=== 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: d71249dc44ddccc12d8161011fa5955b1ce8fead ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, Cray-FFTW 3.3.8.10, # COSMA 2.6.2, ELPA 2022.05.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.0.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.0, # SIRIUS 7.3.2, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Author: Matthias Krack (26.10.2022) # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2022.05.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.0.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.0 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.3.2 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MPI_VERSION=3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta_prefixed_scalapack.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d71249dc44ddccc12d8161011fa5955b1ce8fead_performance_tests/01 job id: 42401274 --- 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/d71249dc44ddccc12d8161011fa5955b1ce8fead_performance_tests/02 job id: 42401275 --- 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/d71249dc44ddccc12d8161011fa5955b1ce8fead_performance_tests/03 job id: 42401276 --- 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/d71249dc44ddccc12d8161011fa5955b1ce8fead_performance_tests/04 job id: 42401277 --- 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/d71249dc44ddccc12d8161011fa5955b1ce8fead_performance_tests/05 job id: 42401278 --- 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/d71249dc44ddccc12d8161011fa5955b1ce8fead_performance_tests/06 job id: 42401280 --- 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/d71249dc44ddccc12d8161011fa5955b1ce8fead_performance_tests/07 job id: 42401281 --- 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/d71249dc44ddccc12d8161011fa5955b1ce8fead_performance_tests/08 job id: 42401282 --- 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/d71249dc44ddccc12d8161011fa5955b1ce8fead_performance_tests/09 job id: 42401283 --- 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/d71249dc44ddccc12d8161011fa5955b1ce8fead_performance_tests/10 job id: 42401284 --- 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/d71249dc44ddccc12d8161011fa5955b1ce8fead_performance_tests/11 job id: 42401286 --- 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/d71249dc44ddccc12d8161011fa5955b1ce8fead_performance_tests/12 job id: 42401287 --- 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/d71249dc44ddccc12d8161011fa5955b1ce8fead_performance_tests/13 job id: 42401288 --- 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/d71249dc44ddccc12d8161011fa5955b1ce8fead_performance_tests/14 job id: 42401289 --- 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/d71249dc44ddccc12d8161011fa5955b1ce8fead_performance_tests/15 job id: 42401292 --- 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/d71249dc44ddccc12d8161011fa5955b1ce8fead_performance_tests/16 job id: 42401293 --- 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/d71249dc44ddccc12d8161011fa5955b1ce8fead_performance_tests/17 job id: 42401294 --- 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/d71249dc44ddccc12d8161011fa5955b1ce8fead_performance_tests/18 job id: 42401295 --- 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/d71249dc44ddccc12d8161011fa5955b1ce8fead_performance_tests/19 job id: 42401296 --- 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/d71249dc44ddccc12d8161011fa5955b1ce8fead_performance_tests/20 job id: 42401297 --- 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/d71249dc44ddccc12d8161011fa5955b1ce8fead_performance_tests/21 job id: 42401299 --- 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/d71249dc44ddccc12d8161011fa5955b1ce8fead_performance_tests/22 job id: 42401300 --- 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/d71249dc44ddccc12d8161011fa5955b1ce8fead_performance_tests/23 job id: 42401301 --- 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/d71249dc44ddccc12d8161011fa5955b1ce8fead_performance_tests/24 job id: 42401302 --- 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/d71249dc44ddccc12d8161011fa5955b1ce8fead_performance_tests/25 job id: 42401303 --- 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/d71249dc44ddccc12d8161011fa5955b1ce8fead_performance_tests/26 job id: 42401305 --- Point --- name: 510 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/1r/12t) --- Point --- name: 511 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ === END TESTS (description) === ===== PLOTS (description) ===== ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2_mem", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md_mem", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md_mem", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md_mem", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls_mem", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" === END PLOTS (description) === ============ RESULTS ============ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d71249dc44ddccc12d8161011fa5955b1ce8fead_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 344 9. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.024 0.034 132.829 132.829 farming_run 1 2.0 132.031 132.053 132.799 132.803 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.454428E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 228 1113141. MP_Allreduce 489 2263609. MP_Sync 27 MP_Alltoall 38 9316958. MP_SendRecv 30 829726. MP_ISendRecv 135 235435. MP_Wait 281 MP_comm_split 8 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.009 0.029 114.536 114.537 qs_energies 1 2.0 0.000 0.000 114.290 114.293 mp2_main 1 3.0 0.000 0.000 112.456 112.459 mp2_gpw_main 1 4.0 0.019 0.025 111.645 111.647 mp2_ri_gpw_compute_in 1 5.0 0.172 0.173 92.631 93.146 mp2_ri_gpw_compute_in_loop 1 6.0 0.005 0.005 55.478 55.993 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.175 41.785 47.228 get_2c_integrals 1 6.0 0.000 0.000 36.461 36.981 integrate_v_rspace 273 8.0 0.438 0.448 25.171 30.413 pw_transfer 6555 10.6 0.377 0.385 27.312 27.827 grid_integrate_task_list 273 9.0 20.983 26.726 20.983 26.726 fft_wrap_pw1pw2 5465 11.4 0.044 0.046 25.986 26.498 fft_wrap_pw1pw2_100 2178 12.4 1.149 1.226 23.547 24.062 compute_2c_integrals 1 7.0 0.002 0.002 19.169 19.170 rpa_ri_compute_en 1 5.0 0.000 0.000 18.904 19.109 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.820 18.901 mp2_eri_2c_integrate_gpw 1 9.0 2.388 2.429 18.818 18.898 cp_fm_cholesky_decompose 12 8.2 17.225 17.743 17.225 17.743 fft3d_s 5443 13.4 16.132 16.650 16.154 16.672 cholesky_decomp 1 7.0 0.000 0.000 16.142 16.668 ao_to_mo_and_store_B_mult_1 272 7.0 10.856 15.594 10.856 15.594 calculate_wavefunction 272 8.0 5.397 5.565 12.524 13.131 rpa_num_int 1 6.0 0.000 0.000 10.490 10.500 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.457 10.469 calc_mat_Q 8 8.0 0.000 0.000 9.327 9.419 contract_S_to_Q 8 9.0 0.000 0.000 8.748 8.841 calc_potential_gpw 544 9.5 0.005 0.006 8.257 8.632 parallel_gemm_fm 14 9.1 0.000 0.000 8.329 8.402 parallel_gemm_fm_cosma 14 10.1 8.329 8.402 8.329 8.402 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.002 0.002 8.190 8.394 potential_pw2rs 545 10.0 0.108 0.110 7.666 8.366 create_integ_mat 1 6.0 0.014 0.027 7.805 7.814 collocate_single_gaussian 272 10.0 0.041 0.043 7.420 7.662 array2fm 1 7.0 0.000 0.000 6.659 7.252 pw_scatter_s 2720 13.7 4.439 4.634 4.439 4.634 pw_gather_s 2722 13.2 3.898 4.258 3.898 4.258 array2fm_buffer_send 1 8.0 2.916 3.092 2.916 3.092 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.647201, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2727.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d71249dc44ddccc12d8161011fa5955b1ce8fead_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 0.029 0.036 396.649 396.651 farming_run 1 2.0 395.981 395.992 396.607 396.610 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827141120 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788822 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.223533E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 703 408373. MP_Allreduce 1825 23678. MP_Sync 38 MP_Alltoall 77 MP_SendRecv 2171 2843495. MP_ISendRecv 1739 144022. MP_Wait 2051 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.041 210.075 210.076 qs_energies 1 2.0 0.000 0.000 209.847 209.859 scf_env_do_scf 1 3.0 0.000 0.000 106.716 106.717 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.824 105.831 rebuild_ks_matrix 4 6.0 0.000 0.000 105.822 105.829 qs_ks_build_kohn_sham_matrix 4 7.0 0.054 0.061 105.822 105.829 hfx_ks_matrix 4 8.0 0.001 0.001 105.447 105.451 integrate_four_center 4 9.0 0.143 0.454 105.446 105.451 mp2_main 1 3.0 0.000 0.000 102.848 102.860 mp2_gpw_main 1 4.0 0.032 0.046 101.881 101.895 integrate_four_center_main 4 10.0 0.113 0.536 96.669 99.008 integrate_four_center_bin 265 11.0 96.556 99.006 96.556 99.006 init_scf_loop 1 4.0 0.000 0.000 92.319 92.319 mp2_ri_gpw_compute_in 1 5.0 0.064 0.064 75.050 76.218 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.633 55.801 mp2_eri_3c_integrate_gpw 91 7.0 0.143 0.161 42.276 47.445 integrate_v_rspace 95 8.0 0.399 0.573 28.640 33.570 pw_transfer 2240 10.6 0.144 0.161 29.881 30.341 ao_to_mo_and_store_B_mult_1 91 7.0 10.664 30.086 10.664 30.086 fft_wrap_pw1pw2 1868 11.4 0.017 0.019 28.890 29.455 grid_integrate_task_list 95 9.0 23.931 29.056 23.931 29.056 mp2_ri_gpw_compute_en 1 5.0 0.057 0.069 26.683 28.592 fft_wrap_pw1pw2_100 730 12.4 1.269 1.439 26.616 27.218 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.831 1.893 24.874 24.883 get_2c_integrals 1 6.0 0.000 0.000 20.327 20.351 compute_2c_integrals 1 7.0 0.002 0.003 19.304 19.306 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 18.862 19.178 mp2_eri_2c_integrate_gpw 1 9.0 1.746 1.979 18.861 19.177 fft3d_s 1823 13.4 18.413 18.768 18.426 18.781 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.396 14.396 calculate_wavefunction 91 8.0 2.022 2.055 9.752 9.989 mp2_ri_gpw_compute_en_expansio 172 7.0 0.556 0.585 8.791 9.416 potential_pw2rs 186 10.0 0.034 0.035 8.594 9.190 local_gemm 172 8.0 8.235 8.840 8.235 8.840 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.233 8.520 mp2_ri_gpw_compute_en_comm 22 7.0 0.501 0.517 7.861 8.327 calc_potential_gpw 182 9.5 0.002 0.002 7.942 8.230 collocate_single_gaussian 91 10.0 0.017 0.025 7.873 8.080 mp_sync 38 10.4 3.387 7.058 3.387 7.058 mp2_ri_gpw_compute_en_ener 172 7.0 6.346 6.441 6.346 6.441 mp_sendrecv_dm3 2068 8.0 5.892 6.339 5.892 6.339 pw_gather_s 912 13.2 4.894 5.426 4.894 5.426 pw_scatter_s 910 13.7 3.951 4.223 3.951 4.223 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.872864, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1511.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d71249dc44ddccc12d8161011fa5955b1ce8fead_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 451.149824E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62385. MP_Allreduce 10249 271. MP_Sync 580 MP_Alltoall 2083 592243. MP_ISendRecv 45220 5520. MP_Wait 60486 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.054 0.096 52.191 52.192 qs_mol_dyn_low 1 2.0 0.003 0.008 51.841 51.848 qs_forces 11 3.9 0.002 0.003 50.468 50.469 qs_energies 11 4.9 0.003 0.014 49.000 49.007 scf_env_do_scf 11 5.9 0.000 0.001 43.156 43.156 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 41.042 41.042 dbcsr_multiply_generic 2286 12.5 0.093 0.099 32.751 33.145 qs_scf_new_mos 108 7.5 0.000 0.001 31.117 31.387 qs_scf_loop_do_ot 108 8.5 0.001 0.001 31.116 31.386 ot_scf_mini 108 9.5 0.002 0.003 29.482 29.695 multiply_cannon 2286 13.5 0.184 0.189 25.570 27.021 multiply_cannon_loop 2286 14.5 1.476 1.552 24.869 26.320 velocity_verlet 10 3.0 0.001 0.001 26.219 26.236 ot_mini 108 10.5 0.001 0.001 18.734 18.994 qs_ot_get_derivative 108 11.5 0.001 0.001 15.876 16.087 mp_waitall_1 267858 16.1 8.480 14.128 8.480 14.128 multiply_cannon_metrocomm3 54864 15.5 0.066 0.071 5.883 12.435 multiply_cannon_multrec 54864 15.5 4.228 6.576 7.471 10.890 rebuild_ks_matrix 119 8.3 0.000 0.000 7.847 7.994 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.013 7.846 7.994 multiply_cannon_sync_h2d 54864 15.5 5.889 7.048 5.889 7.048 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.914 7.044 mp_sum_l 7207 12.9 5.224 6.773 5.224 6.773 qs_ot_get_p 119 10.4 0.001 0.001 6.186 6.448 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.646 5.754 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.137 5.585 init_scf_run 11 5.9 0.000 0.001 4.599 4.600 scf_env_initial_rho_setup 11 6.9 0.000 0.000 4.599 4.599 sum_up_and_integrate 119 10.3 0.012 0.015 4.475 4.482 integrate_v_rspace 119 11.3 0.002 0.002 4.463 4.470 dbcsr_mm_accdrv_process 76910 16.1 1.165 1.817 3.166 4.416 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.018 4.143 calculate_rho_elec 119 8.7 0.011 0.016 4.017 4.143 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.414 3.461 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.859 3.043 apply_single 119 13.6 0.000 0.000 2.859 3.043 multiply_cannon_metrocomm1 54864 15.5 0.051 0.058 1.742 3.002 