.3 0.005 0.019 2.005 2.007 dbcsr_data_release 9322 10.9 1.298 1.933 1.298 1.933 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=95.089000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2191.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ea454b3df0bf366fae64a409da442cd2660d5a8e_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 3.337249E+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 0.059 0.076 97.229 97.240 qs_energies 1 2.0 0.000 0.000 96.455 96.465 ls_scf 1 3.0 0.004 0.011 94.799 94.811 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.192 78.403 ls_scf_main 1 4.0 0.003 0.020 58.604 58.613 multiply_cannon 111 7.7 0.052 0.214 51.170 55.943 density_matrix_trs4 2 5.0 0.005 0.057 52.464 52.577 multiply_cannon_loop 111 8.7 0.115 0.126 46.125 49.434 ls_scf_init_scf 1 4.0 0.000 0.002 32.959 32.961 ls_scf_init_matrix_S 1 5.0 0.000 0.001 31.387 31.455 mp_waitall_1 6369 11.0 22.111 28.951 22.111 28.951 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.914 28.926 multiply_cannon_multrec 1332 9.7 14.326 17.555 22.189 24.717 make_m2s 222 7.7 0.006 0.008 21.169 22.582 make_images 222 8.7 3.138 3.588 21.119 22.534 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 8.626 16.622 make_images_data 222 9.7 0.004 0.004 11.813 13.380 hybrid_alltoall_any 227 10.6 0.799 3.833 11.129 12.856 dbcsr_mm_accdrv_process 3641 10.4 0.208 0.451 7.496 8.940 dbcsr_mm_accdrv_process_sort 3641 11.4 7.110 8.522 7.110 8.522 mp_sum_l 807 5.4 4.273 7.614 4.273 7.614 multiply_cannon_sync_h2d 1332 9.7 5.565 6.104 5.565 6.104 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.502 6.007 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.239 5.935 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.089 5.923 mp_irecv_dv 3229 10.9 2.065 5.837 2.065 5.837 arnoldi_extremal 4 6.8 0.000 0.000 5.175 5.188 arnoldi_normal_ev 4 7.8 0.001 0.005 5.175 5.188 build_subspace 16 8.4 0.014 0.021 4.832 4.841 apply_matrix_preconditioner 6 5.3 0.001 0.002 4.536 4.750 mp_allgather_i34 111 8.7 2.169 4.547 2.169 4.547 calculate_norms 2376 9.8 4.199 4.537 4.199 4.537 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.558 3.848 dbcsr_matrix_vector_mult_local 304 10.0 3.172 3.665 3.174 3.667 dbcsr_sort_data 658 11.4 3.140 3.449 3.140 3.449 ls_scf_post 1 4.0 0.002 0.006 3.233 3.243 dbcsr_special_finalize 555 9.7 0.006 0.006 2.873 3.197 dbcsr_merge_single_wm 555 10.7 0.535 0.650 2.865 3.190 ls_scf_dm_to_ks 2 5.0 0.000 0.002 3.045 3.078 ls_scf_store_result 1 5.0 0.000 0.000 2.962 3.019 dbcsr_data_release 10477 10.7 1.588 2.527 1.588 2.527 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.172 2.174 rebuild_ks_matrix 3 7.3 0.000 0.000 2.149 2.151 qs_ks_build_kohn_sham_matrix 3 8.3 0.009 0.056 2.149 2.151 dbcsr_finalize 304 7.8 0.049 0.061 1.807 2.059 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.240000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2696.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ea454b3df0bf366fae64a409da442cd2660d5a8e_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 4.638757E+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 0.042 0.058 99.828 99.829 qs_energies 1 2.0 0.000 0.000 98.701 98.707 ls_scf 1 3.0 0.000 0.000 96.736 96.743 dbcsr_multiply_generic 111 6.7 0.018 0.018 77.842 78.118 ls_scf_main 1 4.0 0.000 0.000 62.087 62.088 multiply_cannon 111 7.7 0.090 0.188 54.755 60.011 density_matrix_trs4 2 5.0 0.002 0.003 54.867 54.975 multiply_cannon_loop 111 8.7 0.069 0.074 50.396 52.398 ls_scf_init_scf 1 4.0 0.000 0.000 31.049 31.054 mp_waitall_1 5436 11.0 26.456 30.529 26.456 30.529 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.658 29.695 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.484 