calculate_dm_sparse 119 9.5 0.000 0.000 2.849 2.958 rs_pw_transfer 974 11.9 0.012 0.013 2.732 2.798 cp_dbcsr_syevd 50 12.0 0.002 0.003 2.609 2.609 ot_diis_step 108 11.5 0.006 0.006 2.601 2.601 jit_kernel_multiply 13 15.8 1.938 2.524 1.938 2.524 calculate_first_density_matrix 1 7.0 0.002 0.012 2.453 2.459 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.386 2.438 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.404 2.405 cp_fm_redistribute_end 50 14.0 2.182 2.382 2.186 2.383 cp_fm_diag_elpa_base 50 14.0 0.195 2.348 0.196 2.357 density_rs2pw 119 9.7 0.004 0.004 2.158 2.249 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.008 2.240 2.243 grid_integrate_task_list 119 12.3 2.026 2.164 2.026 2.164 acc_transpose_blocks 54864 15.5 0.224 0.243 1.712 2.140 init_scf_loop 11 6.9 0.001 0.013 2.098 2.098 wfi_extrapolate 11 7.9 0.001 0.001 2.085 2.085 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.925 1.976 potential_pw2rs 119 12.3 0.004 0.004 1.823 1.834 mp_sum_d 4129 12.0 1.245 1.789 1.245 1.789 pw_transfer 1439 11.6 0.051 0.055 1.650 1.725 make_m2s 4572 13.5 0.053 0.055 1.660 1.709 fft_wrap_pw1pw2 1201 12.6 0.006 0.007 1.575 1.651 make_images 4572 14.5 0.133 0.138 1.579 1.626 mp_alltoall_d11v 2130 13.8 1.288 1.464 1.288 1.464 mp_sum_dm 438 4.9 1.389 1.441 1.389 1.441 fft3d_ps 1201 14.6 0.358 0.459 1.354 1.426 mp_waitany 12084 13.8 1.251 1.403 1.251 1.403 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.369 1.386 grid_collocate_task_list 119 9.7 1.289 1.371 1.289 1.371 update_particle_set 20 4.0 0.000 0.000 1.335 1.352 md_write_output 11 3.9 0.017 1.298 0.019 1.338 md_output 10 3.0 0.000 0.000 0.017 1.335 fft_wrap_pw1pw2_140 487 13.2 0.079 0.092 1.213 1.289 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=52.192000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=430.000000, yerr=1.414214 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d71249dc44ddccc12d8161011fa5955b1ce8fead_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 490.020864E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2194560 MPI messages size (bytes): total size 310.646604E+09 min size 0.000000E+00 max size 1.145520E+06 average size 141.553031E+03 MPI breakdown and total messages size (bytes): size <= 128 724648 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 281952 4619501568 32768 < size <= 131072 494448 39143342080 131072 < size <= 4194304 440000 264807943488 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62664. MP_Allreduce 10226 305. MP_Sync 104 MP_Alltoall 2060 1061678. MP_ISendRecv 33558 37093. MP_Wait 40318 MP_comm_split 50 MP_ISend 5720 128509. MP_IRecv 5720 128509. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.059 41.662 41.663 qs_mol_dyn_low 1 2.0 0.003 0.004 41.337 41.390 qs_forces 11 3.9 0.002 0.006 40.341 40.343 qs_energies 11 4.9 0.002 0.010 38.662 38.667 scf_env_do_scf 11 5.9 0.001 0.001 32.886 32.886 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 30.383 30.385 dbcsr_multiply_generic 2286 12.5 0.100 0.103 23.733 24.085 qs_scf_new_mos 108 7.5 0.001 0.001 21.550 21.793 qs_scf_loop_do_ot 108 8.5 0.001 0.001 21.549 21.792 ot_scf_mini 108 9.5 0.003 0.004 20.659 20.833 multiply_cannon 2286 13.5 0.208 0.217 18.800 20.446 multiply_cannon_loop 2286 14.5 0.901 0.984 17.693 19.018 velocity_verlet 10 3.0 0.001 0.002 18.938 18.958 ot_mini 108 10.5 0.001 0.001 13.592 13.832 mp_waitall_1 217478 16.2 6.824 13.617 6.824 13.617 multiply_cannon_metrocomm3 27432 15.5 0.068 0.070 4.981 11.927 qs_ot_get_derivative 108 11.5 0.001 0.002 11.133 11.309 multiply_cannon_multrec 27432 15.5 2.292 6.238 7.114 10.452 dbcsr_mm_accdrv_process 47894 16.0 3.458 7.203 4.753 7.508 rebuild_ks_matrix 119 8.3 0.000 0.000 7.065 7.205 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.020 7.065 7.204 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.232 6.358 init_scf_run 11 5.9 0.000 0.001 4.532 4.532 scf_env_initial_rho_setup 11 6.9 0.000 0.001 4.532 4.532 qs_ot_get_p 119 10.4 0.001 0.001 4.184 4.420 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.566 4.408 jit_kernel_multiply 13 16.1 1.243 4.303 1.243 4.303 apply_preconditioner_dbcsr 119 12.6 0.000 0.001 3.038 4.115 apply_single 119 13.6 0.000 0.000 3.038 4.114 sum_up_and_integrate 119 10.3 0.024 0.029 4.107 4.114 integrate_v_rspace 119 11.3 0.002 0.003 4.083 4.092 mp_sum_l 7207 12.9 2.031 3.917 2.031 3.917 qs_rho_update_rho_low 119 7.7 0.001 0.002 3.547 3.573 calculate_rho_elec 119 8.7 0.021 0.024 3.546 3.573 calculate_first_density_matrix 1 7.0 0.001 0.006 3.047 3.049 multiply_cannon_sync_h2d 27432 15.5 2.199 2.818 2.199 2.818 qs_ot_p2m_diag 50 11.0 0.009 0.013 2.728 2.747 make_m2s 4572 13.5 0.052 0.054 2.457 2.678 rs_pw_transfer 974 11.9 0.010 0.011 2.481 2.601 make_images 4572 14.5 0.201 0.238 2.369 2.588 init_scf_loop 11 6.9 0.003 0.020 2.481 2.483 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.464 2.467 ot_diis_step 108 11.5 0.011 0.017 2.408 2.408 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.269 2.309 cp_dbcsr_syevd 50 12.0 0.003 0.004 2.300 2.301 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.107 2.199 calculate_dm_sparse 119 9.5 0.000 0.000 2.067 2.139 acc_transpose_blocks 27432 15.5 0.108 0.112 1.561 2.061 density_rs2pw 119 9.7 0.004 0.004 1.911 2.005 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.978 1.979 grid_integrate_task_list 119 12.3 1.841 1.959 1.841 1.959 cp_fm_redistribute_end 50 14.0 1.629 1.955 1.633 1.956 cp_fm_diag_elpa_base 50 14.0 0.314 1.900 0.322 1.931 potential_pw2rs 119 12.3 0.006 0.007 1.854 1.862 pw_transfer 1439 11.6 0.063 0.067 1.748 1.783 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 1.659 1.696 prepare_preconditioner 11 7.9 0.000 0.000 1.565 1.595 make_preconditioner 11 8.9 0.000 0.002 1.565 1.595 make_images_data 4572 15.5 0.045 0.052 1.134 1.552 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.464 1.523 acc_transpose_blocks_kernels 27432 16.5 0.182 0.272 1.030 1.441 hybrid_alltoall_any 4725 16.4 0.051 0.111 0.981 1.440 wfi_extrapolate 11 7.9 0.001 0.003 1.435 1.435 fft3d_ps 1201 14.6 0.497 0.549 1.367 1.398 fft_wrap_pw1pw2_140 487 13.2 0.075 0.083 1.305 1.341 grid_collocate_task_list 119 9.7 1.240 1.327 1.240 1.327 mp_alltoall_d11v 2130 13.8 1.171 1.314 1.171 1.314 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.284 1.292 mp_allgather_i34 2286 14.5 0.552 1.289 0.552 1.289 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.169 1.218 jit_kernel_transpose 5 15.5 0.848 1.170 0.848 1.170 mp_sum_d 4129 12.0 0.593 1.041 0.593 1.041 mp_sum_dm 438 4.9 0.982 1.029 0.982 1.029 md_output 10 3.0 0.000 0.000 0.025 1.024 md_write_output 11 3.9 0.022 1.005 0.023 1.017 update_particle_set 20 4.0 0.000 0.000 0.955 0.976 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 0.968 0.969 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.890 0.903 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=41.663000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=464.545455, yerr=1.616035 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d71249dc44ddccc12d8161011fa5955b1ce8fead_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 523.382784E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62660. MP_Allreduce 10225 303. MP_Sync 104 MP_Alltoall 1821 1607811. MP_ISendRecv 22134 57667. MP_Wait 33054 MP_comm_split 50 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.029 0.059 32.560 32.562 qs_mol_dyn_low 1 2.0 0.003 0.007 32.123 32.131 qs_forces 11 3.9 0.008 0.018 32.061 32.065 qs_energies 11 4.9 0.004 0.014 30.500 30.510 scf_env_do_scf 11 5.9 0.001 0.005 25.788 25.789 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 23.159 23.160 dbcsr_multiply_generic 2286 12.5 0.094 0.096 16.612 16.729 velocity_verlet 10 3.0 0.001 0.001 15.226 15.227 qs_scf_new_mos 108 7.5 0.001 0.001 15.086 15.114 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.086 15.113 ot_scf_mini 108 9.5 0.002 0.003 14.354 14.371 multiply_cannon 2286 13.5 0.195 0.201 13.130 14.034 multiply_cannon_loop 2286 14.5 0.635 0.665 12.302 13.262 ot_mini 108 10.5 0.001 0.001 8.969 8.990 qs_ot_get_derivative 108 11.5 0.001 0.001 7.422 7.438 multiply_cannon_multrec 18288 15.5 1.927 2.897 6.740 7.124 rebuild_ks_matrix 119 8.3 0.000 0.000 6.465 6.487 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.016 6.464 6.487 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.727 5.748 dbcsr_mm_accdrv_process 38222 16.0 4.129 5.486 4.730 5.542 mp_waitall_1 169478 16.3 3.096 4.260 3.096 4.260 sum_up_and_integrate 119 10.3 0.029 0.030 3.994 4.000 integrate_v_rspace 119 11.3 0.002 0.003 3.965 3.974 init_scf_run 11 5.9 0.000 0.001 3.459 3.459 scf_env_initial_rho_setup 11 6.9 0.000 0.001 3.459 3.459 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.739 3.360 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.238 3.244 calculate_rho_elec 119 8.7 0.030 0.031 3.237 3.243 qs_ot_get_p 119 10.4 0.001 0.001 3.197 3.230 init_scf_loop 11 6.9 0.001 0.004 2.607 2.608 multiply_cannon_metrocomm3 18288 15.5 0.045 0.046 1.489 2.468 rs_pw_transfer 974 11.9 0.009 0.010 2.296 2.466 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.064 2.396 apply_single 119 13.6 0.000 0.000 2.063 2.395 calculate_first_density_matrix 1 7.0 0.001 0.005 2.232 2.233 qs_ot_p2m_diag 50 11.0 0.012 0.012 2.107 2.114 make_m2s 4572 13.5 0.044 0.045 1.905 2.038 jit_kernel_multiply 11 15.9 0.549 2.022 0.549 2.022 density_rs2pw 119 9.7 0.004 0.005 1.826 1.998 make_images 4572 14.5 0.191 0.203 1.820 1.953 grid_integrate_task_list 119 12.3 1.802 1.892 1.802 1.892 pw_transfer 1439 11.6 0.063 0.065 1.851 1.872 calculate_dm_sparse 119 9.5 0.000 0.000 1.822 1.833 cp_dbcsr_syevd 50 12.0 0.003 0.005 1.826 1.826 prepare_preconditioner 11 7.9 0.000 0.000 1.786 1.790 make_preconditioner 11 8.9 0.000 0.000 1.786 1.790 potential_pw2rs 119 12.3 0.007 0.008 1.780 1.787 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 1.761 1.785 make_full_inverse_cholesky 11 9.9 0.000 0.001 1.640 1.721 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.687 1.698 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.659 1.661 mp_sum_l 7207 12.9 1.214 1.607 1.214 1.607 cp_fm_diag_elpa 50 13.0 0.000 0.001 1.540 1.543 multiply_cannon_sync_h2d 18288 15.5 1.371 1.531 1.371 1.531 ot_diis_step 108 11.5 0.011 0.012 1.521 1.521 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.510 1.518 cp_fm_redistribute_end 50 14.0 1.143 1.514 1.145 1.515 fft3d_ps 1201 14.6 0.508 0.525 1.448 1.475 cp_fm_diag_elpa_base 50 14.0 0.352 1.427 0.366 1.466 fft_wrap_pw1pw2_140 487 13.2 0.086 0.090 1.400 1.420 grid_collocate_task_list 119 9.7 1.205 1.289 1.205 1.289 acc_transpose_blocks 18288 15.5 0.075 0.076 1.236 1.258 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.183 1.188 wfi_extrapolate 11 7.9 0.001 0.001 1.171 1.171 qs_energies_init_hamiltonians 11 5.9 0.002 0.010 1.010 1.017 make_images_data 4572 15.5 0.045 0.048 0.866 1.008 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.910 0.939 hybrid_alltoall_any 4725 16.4 0.055 0.114 0.740 0.918 mp_alltoall_z22v 1201 16.6 0.770 0.855 0.770 0.855 mp_alltoall_d11v 2130 13.8 0.717 0.833 0.717 0.833 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.808 0.810 acc_transpose_blocks_kernels 18288 16.5 0.210 0.219 0.797 0.809 cp_fm_cholesky_invert 11 10.9 0.748 0.752 0.748 0.752 rs_pw_transfer_RS2PW_140 130 11.5 0.120 0.132 0.553 0.734 mp_waitany 9880 13.7 0.561 0.728 0.561 0.728 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.652 0.713 mp_allgather_i34 2286 14.5 0.289 0.669 0.289 0.669 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=32.562000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=497.181818, yerr=2.587318 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d71249dc44ddccc12d8161011fa5955b1ce8fead_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 545.951744E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62659. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 2412273. MP_ISendRecv 16422 74133. MP_Wait 24482 MP_comm_split 50 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.020 0.037 35.741 35.742 qs_mol_dyn_low 1 2.0 0.003 0.004 35.374 35.381 qs_forces 11 3.9 0.004 0.011 35.312 35.314 qs_energies 11 4.9 0.002 0.004 33.604 33.611 scf_env_do_scf 11 5.9 0.001 0.001 28.457 28.459 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 24.853 24.854 dbcsr_multiply_generic 2286 12.5 0.102 0.106 18.494 18.593 velocity_verlet 10 3.0 0.001 0.001 18.033 18.034 qs_scf_new_mos 108 7.5 0.001 0.001 16.498 16.550 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.498 16.549 ot_scf_mini 108 9.5 0.002 0.003 15.545 15.594 multiply_cannon 2286 13.5 0.232 0.272 14.472 14.912 multiply_cannon_loop 2286 14.5 0.935 0.965 13.444 13.772 ot_mini 108 10.5 0.001 0.001 9.526 9.582 multiply_cannon_multrec 27432 15.5 2.329 2.990 8.600 8.951 qs_ot_get_derivative 108 11.5 0.001 0.001 7.693 7.743 dbcsr_mm_accdrv_process 47916 15.9 5.357 7.300 6.177 7.357 rebuild_ks_matrix 119 8.3 0.000 0.000 6.627 6.673 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 6.626 6.673 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.902 5.942 sum_up_and_integrate 119 10.3 0.034 0.037 3.857 3.865 integrate_v_rspace 119 11.3 0.002 0.003 3.823 3.831 init_scf_run 11 5.9 0.000 0.001 3.721 3.722 scf_env_initial_rho_setup 11 6.9 0.000 0.001 3.721 3.722 init_scf_loop 11 6.9 0.000 0.002 3.580 3.581 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.370 3.410 calculate_rho_elec 119 8.7 0.040 0.046 3.370 3.409 qs_ot_get_p 119 10.4 0.001 0.001 3.332 3.403 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.727 3.214 mp_waitall_1 145218 16.4 2.302 2.841 2.302 2.841 prepare_preconditioner 11 7.9 0.000 0.000 2.718 2.725 make_preconditioner 11 8.9 0.000 0.001 2.718 2.725 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.307 2.643 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.147 2.598 apply_single 119 13.6 0.000 0.000 2.147 2.598 make_m2s 4572 13.5 0.054 0.056 2.414 2.490 make_images 4572 14.5 0.274 0.333 2.306 2.380 calculate_first_density_matrix 1 7.0 0.000 0.003 2.319 2.321 rs_pw_transfer 974 11.9 0.009 0.009 2.125 2.280 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.115 2.143 calculate_dm_sparse 119 9.5 0.000 0.000 2.066 2.124 density_rs2pw 119 9.7 0.004 0.004 1.846 1.990 qs_ot_p2m_diag 50 11.0 0.015 0.023 1.977 1.987 grid_integrate_task_list 119 12.3 1.820 1.912 1.820 1.912 pw_transfer 1439 11.6 0.063 0.067 1.819 1.851 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.841 1.843 jit_kernel_multiply 10 15.9 0.758 1.813 0.758 1.813 ot_diis_step 108 11.5 0.012 0.012 1.793 1.793 mp_sum_l 7207 12.9 1.189 1.793 1.189 1.793 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 1.730 1.764 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.655 1.668 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.651 1.652 potential_pw2rs 119 12.3 0.008 0.010 1.609 1.614 acc_transpose_blocks 27432 15.5 0.112 0.114 1.455 1.470 fft3d_ps 1201 14.6 0.534 0.586 1.411 1.435 fft_wrap_pw1pw2_140 487 13.2 0.084 0.093 1.389 1.427 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.378 1.379 cp_fm_redistribute_end 50 14.0 0.908 1.353 0.909 1.354 wfi_extrapolate 11 7.9 0.001 0.001 1.353 1.354 cp_fm_diag_elpa_base 50 14.0 0.423 1.288 0.442 1.329 grid_collocate_task_list 119 9.7 1.222 1.303 1.222 1.303 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.756 1.292 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.234 1.246 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.157 1.174 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.149 1.153 cp_fm_upper_to_full 72 13.5 0.830 1.152 0.830 1.152 dbcsr_complete_redistribute 329 12.2 0.117 0.146 0.827 1.106 make_images_data 4572 15.5 0.045 0.049 0.966 1.082 multiply_cannon_sync_h2d 27432 15.5 0.983 1.030 0.983 1.030 hybrid_alltoall_any 4725 16.4 0.062 0.150 0.825 0.995 mp_alltoall_d11v 2130 13.8 0.842 0.982 0.842 0.982 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.605 0.875 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.791 0.869 acc_transpose_blocks_kernels 27432 16.5 0.269 0.276 0.828 0.839 cp_fm_cholesky_invert 11 10.9 0.829 0.832 0.829 0.832 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.816 0.820 mp_alltoall_z22v 1201 16.6 0.759 0.785 0.759 0.785 mp_alltoall_i22 627 13.8 0.451 0.757 0.451 0.757 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=35.742000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=518.000000, yerr=3.717282 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d71249dc44ddccc12d8161011fa5955b1ce8fead_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 594.948096E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62658. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 3682667. MP_ISendRecv 10710 94533. MP_Wait 16690 MP_comm_split 50 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.027 0.044 27.992 27.992 qs_mol_dyn_low 1 2.0 0.003 0.007 27.712 27.719 qs_forces 11 3.9 0.002 0.003 27.656 27.656 qs_energies 11 4.9 0.001 0.001 25.985 25.988 scf_env_do_scf 11 5.9 0.000 0.001 20.874 20.875 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 18.451 18.451 velocity_verlet 10 3.0 0.001 0.001 13.861 13.864 dbcsr_multiply_generic 2286 12.5 0.094 0.096 12.376 12.515 qs_scf_new_mos 108 7.5 0.001 0.001 10.900 10.920 qs_scf_loop_do_ot 108 8.5 0.001 0.001 10.900 10.919 multiply_cannon 2286 13.5 0.228 0.235 9.914 10.401 ot_scf_mini 108 9.5 0.002 0.002 10.231 10.255 multiply_cannon_loop 2286 14.5 0.331 0.342 9.016 9.273 multiply_cannon_multrec 9144 15.5 1.652 1.903 6.014 6.210 ot_mini 108 10.5 0.001 0.001 5.840 5.870 rebuild_ks_matrix 119 8.3 0.000 0.000 5.702 5.725 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.701 5.724 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.086 5.107 qs_ot_get_derivative 108 11.5 0.001 0.001 4.556 4.580 dbcsr_mm_accdrv_process 12550 15.8 3.066 4.086 4.262 4.355 init_scf_run 11 5.9 0.000 0.001 3.702 3.702 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.702 3.702 sum_up_and_integrate 119 10.3 0.037 0.041 3.485 3.490 integrate_v_rspace 119 11.3 0.003 0.003 3.448 3.453 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.117 3.123 calculate_rho_elec 119 8.7 0.059 0.061 3.116 3.123 qs_ot_get_p 119 10.4 0.001 0.001 2.627 2.667 calculate_first_density_matrix 1 7.0 0.000 0.000 2.619 2.619 init_scf_loop 11 6.9 0.000 0.000 2.405 2.405 mp_waitall_1 121218 16.5 1.619 2.132 1.619 2.132 jit_kernel_multiply 10 15.8 1.157 1.978 1.157 1.978 grid_integrate_task_list 119 12.3 1.849 1.932 1.849 1.932 make_m2s 4572 13.5 0.034 0.035 1.695 1.860 make_images 4572 14.5 0.268 0.303 1.606 1.769 calculate_dm_sparse 119 9.5 0.000 0.000 1.745 1.763 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.723 1.724 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.711 1.714 prepare_preconditioner 11 7.9 0.000 0.000 1.695 1.700 make_preconditioner 11 8.9 0.000 0.000 1.695 1.700 pw_transfer 1439 11.6 0.063 0.065 1.671 1.678 rs_pw_transfer 974 11.9 0.008 0.008 1.545 1.640 density_rs2pw 119 9.7 0.004 0.004 1.533 1.628 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.584 1.614 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 1.581 1.587 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.563 1.574 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.524 1.537 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.517 1.518 grid_collocate_task_list 119 9.7 1.272 1.339 1.272 1.339 fft_wrap_pw1pw2_140 487 13.2 0.082 0.086 1.292 1.300 potential_pw2rs 119 12.3 0.010 0.010 1.278 1.280 ot_diis_step 108 11.5 0.012 0.013 1.270 1.270 fft3d_ps 1201 14.6 0.535 0.548 1.247 1.255 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.252 1.252 cp_fm_redistribute_end 50 14.0 0.621 1.234 0.622 1.234 cp_fm_diag_elpa_base 50 14.0 0.569 1.167 0.611 1.223 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.204 1.204 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.168 1.175 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.143 1.168 apply_single 119 13.6 0.000 0.000 1.143 1.167 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.111 1.123 wfi_extrapolate 11 7.9 0.001 0.001 1.039 1.039 hybrid_alltoall_any 4725 16.4 0.062 0.176 0.742 0.993 make_images_data 4572 15.5 0.039 0.043 0.781 0.980 acc_transpose_blocks 9144 15.5 0.039 0.040 0.955 0.961 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.866 0.918 mp_alltoall_d11v 2130 13.8 0.752 0.884 0.752 0.884 cp_fm_cholesky_invert 11 10.9 0.851 0.854 0.851 0.854 multiply_cannon_sync_h2d 9144 15.5 0.711 0.788 0.711 0.788 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.737 0.739 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.717 0.726 acc_transpose_blocks_kernels 9144 16.5 0.117 0.120 0.713 0.717 qs_env_update_s_mstruct 11 6.9 0.002 0.003 0.662 0.711 multiply_cannon_metrocomm3 9144 15.5 0.018 0.019 0.414 0.676 mp_allgather_i34 2286 14.5 0.229 0.620 0.229 0.620 mp_alltoall_z22v 1201 16.6 0.586 0.619 0.586 0.619 jit_kernel_transpose 5 15.6 0.596 0.601 0.596 0.601 parallel_gemm_fm 81 9.0 0.000 0.000 0.593 0.594 parallel_gemm_fm_cosma 81 10.0 0.593 0.594 0.593 0.594 qs_create_task_list 11 7.9 0.000 0.000 0.541 0.567 generate_qs_task_list 11 8.9 0.190 0.212 0.541 0.567 make_basis_sm 11 9.8 0.001 0.001 0.564 0.564 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=27.992000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=563.454545, yerr=6.243013 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d71249dc44ddccc12d8161011fa5955b1ce8fead_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 754.462720E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63729. MP_Allreduce 10074 433. MP_Sync 54 MP_Alltoall 1582 7383731. MP_ISendRecv 4998 189067. MP_Wait 8898 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.028 0.047 41.484 41.485 qs_mol_dyn_low 1 2.0 0.003 0.007 41.267 41.274 qs_forces 11 3.9 0.002 0.002 39.591 39.592 qs_energies 11 4.9 0.001 0.001 37.673 37.676 scf_env_do_scf 11 5.9 0.000 0.001 31.760 31.760 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 24.115 24.116 velocity_verlet 10 3.0 0.001 0.001 23.834 24.041 dbcsr_multiply_generic 2286 12.5 0.101 0.102 17.362 17.584 qs_scf_new_mos 108 7.5 0.001 0.001 15.425 15.523 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.425 15.522 multiply_cannon 2286 13.5 0.306 0.312 13.575 14.507 ot_scf_mini 108 9.5 0.002 0.002 14.359 14.458 multiply_cannon_loop 2286 14.5 0.343 0.349 12.333 13.318 ot_mini 108 10.5 0.001 0.001 8.642 8.758 multiply_cannon_multrec 9144 15.5 3.321 4.665 8.669 8.752 init_scf_loop 11 6.9 0.000 0.000 7.619 7.621 rebuild_ks_matrix 119 8.3 0.000 0.000 6.717 6.857 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 6.716 6.856 qs_ot_get_derivative 108 11.5 0.001 0.001 6.646 6.745 prepare_preconditioner 11 7.9 0.000 0.000 6.712 6.725 make_preconditioner 11 8.9 0.000 0.000 6.712 6.725 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.271 6.606 dbcsr_mm_accdrv_process 12550 15.8 4.246 5.659 5.227 6.509 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.075 6.203 cp_fm_upper_to_full 72 14.2 3.148 4.538 3.148 4.538 init_scf_run 11 5.9 0.000 0.001 3.913 3.913 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.913 3.913 sum_up_and_integrate 119 10.3 0.064 0.065 3.601 3.607 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.565 3.569 calculate_rho_elec 119 8.7 0.118 0.121 3.564 3.569 integrate_v_rspace 119 11.3 0.003 0.003 3.537 3.544 mp_waitall_1 97218 16.6 2.327 3.333 2.327 3.333 qs_ot_get_p 119 10.4 0.001 0.001 3.029 3.162 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.455 2.867 dbcsr_complete_redistribute 329 12.2 0.289 0.296 1.994 2.788 calculate_first_density_matrix 1 7.0 0.000 0.000 2.541 2.542 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.688 2.483 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.156 2.431 apply_single 119 13.6 0.000 0.000 2.156 2.431 make_m2s 4572 13.5 0.038 0.038 2.192 2.336 mp_alltoall_i22 627 13.8 1.462 2.286 1.462 2.286 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.436 2.232 make_images 4572 14.5 0.355 0.386 2.071 2.215 calculate_dm_sparse 119 9.5 0.000 0.000 2.172 2.189 multiply_cannon_metrocomm3 9144 15.5 0.019 0.020 1.222 2.133 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.056 2.056 grid_integrate_task_list 119 12.3 1.995 2.004 1.995 2.004 ot_diis_step 108 11.5 0.014 0.015 1.975 1.975 pw_transfer 1439 11.6 0.066 0.067 1.940 1.944 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.845 1.849 qs_ot_p2m_diag 50 11.0 0.042 0.043 1.809 1.811 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.749 1.779 mp_sum_l 7207 12.9 1.052 1.777 1.052 1.777 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.742 1.743 density_rs2pw 119 9.7 0.003 0.004 1.694 1.706 mp_sum_dm 438 4.9 1.472 1.693 1.472 1.693 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.637 1.686 md_write_output 11 3.9 0.207 1.649 0.209 1.662 md_output 10 3.0 0.000 0.000 0.210 1.658 update_particle_set 20 4.0 0.000 0.000 1.450 1.658 fft_wrap_pw1pw2_140 487 13.2 0.087 0.089 1.560 1.564 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.544 1.545 fft3d_ps 1201 14.6 0.565 0.577 1.480 1.483 grid_collocate_task_list 119 9.7 1.437 1.450 1.437 1.450 cp_fm_cholesky_invert 11 10.9 1.378 1.381 1.378 1.381 rs_pw_transfer 974 11.9 0.009 0.009 1.326 1.353 wfi_extrapolate 11 7.9 0.001 0.001 1.318 1.318 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.273 1.274 cp_fm_diag_elpa_base 50 14.0 1.131 1.183 1.272 1.272 hybrid_alltoall_any 4725 16.4 0.087 0.146 1.067 1.267 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.251 1.257 make_images_data 4572 15.5 0.043 0.046 1.003 1.190 potential_pw2rs 119 12.3 0.014 0.015 1.160 1.164 jit_kernel_multiply 6 15.7 0.953 1.139 0.953 1.139 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.114 1.132 mp_alltoall_d11v 2130 13.8 1.094 1.110 1.094 1.110 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.080 1.095 multiply_cannon_sync_h2d 9144 15.5 1.040 1.046 1.040 1.046 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.979 1.032 qs_create_task_list 11 7.9 0.000 0.000 0.936 0.947 generate_qs_task_list 11 8.9 0.373 0.391 0.936 0.947 acc_transpose_blocks 9144 15.5 0.039 0.039 0.898 0.902 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.852 0.865 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=41.485000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=702.545455, yerr=16.422621 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d71249dc44ddccc12d8161011fa5955b1ce8fead_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 500.981760E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65372. MP_Allreduce 9840 486. MP_Sync 100 MP_Alltoall 1938 327211. MP_ISendRecv 41800 9096. MP_Wait 58168 MP_comm_split 48 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.034 0.058 83.971 83.973 qs_mol_dyn_low 1 2.0 0.004 0.011 83.529 83.538 qs_forces 11 3.9 0.003 0.005 83.446 83.447 qs_energies 11 4.9 0.002 0.006 80.602 80.612 scf_env_do_scf 11 5.9 0.000 0.001 71.665 71.667 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 65.813 65.813 dbcsr_multiply_generic 2055 12.4 0.105 0.107 52.687 52.975 qs_scf_new_mos 99 7.5 0.000 0.001 48.633 48.741 qs_scf_loop_do_ot 99 8.5 0.001 0.001 48.632 48.741 ot_scf_mini 99 9.5 0.002 0.003 46.211 46.301 multiply_cannon 2055 13.4 0.178 0.183 42.716 43.614 velocity_verlet 10 3.0 0.001 0.002 42.589 42.590 multiply_cannon_loop 2055 14.4 1.552 1.589 41.691 42.370 ot_mini 99 10.5 0.001 0.001 28.074 28.173 qs_ot_get_derivative 99 11.5 0.001 0.002 21.193 21.308 multiply_cannon_multrec 49320 15.4 12.071 12.812 17.122 17.860 rebuild_ks_matrix 110 8.3 0.000 0.001 14.714 14.862 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.016 14.714 14.861 mp_waitall_1 241148 16.1 12.169 13.236 12.169 13.236 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.933 13.061 multiply_cannon_sync_h2d 49320 15.4 9.913 10.740 9.913 10.740 qs_ot_get_p 110 10.4 0.001 0.001 9.482 9.569 multiply_cannon_metrocomm3 49320 15.4 0.075 0.080 7.018 8.129 apply_preconditioner_dbcsr 110 12.6 0.000 0.001 7.197 7.760 apply_single 110 13.6 0.000 0.000 7.197 7.760 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.191 7.728 sum_up_and_integrate 110 10.3 0.037 0.045 7.243 7.257 integrate_v_rspace 110 11.3 0.003 0.003 7.206 7.229 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.795 6.850 init_scf_run 11 5.9 0.000 0.001 6.784 6.784 scf_env_initial_rho_setup 11 6.9 0.000 0.002 6.784 6.784 ot_diis_step 99 11.5 0.006 0.008 6.572 6.574 qs_rho_update_rho_low 110 7.6 0.001 0.002 6.246 6.388 calculate_rho_elec 110 8.6 0.020 0.025 6.246 6.387 qs_ot_p2m_diag 48 11.0 0.012 0.018 6.164 6.188 mp_sum_l 6514 12.8 5.371 6.075 5.371 6.075 init_scf_loop 11 6.9 0.001 0.004 5.824 5.826 dbcsr_mm_accdrv_process 87628 16.1 2.107 2.200 4.930 5.235 cp_dbcsr_syevd 48 12.0 0.002 0.002 5.112 5.112 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.623 4.625 cp_fm_redistribute_end 48 14.0 4.020 4.593 4.024 4.594 cp_fm_diag_elpa_base 48 14.0 0.564 4.515 0.567 4.540 make_m2s 4110 13.4 0.060 0.066 4.182 4.310 make_images 4110 14.4 0.178 0.195 4.087 4.218 wfi_extrapolate 11 7.9 0.001 0.004 4.090 4.091 rs_pw_transfer 902 11.9 0.012 0.013 3.858 4.063 calculate_dm_sparse 110 9.5 0.000 0.001 3.789 3.879 multiply_cannon_metrocomm1 49320 15.4 0.058 0.062 2.840 3.845 prepare_preconditioner 11 7.9 0.000 0.000 3.610 3.625 make_preconditioner 11 8.9 0.000 0.001 3.610 3.625 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.580 3.584 density_rs2pw 110 9.6 0.004 0.004 3.289 3.476 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.379 3.416 grid_integrate_task_list 110 12.3 3.246 3.415 3.246 3.415 make_full_inverse_cholesky 11 9.9 0.000 0.001 3.355 3.401 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.200 3.261 pw_transfer 1331 11.6 0.053 0.062 2.995 3.059 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 2.909 2.976 potential_pw2rs 110 12.3 0.006 0.007 2.754 2.775 mp_alltoall_d11v 2046 13.8 2.361 2.735 2.361 2.735 calculate_first_density_matrix 1 7.0 0.000 0.003 2.601 2.606 jit_kernel_multiply 13 15.9 2.535 2.557 2.535 2.557 fft_wrap_pw1pw2_140 451 13.1 0.169 0.188 2.443 2.513 fft3d_ps 1111 14.6 0.752 0.834 2.446 2.503 acc_transpose_blocks 49320 15.4 0.216 0.226 2.170 2.237 grid_collocate_task_list 110 9.6 2.083 2.201 2.083 2.201 mp_waitany 14300 13.8 1.903 2.164 1.903 2.164 mp_sum_d 3883 11.9 1.544 2.090 1.544 2.090 make_images_data 4110 15.4 0.043 0.046 1.936 2.081 cp_fm_cholesky_invert 11 10.9 1.972 1.977 1.972 1.977 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.926 1.945 hybrid_alltoall_any 4261 16.3 0.082 0.484 1.674 1.926 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=83.973000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=475.818182, yerr=2.289032 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d71249dc44ddccc12d8161011fa5955b1ce8fead_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 587.743232E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1972800 MPI messages size (bytes): total size 1.077520E+12 min size 0.000000E+00 max size 4.537280E+06 average size 546.188250E+03 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 222984 1826684928 8192 < size <= 32768 520356 13399818240 32768 < size <= 131072 372336 35386294272 131072 < size <= 4194304 787758 788321309808 4194304 < size <= 16777216 54450 238588003280 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65587. MP_Allreduce 9839 562. MP_Sync 100 MP_Alltoall 1717 3599700. MP_ISendRecv 20680 26400. MP_Wait 32692 MP_comm_split 48 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.083 0.201 71.253 71.255 qs_mol_dyn_low 1 2.0 0.003 0.005 70.402 70.410 qs_forces 11 3.9 0.017 0.045 70.062 70.069 qs_energies 11 4.9 0.002 0.006 66.595 66.604 scf_env_do_scf 11 5.9 0.001 0.001 57.988 57.992 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 50.232 50.233 dbcsr_multiply_generic 2055 12.4 0.114 0.117 39.542 39.687 velocity_verlet 10 3.0 0.001 0.002 36.307 36.315 qs_scf_new_mos 99 7.5 0.001 0.001 34.331 34.481 qs_scf_loop_do_ot 99 8.5 0.001 0.001 34.331 34.481 multiply_cannon 2055 13.4 0.221 0.244 32.692 34.030 ot_scf_mini 99 9.5 0.011 0.038 32.656 32.788 multiply_cannon_loop 2055 14.4 0.932 0.953 31.393 32.421 ot_mini 99 10.5 0.001 0.001 20.004 20.142 multiply_cannon_multrec 24660 15.4 7.632 9.458 14.170 16.621 qs_ot_get_derivative 99 11.5 0.001 0.005 14.202 14.330 rebuild_ks_matrix 110 8.3 0.000 0.001 13.813 13.912 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.019 13.812 13.911 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.066 12.158 mp_waitall_1 186928 16.3 9.225 11.602 9.225 11.602 multiply_cannon_metrocomm3 24660 15.4 0.070 0.072 6.514 9.579 apply_preconditioner_dbcsr 110 12.6 0.000 0.001 6.580 9.056 apply_single 110 13.6 0.000 0.001 6.580 9.056 multiply_cannon_sync_h2d 24660 15.4 7.037 8.210 7.037 8.210 dbcsr_mm_accdrv_process 52282 16.1 4.712 6.144 6.378 8.049 init_scf_loop 11 6.9 0.001 0.004 7.722 7.723 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 6.416 7.187 sum_up_and_integrate 110 10.3 0.052 0.060 6.662 6.709 integrate_v_rspace 110 11.3 0.002 0.002 6.610 6.656 qs_ot_get_p 110 10.4 0.001 0.002 6.019 6.193 init_scf_run 11 5.9 0.000 0.001 6.164 6.165 scf_env_initial_rho_setup 11 6.9 0.000 0.002 6.164 6.165 qs_rho_update_rho_low 110 7.6 0.001 0.002 5.794 5.805 calculate_rho_elec 110 8.6 0.039 0.047 5.793 5.804 ot_diis_step 99 11.5 0.010 0.012 5.754 5.755 prepare_preconditioner 11 7.9 0.000 0.000 5.660 5.680 make_preconditioner 11 8.9 0.000 0.002 5.660 5.680 make_full_inverse_cholesky 11 9.9 0.000 0.001 5.206 5.367 make_m2s 4110 13.4 0.057 0.061 4.312 4.791 make_images 4110 14.4 0.402 0.448 4.203 4.679 mp_sum_l 6514 12.8 1.999 4.211 1.999 4.211 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.125 4.145 cp_dbcsr_syevd 48 12.0 0.003 0.004 3.678 3.678 pw_transfer 1331 11.6 0.065 0.072 3.454 3.627 wfi_extrapolate 11 7.9 0.001 0.004 3.531 3.531 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.348 3.520 grid_integrate_task_list 110 12.3 3.154 3.340 3.154 3.340 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.309 3.312 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.189 3.264 density_rs2pw 110 9.6 0.004 0.004 3.053 3.246 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.141 3.143 rs_pw_transfer 902 11.9 0.012 0.014 2.935 3.139 jit_kernel_multiply 10 16.3 1.310 3.125 1.310 3.125 cp_fm_redistribute_end 48 14.0 2.342 3.117 2.344 3.117 cp_fm_diag_elpa_base 48 14.0 0.739 2.997 0.770 3.083 fft_wrap_pw1pw2_140 451 13.1 0.201 0.219 2.845 3.017 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.962 3.011 calculate_dm_sparse 110 9.5 0.001 0.001 2.973 3.001 make_images_data 4110 15.4 0.047 0.051 2.406 2.909 hybrid_alltoall_any 4261 16.3 0.101 0.441 2.129 2.899 fft3d_ps 1111 14.6 1.072 1.273 2.705 2.860 cp_fm_cholesky_invert 11 10.9 2.586 2.593 2.586 2.593 calculate_first_density_matrix 1 7.0 0.001 0.003 2.545 2.548 potential_pw2rs 110 12.3 0.008 0.009 2.488 2.533 grid_collocate_task_list 110 9.6 2.118 2.269 2.118 2.269 mp_alltoall_d11v 2046 13.8 1.923 2.153 1.923 2.153 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.957 1.974 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.911 1.921 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.878 1.891 multiply_cannon_metrocomm4 22605 15.4 0.073 0.079 0.771 1.758 mp_irecv_dv 57340 16.2 0.646 1.643 0.646 1.643 mp_allgather_i34 2055 14.4 0.633 1.642 0.633 1.642 acc_transpose_blocks 24660 15.4 0.107 0.110 1.524 1.549 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.534 1.545 mp_alltoall_z22v 1111 16.6 1.379 1.488 1.379 1.488 dbcsr_complete_redistribute 325 12.2 0.238 0.299 1.223 1.482 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.368 1.480 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=71.255000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=556.181818, yerr=6.978124 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d71249dc44ddccc12d8161011fa5955b1ce8fead_performance_tests/11/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 404.681598E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 3346752 0.0% 0.0% 100.0% average stack size 0.0 0.0 294.1 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 660.135936E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65578. MP_Allreduce 9838 559. MP_Sync 100 MP_Alltoall 1496 4511006. MP_ISendRecv 13640 27424. MP_Wait 32318 MP_comm_split 48 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.022 0.032 61.536 61.537 qs_mol_dyn_low 1 2.0 0.003 0.005 61.228 61.236 qs_forces 11 3.9 0.004 0.011 59.412 59.413 qs_energies 11 4.9 0.002 0.006 56.270 56.275 scf_env_do_scf 11 5.9 0.000 0.001 48.380 48.381 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 39.736 39.736 velocity_verlet 10 3.0 0.001 0.001 34.081 34.137 dbcsr_multiply_generic 2055 12.4 0.108 0.112 28.463 28.745 qs_scf_new_mos 99 7.5 0.001 0.001 25.301 25.406 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.301 25.405 ot_scf_mini 99 9.5 0.002 0.003 24.057 24.183 multiply_cannon 2055 13.4 0.212 0.224 22.125 23.443 multiply_cannon_loop 2055 14.4 0.618 0.636 20.951 21.940 ot_mini 99 10.5 0.001 0.001 13.660 13.791 rebuild_ks_matrix 110 8.3 0.000 0.000 12.101 12.258 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.101 12.258 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.688 10.823 mp_waitall_1 146766 16.3 7.318 10.585 7.318 10.585 multiply_cannon_multrec 16440 15.4 3.787 4.725 9.552 10.449 qs_ot_get_derivative 99 11.5 0.001 0.001 9.144 9.271 init_scf_loop 11 6.9 0.001 0.003 8.608 8.608 multiply_cannon_metrocomm3 16440 15.4 0.043 0.045 4.524 7.617 prepare_preconditioner 11 7.9 0.000 0.000 6.842 6.863 make_preconditioner 11 8.9 0.000 0.000 6.842 6.863 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.154 6.497 sum_up_and_integrate 110 10.3 0.059 0.061 6.375 6.388 integrate_v_rspace 110 11.3 0.003 0.003 6.315 6.329 dbcsr_mm_accdrv_process 34862 16.1 4.681 5.265 5.621 5.781 qs_ot_get_p 110 10.4 0.001 0.001 5.592 5.762 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.522 5.530 calculate_rho_elec 110 8.6 0.058 0.058 5.521 5.529 init_scf_run 11 5.9 0.000 0.001 5.465 5.466 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.465 5.465 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.015 5.445 apply_single 110 13.6 0.000 0.000 5.015 5.445 make_m2s 4110 13.4 0.050 0.051 4.166 4.510 ot_diis_step 99 11.5 0.010 0.011 4.481 4.481 make_images 4110 14.4 0.393 0.510 4.052 4.396 multiply_cannon_sync_h2d 16440 15.4 3.639 4.326 3.639 4.326 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.930 3.936 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.104 3.742 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.580 3.580 grid_integrate_task_list 110 12.3 3.187 3.382 3.187 3.382 pw_transfer 1331 11.6 0.064 0.072 3.187 3.196 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.080 3.093 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.039 3.040 cp_fm_redistribute_end 48 14.0 1.887 3.012 1.889 3.012 density_rs2pw 110 9.6 0.004 0.004 2.816 3.003 cp_fm_diag_elpa_base 48 14.0 1.060 2.874 1.118 2.987 wfi_extrapolate 11 7.9 0.001 0.001 2.965 2.965 rs_pw_transfer 902 11.9 0.010 0.011 2.695 2.889 hybrid_alltoall_any 4261 16.3 0.105 0.372 2.168 2.814 make_images_data 4110 15.4 0.044 0.048 2.392 2.811 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.759 2.761 cp_fm_cholesky_invert 11 10.9 2.727 2.733 2.727 2.733 fft_wrap_pw1pw2_140 451 