27.492 multiply_cannon_multrec 444 9.7 13.960 16.405 20.975 22.708 make_m2s 222 7.7 0.004 0.005 18.206 20.584 make_images 222 8.7 3.715 4.390 18.145 20.525 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.883 16.658 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.146 13.542 make_images_data 222 9.7 0.003 0.004 10.225 12.747 hybrid_alltoall_any 227 10.6 0.789 3.782 9.894 12.660 multiply_cannon_sync_h2d 444 9.7 6.494 7.903 6.494 7.903 dbcsr_mm_accdrv_process 3003 10.4 0.188 0.345 6.719 7.857 dbcsr_mm_accdrv_process_sort 3003 11.4 6.404 7.510 6.404 7.510 mp_allgather_i34 111 8.7 2.598 7.008 2.598 7.008 arnoldi_extremal 4 6.8 0.000 0.000 5.969 5.979 arnoldi_normal_ev 4 7.8 0.001 0.004 5.969 5.979 build_subspace 16 8.4 0.015 0.020 5.571 5.581 apply_matrix_preconditioner 6 5.3 0.001 0.002 4.576 4.735 mp_sum_l 807 5.4 2.937 4.726 2.937 4.726 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.204 4.410 dbcsr_matrix_vector_mult_local 304 10.0 3.664 4.137 3.666 4.139 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.659 3.862 make_images_sizes 222 9.7 0.000 0.000 1.346 3.846 mp_alltoall_i44 222 10.7 1.346 3.845 1.346 3.845 mp_irecv_dv 1241 11.2 1.641 3.840 1.641 3.840 calculate_norms 792 9.8 3.574 3.720 3.574 3.720 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.014 3.703 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.623 3.678 ls_scf_post 1 4.0 0.000 0.000 3.601 3.607 ls_scf_store_result 1 5.0 0.000 0.000 3.373 3.428 dbcsr_finalize 304 7.8 0.062 0.078 2.200 2.259 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.191 2.193 rebuild_ks_matrix 3 7.3 0.000 0.000 2.158 2.160 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.158 2.160 dbcsr_merge_all 275 8.9 0.474 0.524 2.045 2.095 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=99.829000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3680.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ea454b3df0bf366fae64a409da442cd2660d5a8e_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 8.723292E+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 0.090 0.142 109.591 109.604 qs_energies 1 2.0 0.000 0.000 108.012 108.025 ls_scf 1 3.0 0.000 0.000 105.028 105.039 dbcsr_multiply_generic 111 6.7 0.023 0.027 77.734 77.877 ls_scf_main 1 4.0 0.000 0.000 66.494 66.495 density_matrix_trs4 2 5.0 0.002 0.003 56.994 57.066 multiply_cannon 111 7.7 0.126 0.204 49.567 51.315 multiply_cannon_loop 111 8.7 0.067 0.070 46.110 47.308 ls_scf_init_scf 1 4.0 0.000 0.000 34.853 34.853 ls_scf_init_matrix_S 1 5.0 0.000 0.001 33.074 33.093 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 30.144 30.153 mp_waitall_1 4527 11.1 22.189 26.238 22.189 26.238 make_m2s 222 7.7 0.005 0.005 24.336 25.326 make_images 222 8.7 4.591 5.006 24.230 25.218 multiply_cannon_multrec 444 9.7 17.888 18.621 22.487 23.207 hybrid_alltoall_any 227 10.6 1.663 3.624 13.212 16.012 make_images_data 222 9.7 0.003 0.003 13.459 15.972 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.142 10.750 multiply_cannon_sync_h2d 444 9.7 8.841 8.890 8.841 8.890 arnoldi_extremal 4 6.8 0.000 0.000 7.568 7.576 arnoldi_normal_ev 4 7.8 0.002 0.008 7.568 7.576 build_subspace 16 8.4 0.026 0.036 7.004 7.011 dbcsr_matrix_vector_mult 304 9.0 0.017 0.032 5.531 5.671 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.386 5.478 apply_matrix_preconditioner 6 5.3 0.001 0.001 5.146 5.435 dbcsr_matrix_vector_mult_local 304 10.0 4.987 5.294 4.989 5.297 dbcsr_mm_accdrv_process 1814 10.4 0.283 0.334 4.425 4.546 dbcsr_mm_accdrv_process_sort 1814 11.4 4.119 4.244 4.119 4.244 make_images_sizes 222 9.7 0.000 0.000 1.520 3.800 mp_alltoall_i44 222 10.7 1.519 3.800 1.519 3.800 ls_scf_post 1 4.0 0.000 0.000 3.681 3.693 