13.1 0.210 0.216 2.613 2.626 calculate_dm_sparse 110 9.5 0.001 0.001 2.543 2.573 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.431 2.489 multiply_cannon_metrocomm4 14385 15.4 0.045 0.049 0.917 2.488 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.370 2.441 calculate_first_density_matrix 1 7.0 0.003 0.025 2.421 2.422 fft3d_ps 1111 14.6 1.064 1.076 2.394 2.407 mp_irecv_dv 48980 15.7 0.846 2.363 0.846 2.363 grid_collocate_task_list 110 9.6 2.179 2.299 2.179 2.299 potential_pw2rs 110 12.3 0.010 0.010 2.213 2.221 mp_sum_l 6514 12.8 1.554 2.167 1.554 2.167 mp_alltoall_d11v 2046 13.8 1.804 2.072 1.804 2.072 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 1.987 1.989 jit_kernel_multiply 10 16.3 0.546 1.922 0.546 1.922 dbcsr_complete_redistribute 325 12.2 0.324 0.359 1.458 1.903 mp_sum_dm 438 4.9 1.787 1.860 1.787 1.860 md_output 10 3.0 0.000 0.000 0.060 1.803 md_write_output 11 3.9 0.056 1.785 0.058 1.803 update_particle_set 20 4.0 0.000 0.000 1.743 1.799 cp_fm_upper_to_full 70 13.6 1.353 1.781 1.353 1.781 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.615 1.638 cp_fm_cholesky_decompose 22 10.9 1.614 1.634 1.614 1.634 mp_allgather_i34 2055 14.4 0.500 1.517 0.500 1.517 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.484 1.500 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.343 1.457 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.000 1.427 mp_waitany 17072 13.8 1.175 1.352 1.175 1.352 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.285 1.292 acc_transpose_blocks 16440 15.4 0.073 0.078 1.212 1.235 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=61.537000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=623.454545, yerr=7.738495 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d71249dc44ddccc12d8161011fa5955b1ce8fead_performance_tests/12/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 601.317074E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 4916280 0.0% 0.0% 100.0% average stack size 0.0 0.0 200.2 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 737.505280E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65576. MP_Allreduce 9838 600. MP_Sync 100 MP_Alltoall 1496 5863162. MP_ISendRecv 10120 43184. MP_Wait 25102 MP_comm_split 48 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.035 0.049 64.838 64.839 qs_mol_dyn_low 1 2.0 0.003 0.003 64.557 64.566 qs_forces 11 3.9 0.008 0.010 63.261 63.262 qs_energies 11 4.9 0.001 0.002 59.920 59.928 scf_env_do_scf 11 5.9 0.000 0.001 51.849 51.852 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 40.168 40.169 velocity_verlet 10 3.0 0.001 0.001 37.482 37.535 dbcsr_multiply_generic 2055 12.4 0.116 0.119 29.128 29.314 qs_scf_new_mos 99 7.5 0.001 0.001 25.955 26.050 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.955 26.049 ot_scf_mini 99 9.5 0.003 0.003 24.331 24.430 multiply_cannon 2055 13.4 0.242 0.260 22.279 23.207 multiply_cannon_loop 2055 14.4 0.884 0.906 20.935 21.494 ot_mini 99 10.5 0.001 0.001 13.889 14.005 multiply_cannon_multrec 24660 15.4 4.197 6.901 12.635 13.786 rebuild_ks_matrix 110 8.3 0.000 0.000 11.784 11.900 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.783 11.899 init_scf_loop 11 6.9 0.000 0.000 11.640 11.641 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.429 10.531 prepare_preconditioner 11 7.9 0.000 0.000 9.955 9.968 make_preconditioner 11 8.9 0.000 0.000 9.955 9.968 qs_ot_get_derivative 99 11.5 0.001 0.001 9.749 9.851 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.148 9.637 dbcsr_mm_accdrv_process 52304 16.0 6.851 8.309 8.291 9.272 sum_up_and_integrate 110 10.3 0.067 0.070 6.259 6.275 integrate_v_rspace 110 11.3 0.003 0.003 6.192 6.205 mp_waitall_1 126806 16.4 4.280 6.072 4.280 6.072 qs_ot_get_p 110 10.4 0.001 0.001 5.480 5.619 make_m2s 4110 13.4 0.059 0.061 5.283 5.581 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.517 5.524 calculate_rho_elec 110 8.6 0.077 0.081 5.516 5.524 make_images 4110 14.4 0.578 0.696 5.142 5.438 init_scf_run 11 5.9 0.000 0.001 5.365 5.365 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.364 5.365 cp_fm_upper_to_full 70 13.8 3.382 4.809 3.382 4.809 ot_diis_step 99 11.5 0.011 0.011 4.101 4.101 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.021 4.100 apply_single 110 13.6 0.000 0.000 4.020 4.100 dbcsr_complete_redistribute 325 12.2 0.422 0.469 2.645 3.800 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.747 3.761 grid_integrate_task_list 110 12.3 3.278 3.420 3.278 3.420 multiply_cannon_sync_h2d 24660 15.4 3.196 3.354 3.196 3.354 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.147 3.297 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.236 3.289 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.271 3.272 multiply_cannon_metrocomm3 24660 15.4 0.035 0.036 1.391 3.216 pw_transfer 1331 11.6 0.064 0.071 3.180 3.204 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.075 3.101 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.984 3.044 hybrid_alltoall_any 4261 16.3 0.119 0.457 2.315 2.965 make_images_data 4110 15.4 0.046 0.050 2.623 2.945 calculate_dm_sparse 110 9.5 0.001 0.001 2.908 2.938 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.798 2.926 wfi_extrapolate 11 7.9 0.001 0.001 2.908 2.908 mp_alltoall_i22 605 13.7 1.673 2.882 1.673 2.882 density_rs2pw 110 9.6 0.004 0.004 2.692 2.856 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.768 2.770 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.743 2.745 cp_fm_redistribute_end 48 14.0 1.365 2.719 1.366 2.720 cp_fm_diag_elpa_base 48 14.0 1.271 2.592 1.351 2.703 cp_fm_cholesky_invert 11 10.9 2.677 2.685 2.677 2.685 fft_wrap_pw1pw2_140 451 13.1 0.202 0.212 2.615 2.644 rs_pw_transfer 902 11.9 0.010 0.011 2.353 2.530 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.420 2.453 fft3d_ps 1111 14.6 1.063 1.093 2.378 2.396 grid_collocate_task_list 110 9.6 2.218 2.392 2.218 2.392 calculate_first_density_matrix 1 7.0 0.000 0.000 2.372 2.374 qs_energies_init_hamiltonians 11 5.9 0.006 0.010 2.225 2.232 jit_kernel_multiply 12 15.7 1.106 2.178 1.106 2.178 potential_pw2rs 110 12.3 0.013 0.013 2.060 2.069 mp_alltoall_d11v 2046 13.8 1.736 1.947 1.736 1.947 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.695 1.727 cp_fm_cholesky_decompose 22 10.9 1.640 1.701 1.640 1.701 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.599 1.700 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.599 1.612 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.519 1.534 multiply_cannon_metrocomm4 20550 15.4 0.057 0.060 0.829 1.530 acc_transpose_blocks 24660 15.4 0.105 0.108 1.499 1.518 mp_irecv_dv 62702 16.1 0.731 1.452 0.731 1.452 mp_sum_l 6514 12.8 0.877 1.446 0.877 1.446 mp_allgather_i34 2055 14.4 0.448 1.370 0.448 1.370 mp_sum_dm 438 4.9 1.258 1.331 1.258 1.331 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=64.839000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=695.636364, yerr=10.394690 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d71249dc44ddccc12d8161011fa5955b1ce8fead_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 830.791680E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65574. MP_Allreduce 9838 640. MP_Sync 100 MP_Alltoall 1496 8504061. MP_ISendRecv 6600 54848. MP_Wait 17226 MP_comm_split 48 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.029 54.956 54.957 qs_mol_dyn_low 1 2.0 0.003 0.005 54.626 54.635 qs_forces 11 3.9 0.005 0.006 54.561 54.562 qs_energies 11 4.9 0.001 0.002 50.952 50.957 scf_env_do_scf 11 5.9 0.000 0.001 42.050 42.050 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 34.323 34.324 velocity_verlet 10 3.0 0.001 0.001 30.456 30.459 dbcsr_multiply_generic 2055 12.4 0.110 0.114 22.822 23.026 qs_scf_new_mos 99 7.5 0.001 0.001 20.410 20.471 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.409 20.471 ot_scf_mini 99 9.5 0.002 0.002 19.180 19.232 multiply_cannon 2055 13.4 0.246 0.263 17.330 18.607 multiply_cannon_loop 2055 14.4 0.323 0.337 15.981 16.377 rebuild_ks_matrix 110 8.3 0.000 0.000 11.331 11.372 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.330 11.372 ot_mini 99 10.5 0.001 0.001 10.438 10.488 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.088 10.123 multiply_cannon_multrec 8220 15.4 3.187 4.864 7.403 8.656 mp_waitall_1 106626 16.5 6.157 7.737 6.157 7.737 init_scf_loop 11 6.9 0.000 0.000 7.680 7.682 qs_ot_get_derivative 99 11.5 0.001 0.001 6.553 6.605 sum_up_and_integrate 110 10.3 0.079 0.081 6.109 6.120 prepare_preconditioner 11 7.9 0.000 0.000 6.069 6.076 make_preconditioner 11 8.9 0.000 0.000 6.069 6.076 integrate_v_rspace 110 11.3 0.003 0.003 6.029 6.040 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.658 5.729 init_scf_run 11 5.9 0.000 0.001 5.702 5.702 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.701 5.701 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.521 5.537 calculate_rho_elec 110 8.6 0.115 0.115 5.520 5.536 dbcsr_mm_accdrv_process 17442 15.9 2.803 3.936 4.084 4.974 qs_ot_get_p 110 10.4 0.001 0.001 4.855 4.914 multiply_cannon_metrocomm3 8220 15.4 0.017 0.017 3.175 4.495 make_m2s 4110 13.4 0.039 0.040 4.201 4.475 make_images 4110 14.4 0.638 0.693 4.071 4.341 ot_diis_step 99 11.5 0.012 0.012 3.864 3.864 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.773 3.834 apply_single 110 13.6 0.000 0.000 3.773 3.834 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.518 3.521 grid_integrate_task_list 110 12.3 3.363 3.509 3.363 3.509 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.220 3.220 pw_transfer 1331 11.6 0.065 0.069 3.187 3.193 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.081 3.091 multiply_cannon_sync_h2d 8220 15.4 2.901 3.038 2.901 3.038 cp_fm_cholesky_invert 11 10.9 2.975 2.979 2.975 2.979 calculate_first_density_matrix 1 7.0 0.000 0.000 2.950 2.951 make_images_data 4110 15.4 0.038 0.044 2.345 2.746 qs_energies_init_hamiltonians 11 5.9 0.007 0.010 2.732 2.732 density_rs2pw 110 9.6 0.004 0.004 2.592 2.711 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.709 2.709 cp_fm_redistribute_end 48 14.0 0.683 2.685 0.687 2.686 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.677 2.679 wfi_extrapolate 11 7.9 0.001 0.001 2.667 2.667 cp_fm_diag_elpa_base 48 14.0 1.818 2.499 1.992 2.660 fft_wrap_pw1pw2_140 451 13.1 0.212 0.215 2.646 2.658 hybrid_alltoall_any 4261 16.3 0.199 0.862 2.262 2.619 grid_collocate_task_list 110 9.6 2.324 2.511 2.324 2.511 calculate_dm_sparse 110 9.5 0.001 0.001 2.460 2.494 fft3d_ps 1111 14.6 1.112 1.125 2.340 2.350 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.252 2.270 rs_pw_transfer 902 11.9 0.010 0.010 2.106 2.241 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.770 2.000 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.882 1.918 potential_pw2rs 110 12.3 0.015 0.016 1.866 1.871 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.766 1.790 mp_alltoall_d11v 2046 13.8 1.591 1.783 1.591 1.783 cp_fm_cholesky_decompose 22 10.9 1.660 1.671 1.660 1.671 qs_env_update_s_mstruct 11 6.9 0.018 0.021 1.524 1.639 dbcsr_complete_redistribute 325 12.2 0.617 0.655 1.499 1.580 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.573 1.580 mp_allgather_i34 2055 14.4 0.524 1.574 0.524 1.574 jit_kernel_multiply 9 15.8 0.969 1.515 0.969 1.515 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.431 1.443 multiply_cannon_metrocomm4 6165 15.4 0.017 0.019 0.484 1.372 mp_irecv_dv 24056 15.7 0.459 1.329 0.459 1.329 qs_create_task_list 11 7.9 0.000 0.000 1.221 1.319 generate_qs_task_list 11 8.9 0.379 0.447 1.221 1.319 make_basis_sm 11 9.8 0.009 0.011 1.192 1.194 mp_waitany 9240 13.8 1.031 1.165 1.031 1.165 parallel_gemm_fm 81 9.0 0.000 0.000 1.139 1.142 parallel_gemm_fm_cosma 81 10.0 1.139 1.142 1.139 1.142 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.113 1.139 multiply_cannon_metrocomm1 8220 15.4 0.021 0.022 0.719 1.107 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=54.957000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=780.636364, yerr=10.092956 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d71249dc44ddccc12d8161011fa5955b1ce8fead_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 1.403441E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67104. MP_Allreduce 9672 819. MP_Sync 52 MP_Alltoall 1474 16505187. MP_ISendRecv 4620 360267. MP_Wait 7524 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.020 0.050 86.714 86.714 qs_mol_dyn_low 1 2.0 0.003 0.003 86.405 86.413 qs_forces 11 3.9 0.002 0.003 86.343 86.344 qs_energies 11 4.9 0.001 0.001 82.317 82.319 scf_env_do_scf 11 5.9 0.001 0.001 71.946 71.946 velocity_verlet 10 3.0 0.001 0.001 54.271 54.277 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 43.771 43.773 dbcsr_multiply_generic 2055 12.4 0.120 0.122 30.770 30.803 init_scf_loop 11 6.9 0.000 0.000 28.100 28.103 qs_scf_new_mos 99 7.5 0.001 0.001 27.551 27.576 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.550 27.576 prepare_preconditioner 11 7.9 0.000 0.000 26.175 26.182 make_preconditioner 11 8.9 0.000 0.000 26.175 26.182 ot_scf_mini 99 9.5 0.002 0.002 25.783 25.807 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.444 25.633 multiply_cannon 2055 13.4 0.346 0.362 23.562 24.508 multiply_cannon_loop 2055 14.4 0.342 0.344 21.802 22.493 cp_fm_upper_to_full 70 14.2 12.763 18.194 12.763 18.194 ot_mini 99 10.5 0.001 0.001 15.168 15.181 rebuild_ks_matrix 110 8.3 0.001 0.001 13.036 13.071 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 13.036 13.070 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.857 11.889 qs_ot_get_derivative 99 11.5 0.001 0.001 10.648 10.671 dbcsr_complete_redistribute 325 12.2 1.032 1.064 7.285 10.407 mp_waitall_1 87304 16.6 8.623 10.219 8.623 10.219 multiply_cannon_multrec 8220 15.4 4.342 4.529 9.595 9.720 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.274 9.400 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.716 8.814 mp_alltoall_i22 605 13.7 5.316 8.460 5.316 8.460 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 5.576 6.968 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.868 6.943 apply_single 110 13.6 0.000 0.000 4.868 6.943 sum_up_and_integrate 110 10.3 0.151 0.152 6.545 6.561 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.384 6.417 calculate_rho_elec 110 8.6 0.227 0.227 6.383 6.417 integrate_v_rspace 110 11.3 0.004 0.004 6.394 6.409 init_scf_run 11 5.9 0.000 0.001 6.152 6.152 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.152 6.152 make_m2s 4110 13.4 0.043 0.043 5.308 5.807 make_images 4110 14.4 0.876 0.913 5.120 5.617 qs_ot_get_p 110 10.4 0.001 0.001 5.406 5.427 dbcsr_mm_accdrv_process 11614 15.7 3.320 3.659 5.109 5.348 cp_fm_cholesky_invert 11 10.9 5.181 5.185 5.181 5.185 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.427 5.102 ot_diis_step 99 11.5 0.015 0.015 4.497 4.497 multiply_cannon_sync_h2d 8220 15.4 3.946 3.948 3.946 3.948 qs_ot_p2m_diag 48 11.0 0.151 0.156 3.910 3.917 hybrid_alltoall_any 4261 16.3 0.256 0.552 2.970 3.735 grid_integrate_task_list 110 12.3 3.663 3.722 3.663 3.722 pw_transfer 1331 11.6 0.073 0.074 3.669 3.675 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.631 3.633 make_images_data 4110 15.4 0.042 0.045 2.962 3.575 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 3.553 3.560 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.472 3.472 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.427 3.431 acc_transpose_blocks 8220 15.4 0.035 0.036 2.166 3.349 calculate_dm_sparse 110 9.5 0.001 0.001 3.174 3.199 wfi_extrapolate 11 7.9 0.001 0.001 3.192 3.192 acc_transpose_blocks_kernels 8220 16.4 0.110 0.110 1.928 3.111 fft_wrap_pw1pw2_140 451 13.1 0.214 0.216 3.094 3.101 mp_sum_l 6514 12.8 1.022 3.066 1.022 3.066 jit_kernel_transpose 5 15.6 1.818 3.002 1.818 3.002 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.932 2.943 density_rs2pw 110 9.6 0.004 0.004 2.887 2.904 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.886 2.886 cp_fm_diag_elpa_base 48 14.0 2.358 2.553 2.885 2.885 calculate_first_density_matrix 1 7.0 0.000 0.000 2.859 2.859 fft3d_ps 1111 14.6 1.267 1.279 2.764 2.773 grid_collocate_task_list 110 9.6 2.626 2.648 2.626 2.648 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.338 2.350 qs_env_update_s_mstruct 11 6.9 0.001 0.001 2.158 2.197 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.080 2.192 rs_pw_transfer 902 11.9 0.010 0.011 2.037 2.086 mp_alltoall_d11v 2046 13.8 2.025 2.066 2.025 2.066 potential_pw2rs 110 12.3 0.022 0.022 1.988 1.992 cp_fm_cholesky_decompose 22 10.9 1.967 1.984 1.967 1.984 qs_create_task_list 11 7.9 0.000 0.000 1.898 1.944 generate_qs_task_list 11 8.9 0.739 0.794 1.898 1.943 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.782 1.786 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.751 1.780 jit_kernel_multiply 10 15.4 1.586 1.772 1.586 1.772 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=86.714000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1250.727273, yerr=69.695560 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d71249dc44ddccc12d8161011fa5955b1ce8fead_performance_tests/15/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 1.094965E+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 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 630.575104E+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 3175954870160 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4077 56870. MP_Allreduce 11233 786. MP_Sync 170 MP_Alltoall 2226 1851561. MP_ISendRecv 48640 18752. MP_Wait 66796 MP_comm_split 83 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.023 0.043 203.784 203.788 qs_mol_dyn_low 1 2.0 0.003 0.003 203.363 203.377 qs_forces 11 3.9 0.003 0.004 203.281 203.282 qs_energies 11 4.9 0.016 0.077 197.824 197.838 scf_env_do_scf 11 5.9 0.001 0.001 181.379 181.383 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 160.856 160.858 dbcsr_multiply_generic 2507 12.6 0.176 0.180 125.399 125.899 qs_scf_new_mos 117 7.6 0.001 0.001 122.354 122.768 qs_scf_loop_do_ot 117 8.6 0.005 0.038 122.353 122.768 velocity_verlet 10 3.0 0.001 0.001 122.198 122.199 ot_scf_mini 117 9.6 0.003 0.004 115.763 116.155 multiply_cannon 2507 13.6 0.237 0.245 102.036 103.762 multiply_cannon_loop 2507 14.6 2.095 2.162 99.896 101.719 ot_mini 117 10.6 0.001 0.001 67.029 67.415 multiply_cannon_multrec 60168 15.6 33.154 35.788 42.517 45.197 qs_ot_get_derivative 117 11.6 0.001 0.001 42.237 42.603 rebuild_ks_matrix 128 8.3 0.001 0.001 33.071 33.387 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.018 33.070 33.387 mp_waitall_1 291448 16.2 28.865 31.864 28.865 31.864 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.774 30.045 multiply_cannon_sync_h2d 60168 15.6 27.331 29.368 27.331 29.368 qs_ot_get_p 128 10.4 0.001 0.001 26.900 27.299 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.260 24.991 apply_single 128 13.6 0.001 0.001 24.260 24.991 ot_diis_step 117 11.6 0.007 0.008 24.494 24.495 init_scf_loop 11 6.9 0.001 0.010 20.446 20.448 qs_ot_p2m_diag 83 11.4 0.077 0.091 20.120 20.227 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 18.919 19.205 multiply_cannon_metrocomm3 60168 15.6 0.109 0.114 15.782 18.597 cp_dbcsr_syevd 83 12.4 0.004 0.005 17.540 17.541 prepare_preconditioner 11 7.9 0.000 0.000 15.886 15.922 make_preconditioner 11 8.9 0.000 0.002 15.886 15.921 make_full_inverse_cholesky 11 9.9 0.000 0.001 15.123 15.276 cp_fm_diag_elpa 83 13.4 0.000 0.001 14.714 14.718 cp_fm_redistribute_end 83 14.4 11.597 14.643 11.607 14.646 cp_fm_diag_elpa_base 83 14.4 2.998 14.427 3.029 14.547 make_m2s 5014 13.6 0.103 0.111 13.799 14.146 make_images 5014 14.6 0.403 0.422 13.619 13.973 sum_up_and_integrate 128 10.3 0.089 0.105 13.861 13.878 integrate_v_rspace 128 11.3 0.003 0.004 13.772 13.791 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.548 12.668 calculate_rho_elec 128 8.7 0.045 0.063 12.547 12.668 init_scf_run 11 5.9 0.000 0.001 12.224 12.224 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.223 12.224 mp_sum_l 7870 13.0 8.326 10.169 8.326 10.169 dbcsr_mm_accdrv_process 124484 16.2 3.328 3.483 8.925 10.005 cp_fm_cholesky_invert 11 10.9 9.084 9.092 9.084 9.092 wfi_extrapolate 11 7.9 0.001 0.001 9.016 9.016 multiply_cannon_metrocomm1 60168 15.6 0.085 0.090 6.226 8.733 calculate_dm_sparse 128 9.5 0.001 0.001 8.383 8.452 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.954 8.119 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.799 7.870 make_images_data 5014 15.6 0.069 0.075 6.788 7.687 grid_integrate_task_list 128 12.3 7.101 7.499 7.101 7.499 hybrid_alltoall_any 5200 16.5 0.290 2.266 5.940 7.205 pw_transfer 1547 11.6 0.074 0.107 6.724 6.994 density_rs2pw 128 9.7 0.006 0.007 6.348 6.892 fft_wrap_pw1pw2 1291 12.7 0.009 0.013 6.521 6.761 cp_dbcsr_sm_fm_multiply 37 9.5 0.003 0.003 6.649 6.657 rs_pw_transfer 1046 11.9 0.017 0.019 5.559 6.139 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.816 5.899 fft_wrap_pw1pw2_140 523 13.2 0.445 0.504 5.637 5.809 mp_alltoall_d11v 2415 14.1 4.146 5.732 4.146 5.732 fft3d_ps 1291 14.7 2.095 2.582 5.315 5.499 grid_collocate_task_list 128 9.7 4.722 5.083 4.722 5.083 cp_fm_cholesky_decompose 22 10.9 4.629 4.642 4.629 4.642 potential_pw2rs 128 12.3 0.009 0.010 4.375 4.394 mp_sum_d 4460 12.1 3.601 4.391 3.601 4.391 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=203.788000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=595.000000, yerr=6.993503 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d71249dc44ddccc12d8161011fa5955b1ce8fead_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 2.199914E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 6022464 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.2 marketing flops 145.647559E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 827.129856E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2425920 MPI messages size (bytes): total size 4.132350E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703416E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 71436 2336489472 32768 < size <= 131072 728832 55956209664 131072 < size <= 4194304 1386864 1409906900992 4194304 < size <= 16777216 155760 1473826772352 16777216 < size 68112 1190343475200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4107 56904. MP_Allreduce 11307 944. MP_Sync 170 MP_Alltoall 1983 5036734. MP_ISendRecv 24252 47072. MP_Wait 38240 MP_comm_split 83 MP_ISend 11836 212447. MP_IRecv 11836 212447. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.030 193.596 193.598 qs_mol_dyn_low 1 2.0 0.003 0.003 193.200 193.224 qs_forces 11 3.9 0.004 0.007 193.120 193.121 qs_energies 11 4.9 0.001 0.002 186.436 186.450 scf_env_do_scf 11 5.9 0.001 0.001 169.507 169.517 scf_env_do_scf_inner_loop 118 6.6 0.003 0.010 136.459 136.462 velocity_verlet 10 3.0 0.001 0.001 120.645 120.647 dbcsr_multiply_generic 2527 12.6 0.186 0.192 100.203 101.500 qs_scf_new_mos 118 7.6 0.001 0.001 97.458 97.920 qs_scf_loop_do_ot 118 8.6 0.001 0.001 97.457 97.919 ot_scf_mini 118 9.6 0.004 0.004 92.609 93.134 multiply_cannon 2527 13.6 0.480 0.534 79.810 84.752 multiply_cannon_loop 2527 14.6 1.269 1.310 76.484 79.115 ot_mini 118 10.6 0.001 0.001 51.754 52.206 mp_waitall_1 228564 16.4 26.213 39.933 26.213 39.933 multiply_cannon_multrec 30324 15.6 22.295 26.559 32.566 37.060 rebuild_ks_matrix 129 8.3 0.001 0.001 33.037 33.538 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.019 33.037 33.538 init_scf_loop 11 6.9 0.000 0.000 32.953 32.955 qs_ks_update_qs_env 129 7.6 0.001 0.001 29.817 30.287 qs_ot_get_derivative 118 11.6 0.001 0.002 29.541 30.055 multiply_cannon_metrocomm3 30324 15.6 0.095 0.100 16.716 29.134 prepare_preconditioner 11 7.9 0.000 0.000 28.581 28.639 make_preconditioner 11 8.9 0.000 0.000 28.581 28.639 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.237 27.773 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 22.350 23.473 apply_single 129 13.6 0.001 0.001 22.350 23.473 qs_ot_get_p 129 10.4 0.001 0.001 22.414 22.946 multiply_cannon_sync_h2d 30324 15.6 19.476 22.857 19.476 22.857 ot_diis_step 118 11.6 0.014 0.015 22.036 22.038 qs_ot_p2m_diag 83 11.4 0.187 0.216 17.603 17.641 cp_fm_cholesky_invert 11 10.9 16.743 16.756 16.743 16.756 cp_dbcsr_syevd 83 12.4 0.005 0.005 16.444 16.445 make_m2s 5054 13.6 0.090 0.095 14.561 16.112 make_images 5054 14.6 1.170 1.362 14.351 15.904 sum_up_and_integrate 129 10.3 0.114 0.134 14.506 14.532 integrate_v_rspace 129 11.3 0.003 0.004 14.392 14.423 cp_fm_diag_elpa 83 13.4 0.000 0.001 13.275 13.285 cp_fm_redistribute_end 83 14.4 7.717 13.205 7.728 13.205 cp_fm_diag_elpa_base 83 14.4 5.240 12.739 5.457 13.108 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.053 13.086 calculate_rho_elec 129 8.7 0.088 0.106 13.052 13.085 init_scf_run 11 5.9 0.000 0.001 12.172 12.173 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.172 12.173 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 11.030 11.381 multiply_cannon_metrocomm4 27797 15.6 0.096 0.113 3.735 10.569 make_images_data 5054 15.6 0.067 0.074 8.877 10.562 dbcsr_mm_accdrv_process 62734 16.2 4.615 5.279 9.731 10.335 mp_irecv_dv 70031 16.3 3.540 10.183 3.540 10.183 hybrid_alltoall_any 5240 16.5 0.344 1.529 7.622 9.801 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 7.674 8.426 wfi_extrapolate 11 7.9 0.001 0.001 8.333 8.333 pw_transfer 1559 11.6 0.085 0.099 7.827 7.895 grid_integrate_task_list 129 12.3 7.202 7.677 7.202 7.677 fft_wrap_pw1pw2 1301 12.7 0.010 0.010 7.603 7.669 density_rs2pw 129 9.7 0.006 0.007 6.815 7.492 cp_fm_cholesky_decompose 22 10.9 6.900 6.982 6.900 6.982 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.694 6.703 fft_wrap_pw1pw2_140 527 13.2 0.471 0.520 6.611 6.677 calculate_dm_sparse 129 9.5 0.001 0.001 6.511 6.673 rs_pw_transfer 1054 12.0 0.014 0.016 5.709 6.448 fft3d_ps 1301 14.7 2.799 2.948 5.963 5.994 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.850 5.969 mp_sum_l 7930 13.1 4.046 5.642 4.046 5.642 qs_ot_get_orbitals 118 10.6 0.001 0.001 5.400 5.464 grid_collocate_task_list 129 9.7 4.945 5.331 4.945 5.331 mp_allgather_i34 2527 14.6 1.903 5.237 1.903 5.237 mp_alltoall_d11v 2423 14.1 4.229 4.935 4.229 4.935 potential_pw2rs 129 12.3 0.015 0.016 4.912 4.935 dbcsr_complete_redistribute 395 12.7 0.774 0.858 3.228 4.070 mp_sum_d 4493 12.2 2.609 4.052 2.609 4.052 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=193.598000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=787.181818, yerr=4.549089 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d71249dc44ddccc12d8161011fa5955b1ce8fead_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420241154048 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528896499712 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514757E+12 0.0% 0.0% 100.0% flops max/rank 2.928533E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755939872 0.0% 0.0% 100.0% number of processed stacks 3984192 0.0% 0.0% 100.0% average stack size 0.0 0.0 1695.