ls_scf_store_result 1 5.0 0.000 0.000 3.408 3.419 mp_allgather_i34 111 8.7 1.098 3.415 1.098 3.415 calculate_norms 792 9.8 3.246 3.288 3.246 3.288 dbcsr_finalize 304 7.8 0.082 0.089 3.090 3.183 dbcsr_merge_all 275 8.9 0.891 0.918 2.876 2.964 qs_energies_init_hamiltonians 1 3.0 0.001 0.002 2.954 2.954 dbcsr_complete_redistribute 5 7.6 1.432 1.462 2.799 2.910 dbcsr_data_release 12724 10.6 2.338 2.849 2.338 2.849 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.694 2.696 rebuild_ks_matrix 3 7.3 0.000 0.000 2.625 2.627 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.625 2.627 matrix_ls_to_qs 2 6.0 0.000 0.000 2.462 2.591 dbcsr_sort_data 325 11.1 2.447 2.519 2.447 2.519 dbcsr_new_transposed 4 7.5 0.256 0.273 2.454 2.475 mp_sum_l 807 5.4 1.516 2.357 1.516 2.357 dbcsr_frobenius_norm 74 6.6 2.055 2.131 2.200 2.247 mp_alltoall_d11v 48 9.2 2.194 2.244 2.194 2.244 dbcsr_add_d 103 6.2 0.000 0.000 2.133 2.205 dbcsr_add_anytype 103 7.2 0.860 0.894 2.132 2.205 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=109.604000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6841.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ea454b3df0bf366fae64a409da442cd2660d5a8e_performance_tests/27/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 7009386627072 0.0% 0.0% 100.0% flops 9 x 9 x 32 7335108845568 0.0% 0.0% 100.0% flops 9 x 22 x 32 9866241589248 0.0% 0.0% 100.0% flops 22 x 9 x 32 9884108906496 0.0% 0.0% 100.0% flops 22 x 22 x 32 13354440523776 0.0% 0.0% 100.0% flops 32 x 32 x 9 20607185977344 0.0% 0.0% 100.0% flops 32 x 32 x 22 25186560638976 0.0% 0.0% 100.0% flops 9 x 32 x 32 28458319085568 0.0% 0.0% 100.0% flops 22 x 32 x 32 34782389993472 0.0% 0.0% 100.0% flops 9 x 32 x 9 42881542373376 0.0% 0.0% 100.0% flops 22 x 32 x 9 55680402235392 0.0% 0.0% 100.0% flops 9 x 32 x 22 55680402235392 0.0% 0.0% 100.0% flops 22 x 32 x 22 72328573419520 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 383.054662E+12 0.0% 0.0% 100.0% flops max/rank 733.641090E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 26899403712 0.0% 0.0% 100.0% number of processed stacks 118860288 0.0% 0.0% 100.0% average stack size 0.0 0.0 226.3 marketing flops 780.439111E+12 ------------------------------------------------------------------------------- # multiplications 1445 max memory usage/rank 590.618624E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 102097920 MPI messages size (bytes): total size 37.227590E+12 min size 0.000000E+00 max size 4.551360E+06 average size 364.626312E+03 MPI breakdown and total messages size (bytes): size <= 128 731472 0 128 < size <= 8192 11922720 97670922240 8192 < size <= 32768 24718992 614677610496 32768 < size <= 131072 20000256 1970081366016 131072 < size <= 4194304 42515668 24886801223040 4194304 < size <= 16777216 2208812 9656099886720 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4640 76559. MP_Allreduce 13152 2094. MP_Sync 1064 MP_Alltoall 2588 694801483. MP_SendRecv 168740 11136. MP_ISendRecv 92040 11136. MP_Wait 102830 MP_comm_split 40 MP_ISend 26090 85106. MP_IRecv 37890 59644. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.058 0.206 290.184 290.186 qs_mol_dyn_low 1 2.0 0.004 0.040 288.664 288.682 qs_forces 5 3.8 0.005 0.036 288.498 288.501 qs_energies 5 4.8 0.003 0.034 284.462 284.511 scf_env_do_scf 5 5.8 0.000 0.010 266.268 266.271 scf_env_do_scf_inner_loop 105 6.6 0.003 0.045 227.748 227.751 qs_scf_new_mos 105 7.6 0.000 0.001 170.926 171.330 qs_scf_loop_do_ot 105 8.6 0.001 0.001 170.926 171.329 ot_scf_mini 105 9.6 0.004 0.051 160.789 161.100 dbcsr_multiply_generic 1445 12.2 0.125 0.133 139.468 140.921 velocity_verlet 4 3.0 0.003 0.061 133.215 133.217 multiply_cannon 1445 13.2 0.284 0.297 116.962 121.466 multiply_cannon_loop 1445 14.2 2.388 2.523 113.989 117.885 qs_ot_get_p 112 10.4 0.001 0.001 66.231 66.703 ot_mini 105 10.6 0.001 0.009 63.757 63.894 qs_ot_p2m_diag 40 11.0 0.020 0.030 53.814 53.961 mp_waitall_1 488190 16.1 41.314 53.860 41.314 53.860 cp_dbcsr_syevd 40 12.0 0.002 0.004 50.256 50.258 multiply_cannon_multrec 69360 15.2 31.277 36.821 39.145 44.791 cp_fm_syevd 40 13.0 0.000 0.003 44.167 44.322 qs_ot_get_derivative 55 11.6 0.001 0.001 41.802 42.007 multiply_cannon_metrocomm3 69360 15.2 0.188 0.200 26.699 40.861 cp_fm_redistribute_end 40 14.0 19.383 38.651 19.391 38.655 cp_fm_syevd_base 40 14.0 19.248 38.517 19.248 38.517 init_scf_loop 7 6.6 0.000 0.005 38.444 38.446 rebuild_ks_matrix 110 8.4 0.000 0.000 37.026 37.198 qs_ks_build_kohn_sham_matrix 110 9.4 0.027 0.066 37.025 37.198 multiply_cannon_sync_h2d 69360 15.2 29.775 34.957 29.775 34.957 qs_ks_update_qs_env 112 7.6 0.001 0.001 33.783 33.935 prepare_preconditioner 7 7.6 0.000 0.001 32.453 32.485 make_preconditioner 7 8.6 0.001 0.007 32.453 32.485 qs_rho_update_rho_low 110 7.6 0.001 0.002 27.971 28.272 calculate_rho_elec 110 8.6 0.029 0.032 27.970 28.272 rs_pw_transfer 690 11.5 0.011 0.013 24.964 26.154 make_full_inverse_cholesky 7 9.6 0.000 0.001 23.788 23.857 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 23.236 23.469 apply_single 62 13.6 0.000 0.000 23.236 23.469 density_rs2pw 110 9.6 0.005 0.007 21.133 22.278 ot_new_cg_direction 55 11.6 0.001 0.003 21.191 21.193 pw_transfer 1645 12.4 0.079 0.098 15.484 15.999 cp_fm_cholesky_invert 7 10.6 15.947 15.964 15.947 15.964 fft_wrap_pw1pw2 1425 13.5 0.012 0.015 15.346 15.856 qs_ot_get_orbitals 105 10.6 0.001 0.001 15.406 15.780 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 15.511 15.675 init_scf_run 5 5.8 0.000 0.001 14.720 14.722 scf_env_initial_rho_setup 5 6.8 0.001 0.008 14.720 14.721 sum_up_and_integrate 60 10.3 0.028 0.033 14.229 14.239 integrate_v_rspace 60 11.3 0.002 0.003 14.201 14.212 mp_sum_l 4684 12.4 12.774 13.765 12.774 13.765 fft_wrap_pw1pw2_240 915 15.0 0.417 0.464 13.107 13.504 qs_vxc_create 110 10.4 0.002 0.005 12.379 12.419 fft3d_pb 915 16.0 2.319 2.650 11.939 12.402 calculate_dm_sparse 110 9.5 0.000 0.001 11.835 12.020 check_diag 80 13.5 8.584 8.852 10.891 11.043 multiply_cannon_metrocomm1 69360 15.2 0.096 0.102 5.204 10.996 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 10.598 10.672 rs_pw_transfer_RS2PW_30 110 11.6 1.547 1.644 9.528 10.549 mp_sendrecv_dv 168740 12.6 10.009 10.188 10.009 10.188 make_m2s 2890 13.2 0.076 0.082 9.354 10.054 potential_pw2rs 60 12.3 0.003 0.003 9.889 9.968 make_images 2890 14.2 0.249 0.269 9.252 9.951 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.001 8.903 8.920 xc_rho_set_and_dset_create 110 12.4 0.085 0.109 8.363 8.633 dbcsr_mm_accdrv_process 154766 15.8 4.055 4.224 7.742 8.529 calculate_first_density_matrix 1 7.0 0.000 0.005 8.491 8.504 xc_vxc_pw_create 60 11.3 0.041 0.065 8.462 8.503 acc_transpose_blocks 69360 15.2 0.367 0.386 7.940 8.499 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 8.278 8.335 make_full_single_inverse 7 9.6 0.001 0.002 8.253 8.285 mp_alltoall_z22v 2340 17.7 7.109 8.116 7.109 8.116 xc_pw_derive 510 13.4 0.005 0.006 7.872 7.954 cp_fm_cholesky_decompose 14 10.2 7.230 7.241 7.230 7.241 acc_transpose_blocks_kernels 69360 16.2 0.851 0.906 6.027 6.501 mp_alltoall_d11v 1300 13.8 5.270 6.081 5.270 6.081 mp_waitany 7680 13.5 4.836 6.064 4.836 6.064 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=290.186000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=559.600000, yerr=4.029888 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: ea454b3df0bf366fae64a409da442cd2660d5a8e Summary: empty Status: OK