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 941.658112E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1042912 MPI messages size (bytes): total size 2.716210E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.604448E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 281856 36943429632 131072 < size <= 4194304 660064 996105256960 4194304 < size <= 16777216 65632 931531083728 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 57335. MP_Allreduce 11226 986. MP_Sync 170 MP_Alltoall 1712 9388896. MP_ISendRecv 15872 75008. MP_Wait 29756 MP_comm_split 83 MP_ISend 11748 275205. MP_IRecv 11748 275205. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.047 0.102 176.998 177.001 qs_mol_dyn_low 1 2.0 0.003 0.007 176.373 176.386 qs_forces 11 3.9 0.003 0.007 176.280 176.284 qs_energies 11 4.9 0.003 0.007 169.813 169.822 scf_env_do_scf 11 5.9 0.001 0.001 153.766 153.767 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 118.777 118.778 velocity_verlet 10 3.0 0.001 0.001 112.033 112.035 dbcsr_multiply_generic 2507 12.6 0.182 0.186 83.289 84.474 qs_scf_new_mos 117 7.6 0.001 0.001 82.534 82.882 qs_scf_loop_do_ot 117 8.6 0.001 0.002 82.533 82.881 ot_scf_mini 117 9.6 0.004 0.005 78.417 78.790 multiply_cannon 2507 13.6 0.500 0.521 63.427 68.507 multiply_cannon_loop 2507 14.6 0.860 0.886 60.279 62.871 ot_mini 117 10.6 0.001 0.001 43.677 44.039 mp_waitall_1 178456 16.5 26.224 35.876 26.224 35.876 init_scf_loop 11 6.9 0.001 0.006 34.888 34.890 prepare_preconditioner 11 7.9 0.000 0.000 30.889 30.933 make_preconditioner 11 8.9 0.000 0.002 30.889 30.933 rebuild_ks_matrix 128 8.3 0.001 0.001 29.926 30.475 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.018 29.926 30.475 make_full_inverse_cholesky 11 9.9 0.000 0.001 28.541 29.894 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.039 27.545 multiply_cannon_metrocomm3 20056 15.6 0.058 0.061 16.020 26.019 multiply_cannon_multrec 20056 15.6 13.562 16.294 22.754 25.605 qs_ot_get_derivative 117 11.6 0.001 0.002 24.002 24.359 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.756 20.872 apply_single 128 13.6 0.001 0.001 19.755 20.871 qs_ot_get_p 128 10.4 0.001 0.001 20.077 20.523 ot_diis_step 117 11.6 0.017 0.018 19.572 19.573 qs_ot_p2m_diag 83 11.4 0.265 0.272 15.840 15.852 make_m2s 5014 13.6 0.081 0.086 14.806 15.762 multiply_cannon_sync_h2d 20056 15.6 14.238 15.737 14.238 15.737 make_images 5014 14.6 1.178 1.278 14.573 15.526 cp_dbcsr_syevd 83 12.4 0.005 0.005 14.836 14.838 cp_fm_cholesky_invert 11 10.9 14.443 14.452 14.443 14.452 sum_up_and_integrate 128 10.3 0.131 0.143 14.000 14.024 integrate_v_rspace 128 11.3 0.003 0.004 13.869 13.897 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.834 12.880 calculate_rho_elec 128 8.7 0.132 0.147 12.833 12.880 cp_fm_diag_elpa 83 13.4 0.000 0.001 11.715 11.715 cp_fm_redistribute_end 83 14.4 4.400 11.656 4.414 11.659 cp_fm_diag_elpa_base 83 14.4 6.817 11.087 7.226 11.575 init_scf_run 11 5.9 0.000 0.001 11.058 11.058 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.058 11.058 make_images_data 5014 15.6 0.060 0.070 9.262 10.494 hybrid_alltoall_any 5200 16.5 0.433 1.968 8.012 9.814 multiply_cannon_metrocomm4 17549 15.6 0.063 0.072 3.428 9.287 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 8.920 9.178 mp_irecv_dv 50230 16.2 3.303 9.036 3.303 9.036 dbcsr_mm_accdrv_process 41502 16.2 4.312 5.361 8.641 8.851 pw_transfer 1547 11.6 0.084 0.105 7.583 7.687 grid_integrate_task_list 128 12.3 7.308 7.672 7.308 7.672 cp_fm_cholesky_decompose 22 10.9 7.446 7.482 7.446 7.482 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.360 7.469 cp_fm_upper_to_full 105 14.5 5.613 7.393 5.613 7.393 wfi_extrapolate 11 7.9 0.001 0.001 7.349 7.349 density_rs2pw 128 9.7 0.006 0.006 6.356 6.811 fft_wrap_pw1pw2_140 523 13.2 0.479 0.537 6.466 6.584 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 5.798 6.509 dbcsr_complete_redistribute 395 12.7 1.169 1.204 4.562 6.299 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.019 6.024 calculate_dm_sparse 128 9.5 0.001 0.001 5.727 5.800 fft3d_ps 1291 14.7 2.694 2.908 5.629 5.714 rs_pw_transfer 1046 11.9 0.014 0.014 4.999 5.483 grid_collocate_task_list 128 9.7 5.066 5.398 5.066 5.398 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.115 5.284 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.387 5.119 mp_allgather_i34 2507 14.6 1.683 4.763 1.683 4.763 mp_alltoall_d11v 2415 14.1 4.157 4.663 4.157 4.663 mp_sum_l 7870 13.0 3.279 4.523 3.279 4.523 potential_pw2rs 128 12.3 0.021 0.023 4.414 4.429 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.024 4.046 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.329 4.009 mp_alltoall_i22 716 14.1 1.934 3.834 1.934 3.834 qs_energies_init_hamiltonians 11 5.9 0.001 0.011 3.745 3.747 calculate_first_density_matrix 1 7.0 0.000 0.002 3.536 3.540 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=177.001000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=894.545455, yerr=14.086510 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d71249dc44ddccc12d8161011fa5955b1ce8fead_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410023282688 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444707676160 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796579E+12 0.0% 0.0% 100.0% flops max/rank 4.320339E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500928 0.0% 0.0% 100.0% number of processed stacks 5927808 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.2 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.165255E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1133160 MPI messages size (bytes): total size 2.008142E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.772161E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 315952 35695099904 131072 < size <= 4194304 709496 778939400192 4194304 < size <= 16777216 69840 660837789680 16777216 < size 30480 532676608000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4079 57274. MP_Allreduce 11236 1068. MP_Sync 168 MP_Alltoall 1700 12496381. MP_ISendRecv 11684 75008. MP_Wait 28114 MP_comm_split 82 MP_ISend 14840 244848. MP_IRecv 14840 244848. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.048 187.705 187.706 qs_mol_dyn_low 1 2.0 0.003 0.004 187.006 187.019 qs_forces 11 3.9 0.005 0.006 186.893 186.898 qs_energies 11 4.9 0.002 0.009 179.919 179.929 scf_env_do_scf 11 5.9 0.001 0.001 163.115 163.124 velocity_verlet 10 3.0 0.001 0.001 122.743 122.745 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 116.223 116.225 qs_scf_new_mos 116 7.6 0.001 0.001 81.094 81.323 qs_scf_loop_do_ot 116 8.6 0.001 0.002 81.093 81.323 dbcsr_multiply_generic 2485 12.5 0.188 0.192 80.041 80.758 ot_scf_mini 116 9.6 0.003 0.005 76.622 76.898 multiply_cannon 2485 13.5 0.550 0.582 55.681 58.576 multiply_cannon_loop 2485 14.5 1.171 1.199 52.079 53.847 init_scf_loop 11 6.9 0.001 0.004 46.768 46.769 ot_mini 116 10.6 0.001 0.001 43.399 43.640 prepare_preconditioner 11 7.9 0.000 0.000 42.750 42.769 make_preconditioner 11 8.9 0.000 0.001 42.749 42.769 make_full_inverse_cholesky 11 9.9 0.000 0.001 36.246 41.341 multiply_cannon_multrec 29820 15.5 14.083 18.864 27.565 32.070 rebuild_ks_matrix 127 8.3 0.001 0.001 28.693 28.943 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.018 28.693 28.942 mp_waitall_1 152434 16.5 17.158 26.849 17.158 26.849 qs_ks_update_qs_env 127 7.6 0.001 0.001 25.924 26.136 qs_ot_get_derivative 116 11.6 0.001 0.002 24.003 24.270 make_m2s 4970 13.5 0.095 0.100 20.037 21.074 make_images 4970 14.5 1.947 2.279 19.732 20.768 qs_ot_get_p 127 10.4 0.001 0.001 19.614 19.873 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 18.806 19.361 apply_single 127 13.6 0.001 0.001 18.806 19.360 ot_diis_step 116 11.6 0.017 0.018 19.253 19.255 cp_fm_upper_to_full 104 14.7 11.362 16.784 11.362 16.784 cp_fm_cholesky_invert 11 10.9 16.205 16.214 16.205 16.214 qs_ot_p2m_diag 82 11.4 0.339 0.385 15.570 15.621 multiply_cannon_metrocomm3 29820 15.5 0.046 0.049 6.135 14.453 cp_dbcsr_syevd 82 12.4 0.005 0.005 14.307 14.309 sum_up_and_integrate 127 10.3 0.139 0.152 13.939 13.970 integrate_v_rspace 127 11.3 0.003 0.004 13.799 13.835 dbcsr_mm_accdrv_process 61748 16.2 7.804 9.708 13.065 13.621 qs_rho_update_rho_low 127 7.7 0.001 0.001 12.993 13.035 calculate_rho_elec 127 8.7 0.173 0.189 12.992 13.035 dbcsr_complete_redistribute 393 12.7 1.509 1.625 9.046 12.928 multiply_cannon_sync_h2d 29820 15.5 11.703 12.832 11.703 12.832 make_images_data 4970 15.5 0.063 0.068 10.780 12.526 copy_fm_to_dbcsr 208 11.6 0.001 0.002 7.689 11.575 hybrid_alltoall_any 5155 16.4 0.520 2.195 9.647 11.321 cp_fm_diag_elpa 82 13.4 0.000 0.001 11.291 11.293 cp_fm_redistribute_end 82 14.4 1.924 11.220 1.938 11.227 cp_fm_diag_elpa_base 82 14.4 8.673 10.627 9.255 11.116 init_scf_run 11 5.9 0.000 0.001 10.570 10.571 scf_env_initial_rho_setup 11 6.9 0.000 0.001 10.570 10.571 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.481 10.225 mp_alltoall_i22 712 14.1 5.650 9.469 5.650 9.469 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.182 9.367 pw_transfer 1535 11.6 0.084 0.098 7.752 7.828 grid_integrate_task_list 127 12.3 7.437 7.814 7.437 7.814 cp_fm_cholesky_decompose 22 10.9 7.600 7.698 7.600 7.698 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 7.530 7.610 wfi_extrapolate 11 7.9 0.001 0.001 7.516 7.516 multiply_cannon_metrocomm4 24850 15.5 0.074 0.084 2.726 7.075 mp_irecv_dv 75445 16.2 2.583 6.820 2.583 6.820 fft_wrap_pw1pw2_140 519 13.2 0.479 0.494 6.667 6.756 density_rs2pw 127 9.7 0.005 0.006 6.164 6.620 calculate_dm_sparse 127 9.5 0.001 0.001 6.127 6.233 fft3d_ps 1281 14.7 2.776 2.854 5.780 5.841 grid_collocate_task_list 127 9.7 5.167 5.587 5.167 5.587 mp_alltoall_d11v 2401 14.1 4.818 5.472 4.818 5.472 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.383 5.434 rs_pw_transfer 1038 11.9 0.013 0.014 4.689 5.140 qs_energies_init_hamiltonians 11 5.9 0.048 0.112 4.714 4.716 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.323 4.407 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.342 4.392 potential_pw2rs 127 12.3 0.022 0.023 4.203 4.224 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.135 4.207 jit_kernel_multiply 13 15.9 2.521 4.008 2.521 4.008 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=187.706000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1085.000000, yerr=21.549098 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d71249dc44ddccc12d8161011fa5955b1ce8fead_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 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 1963542011904 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 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019182452736 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019182452736 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 92.796564E+12 0.0% 0.0% 100.0% flops max/rank 5.820057E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499488 0.0% 0.0% 100.0% number of processed stacks 1944496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3448.5 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.504625E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 238560 MPI messages size (bytes): total size 1.321104E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.537828E+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 112800 59139686400 4194304 < size <= 16777216 104112 545846722560 16777216 < size 20064 716108490000 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8852 52. MP_Alltoall 9584 804353. MP_ISend 39716 2104723. MP_IRecv 39716 2103824. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3995 58308. MP_Allreduce 10985 1176. MP_Sync 86 MP_Alltoall 1700 18828162. MP_ISendRecv 7620 122880. MP_Wait 19810 MP_ISend 10600 423612. MP_IRecv 10600 423612. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.033 170.505 170.506 qs_mol_dyn_low 1 2.0 0.003 0.003 170.128 170.140 qs_forces 11 3.9 0.003 0.003 169.405 169.409 qs_energies 11 4.9 0.001 0.002 162.149 162.160 scf_env_do_scf 11 5.9 0.001 0.001 144.834 144.849 velocity_verlet 10 3.0 0.001 0.001 110.590 110.633 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 109.237 109.238 dbcsr_multiply_generic 2485 12.5 0.180 0.185 74.792 75.325 qs_scf_new_mos 116 7.6 0.001 0.001 74.172 74.271 qs_scf_loop_do_ot 116 8.6 0.001 0.001 74.171 74.270 ot_scf_mini 116 9.6 0.003 0.004 69.777 69.850 multiply_cannon 2485 13.5 0.586 0.622 55.386 59.820 multiply_cannon_loop 2485 14.5 0.448 0.458 50.881 51.783 ot_mini 116 10.6 0.001 0.001 40.170 40.237 init_scf_loop 11 6.9 0.000 0.000 35.448 35.450 mp_waitall_1 128490 16.6 25.784 32.737 25.784 32.737 prepare_preconditioner 11 7.9 0.000 0.000 31.630 31.659 make_preconditioner 11 8.9 0.000 0.000 31.630 31.659 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.410 29.694 rebuild_ks_matrix 127 8.3 0.001 0.001 28.185 28.261 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.017 28.185 28.260 qs_ks_update_qs_env 127 7.6 0.001 0.001 25.731 25.804 qs_ot_get_derivative 116 11.6 0.002 0.002 20.608 20.685 multiply_cannon_multrec 9940 15.5 10.340 14.243 17.626 20.606 ot_diis_step 116 11.6 0.019 0.020 19.491 19.491 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.099 19.467 apply_single 127 13.6 0.001 0.001 19.099 19.467 multiply_cannon_metrocomm3 9940 15.5 0.022 0.023 12.100 18.827 make_m2s 4970 13.5 0.065 0.069 15.886 18.161 cp_fm_cholesky_invert 11 10.9 18.065 18.071 18.065 18.071 make_images 4970 14.5 2.267 2.663 15.583 17.852 qs_ot_get_p 127 10.4 0.001 0.001 16.865 16.935 sum_up_and_integrate 127 10.3 0.179 0.188 13.820 13.867 integrate_v_rspace 127 11.3 0.004 0.004 13.641 13.695 qs_ot_p2m_diag 82 11.4 0.489 0.495 13.190 13.206 qs_rho_update_rho_low 127 7.7 0.001 0.001 12.922 12.961 calculate_rho_elec 127 8.7 0.256 0.266 12.921 12.960 make_images_data 4970 15.5 0.052 0.061 9.732 12.300 cp_dbcsr_syevd 82 12.4 0.005 0.005 12.074 12.075 multiply_cannon_sync_h2d 9940 15.5 11.547 12.074 11.547 12.074 hybrid_alltoall_any 5155 16.4 0.829 3.776 9.547 11.664 init_scf_run 11 5.9 0.000 0.001 10.574 10.574 scf_env_initial_rho_setup 11 6.9 0.000 0.001 10.573 10.574 cp_fm_diag_elpa 82 13.4 0.000 0.000 9.151 9.152 cp_fm_diag_elpa_base 82 14.4 8.912 8.983 9.148 9.148 grid_integrate_task_list 127 12.3 7.718 7.960 7.718 7.960 cp_fm_cholesky_decompose 22 10.9 7.830 7.904 7.830 7.904 qs_ot_get_derivative_diag 76 12.4 0.002 0.003 7.796 7.847 pw_transfer 1535 11.6 0.082 0.088 7.595 7.605 dbcsr_mm_accdrv_process 20590 16.1 2.582 3.336 6.918 7.543 fft_wrap_pw1pw2 1281 12.7 0.009 0.010 7.376 7.393 mp_allgather_i34 2485 14.5 2.738 7.324 2.738 7.324 multiply_cannon_metrocomm1 9940 15.5 0.029 0.030 4.534 7.230 wfi_extrapolate 11 7.9 0.001 0.001 7.215 7.215 fft_wrap_pw1pw2_140 519 13.2 0.498 0.519 6.511 6.534 calculate_dm_sparse 127 9.5 0.001 0.001 6.063 6.152 density_rs2pw 127 9.7 0.005 0.006 5.813 6.005 dbcsr_complete_redistribute 393 12.7 2.130 2.227 5.319 5.694 grid_collocate_task_list 127 9.7 5.441 5.674 5.441 5.674 fft3d_ps 1281 14.7 2.698 2.764 5.551 5.578 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.491 5.494 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.189 5.191 mp_alltoall_d11v 2401 14.1 4.693 5.042 4.693 5.042 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.570 4.594 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.409 4.441 multiply_cannon_metrocomm4 7455 15.5 0.024 0.026 1.844 4.185 rs_pw_transfer 1038 11.9 0.012 0.013 3.959 4.170 mp_irecv_dv 28618 15.9 1.807 4.107 1.807 4.107 potential_pw2rs 127 12.3 0.026 0.027 3.945 3.959 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.556 3.854 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.488 3.825 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.640 3.685 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.534 3.623 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=170.506000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1394.545455, yerr=43.418395 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d71249dc44ddccc12d8161011fa5955b1ce8fead_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 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 1963542011904 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 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019182452736 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019182452736 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 92.796564E+12 0.0% 0.0% 100.0% flops max/rank 11.606412E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499488 0.0% 0.0% 100.0% number of processed stacks 1947808 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.6 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 3.088998E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 99400 MPI messages size (bytes): total size 1.127422E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.342272E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 44768 34745614336 4194304 < size <= 16777216 43984 376564613120 16777216 < size 10032 716108490000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3991 59299. MP_Allreduce 10975 1515. MP_Sync 86 MP_Alltoall 1700 36954339. MP_ISendRecv 3556 218624. MP_Wait 11506 MP_ISend 6360 1080477. MP_IRecv 6360 1080477. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.026 0.044 285.573 285.574 qs_mol_dyn_low 1 2.0 0.003 0.004 285.069 285.080 qs_forces 11 3.9 0.003 0.003 284.983 284.990 qs_energies 11 4.9 0.002 0.002 276.468 276.485 scf_env_do_scf 11 5.9 0.001 0.001 255.031 255.045 velocity_verlet 10 3.0 0.001 0.001 206.895 206.902 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 129.919 129.921 init_scf_loop 11 6.9 0.000 0.000 124.861 124.865 prepare_preconditioner 11 7.9 0.000 0.000 120.259 120.293 make_preconditioner 11 8.9 0.000 0.000 120.259 120.293 make_full_inverse_cholesky 11 9.9 0.000 0.000 95.909 117.443 qs_scf_new_mos 116 7.6 0.001 0.001 88.597 88.713 qs_scf_loop_do_ot 116 8.6 0.001 0.001 88.596 88.713 ot_scf_mini 116 9.6 0.004 0.004 83.881 83.917 dbcsr_multiply_generic 2485 12.5 0.211 0.217 81.678 82.213 cp_fm_upper_to_full 104 14.8 53.786 77.172 53.786 77.172 multiply_cannon 2485 13.5 0.698 0.761 57.787 58.328 multiply_cannon_loop 2485 14.5 0.469 0.474 54.221 55.543 ot_mini 116 10.6 0.001 0.001 43.453 43.484 dbcsr_complete_redistribute 393 12.7 3.988 4.043 29.890 43.012 copy_fm_to_dbcsr 208 11.6 0.001 0.002 26.576 39.663 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 24.303 37.235 mp_alltoall_i22 712 14.1 22.151 35.086 22.151 35.086 cp_fm_cholesky_invert 11 10.9 33.146 33.152 33.146 33.152 rebuild_ks_matrix 127 8.3 0.001 0.001 33.019 33.058 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.018 33.018 33.058 mp_waitall_1 104546 16.7 27.531 31.390 27.531 31.390 qs_ks_update_qs_env 127 7.6 0.001 0.001 30.802 30.841 qs_ot_get_p 127 10.4 0.001 0.001 25.210 25.220 qs_ot_get_derivative 116 11.6 0.001 0.002 23.643 23.678 qs_ot_p2m_diag 82 11.4 0.867 0.873 21.259 21.286 make_m2s 4970 13.5 0.075 0.076 19.433 20.356 make_images 4970 14.5 3.729 3.856 18.960 19.883 ot_diis_step 116 11.6 0.022 0.022 19.786 19.786 cp_dbcsr_syevd 82 12.4 0.005 0.006 19.581 19.583 multiply_cannon_metrocomm3 9940 15.5 0.023 0.023 18.373 19.466 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.033 19.230 apply_single 127 13.6 0.001 0.001 19.033 19.230 multiply_cannon_multrec 9940 15.5 10.446 12.222 17.796 17.859 cp_fm_diag_elpa 82 13.4 0.000 0.000 16.452 16.452 cp_fm_diag_elpa_base 82 14.4 12.031 13.628 16.448 16.448 multiply_cannon_sync_h2d 9940 15.5 15.538 15.549 15.538 15.549 sum_up_and_integrate 127 10.3 0.319 0.320 15.430 15.517 integrate_v_rspace 127 11.3 0.004 0.004 15.111 15.198 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.701 14.727 calculate_rho_elec 127 8.7 0.479 0.480 14.700 14.726 hybrid_alltoall_any 5155 16.4 1.294 3.013 10.475 12.351 make_images_data 4970 15.5 0.059 0.064 10.315 12.005 init_scf_run 11 5.9 0.000 0.001 11.540 11.540 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.540 11.540 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.202 9.237 dbcsr_mm_accdrv_process 20590 16.0 3.848 5.561 7.117 8.904 cp_fm_cholesky_decompose 22 10.9 8.760 8.798 8.760 8.798 grid_integrate_task_list 127 12.3 8.478 8.660 8.478 8.660 wfi_extrapolate 11 7.9 0.001 0.001 8.519 8.519 pw_transfer 1535 11.6 0.089 0.090 8.158 8.168 fft_wrap_pw1pw2 1281 12.7 0.010 0.010 7.926 7.936 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 7.894 7.895 fft_wrap_pw1pw2_140 519 13.2 0.534 0.538 7.040 7.062 mp_alltoall_d11v 2401 14.1 6.752 6.887 6.752 6.887 calculate_dm_sparse 127 9.5 0.001 0.001 6.546 6.647 grid_collocate_task_list 127 9.7 6.287 6.321 6.287 6.321 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.195 6.285 copy_dbcsr_to_fm 185 11.7 0.004 0.004 5.967 6.023 fft3d_ps 1281 14.7 2.733 2.741 5.988 6.001 density_rs2pw 127 9.7 0.005 0.005 5.872 5.906 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=285.574000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2779.090909, yerr=173.730959 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d71249dc44ddccc12d8161011fa5955b1ce8fead_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 1.261494E+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 6918638. MP_ISendRecv 570 19200. MP_Wait 1302 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.112 0.301 86.094 86.096 qs_energies 1 2.0 0.000 0.000 84.665 84.735 ls_scf 1 3.0 0.002 0.007 83.506 83.577 dbcsr_multiply_generic 111 6.7 0.015 0.019 72.617 72.795 multiply_cannon 111 7.7 0.019 0.037 55.916 56.786 multiply_cannon_loop 111 8.7 0.209 0.220 52.497 53.677 ls_scf_main 1 4.0 0.000 0.002 52.227 52.229 density_matrix_trs4 2 5.0 0.002 0.003 46.760 46.866 ls_scf_init_scf 1 4.0 0.006 0.024 28.147 28.149 ls_scf_init_matrix_S 1 5.0 0.003 0.020 27.071 27.126 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.003 24.935 24.959 mp_waitall_1 11316 10.9 22.582 24.857 22.582 24.857 multiply_cannon_multrec 2664 9.7 8.146 8.801 15.516 17.131 multiply_cannon_sync_h2d 2664 9.7 13.515 15.330 13.515 15.330 make_m2s 222 7.7 0.010 0.021 13.057 13.611 make_images 222 8.7 0.099 0.108 13.033 13.591 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 9.785 12.420 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.496 8.158 make_images_data 222 9.7 0.004 0.005 7.626 8.139 dbcsr_mm_accdrv_process 4760 10.4 0.510 0.617 6.988 7.914 hybrid_alltoall_any 227 10.6 0.215 1.823 6.544 7.845 dbcsr_mm_accdrv_process_sort 4760 11.4 6.280 7.157 6.280 7.157 calculate_norms 4752 9.8 5.499 6.082 5.499 6.082 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.084 5.211 mp_sum_l 807 5.4 3.159 4.643 3.159 4.643 multiply_cannon_metrocomm4 2442 9.7 0.011 0.014 2.072 3.657 mp_irecv_dv 6231 10.9 2.054 3.625 2.054 3.625 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.309 3.361 make_images_sizes 222 9.7 0.000 0.000 0.680 3.227 mp_alltoall_i44 222 10.7 0.680 3.227 0.680 3.227 arnoldi_extremal 4 6.8 0.002 0.014 3.194 3.223 arnoldi_normal_ev 4 7.8 0.003 0.007 3.192 3.223 ls_scf_post 1 4.0 0.004 0.042 3.130 3.205 build_subspace 16 8.4 0.009 0.012 3.084 3.087 ls_scf_store_result 1 5.0 0.000 0.000 2.875 2.912 dbcsr_special_finalize 555 9.7 0.005 0.006 2.392 2.778 dbcsr_merge_single_wm 555 10.7 0.457 0.599 2.384 2.771 make_images_pack 222 9.7 2.210 2.643 2.212 2.644 dbcsr_matrix_vector_mult 304 9.0 0.003 0.010 2.300 2.548 dbcsr_sort_data 658 11.4 2.183 2.492 2.183 2.492 dbcsr_matrix_vector_mult_local 304 10.0 2.063 2.462 2.065 2.464 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.205 2.298 buffer_matrices_ensure_size 222 8.7 1.747 2.030 1.747 2.030 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.730 1.732 rebuild_ks_matrix 3 7.3 0.000 0.000 1.720 1.722 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.005 1.720 1.722 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=86.096000, 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/d71249dc44ddccc12d8161011fa5955b1ce8fead_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 2.107023E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 10339. MP_Sync 4 MP_Alltoall 47 15335933. MP_ISendRecv 282 57600. MP_Wait 828 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.062 0.096 91.423 91.424 qs_energies 1 2.0 0.000 0.000 90.120 90.125 ls_scf 1 3.0 0.000 0.001 88.808 88.813 dbcsr_multiply_generic 111 6.7 0.015 0.016 74.985 75.303 multiply_cannon 111 7.7 0.028 0.043 53.133 56.838 ls_scf_main 1 4.0 0.000 0.000 54.793 54.799 multiply_cannon_loop 111 8.7 0.116 0.123 49.871 52.725 density_matrix_trs4 2 5.0 0.002 0.003 49.165 49.355 ls_scf_init_scf 1 4.0 0.000 0.002 30.535 30.537 mp_waitall_1 9246 10.9 20.997 29.898 20.997 29.898 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.402 29.504 multiply_cannon_multrec 1332 9.7 13.121 16.572 22.416 27.016 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.969 26.982 multiply_cannon_metrocomm3 1332 9.7 0.006 0.007 11.677 20.634 make_m2s 222 7.7 0.006 0.007 15.294 16.030 make_images 222 8.7 1.582 1.952 15.264 16.000 dbcsr_mm_accdrv_process 4041 10.4 0.289 0.436 8.901 10.440 dbcsr_mm_accdrv_process_sort 4041 11.4 8.493 10.020 8.493 10.020 make_images_data 222 9.7 0.004 0.004 8.797 9.783 hybrid_alltoall_any 227 10.6 0.521 2.471 8.282 9.209 mp_sum_l 807 5.4 5.260 8.044 5.260 8.044 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.226 7.772 mp_irecv_dv 3311 11.0 3.207 7.701 3.207 7.701 calculate_norms 2376 9.8 5.999 6.775 5.999 6.775 multiply_cannon_sync_h2d 1332 9.7 4.828 6.637 4.828 6.637 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.023 6.594 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.004 5.214 arnoldi_extremal 4 6.8 0.000 0.000 4.655 4.675 arnoldi_normal_ev 4 7.8 0.001 0.004 4.655 4.674 build_subspace 16 8.4 0.014 0.021 4.401 4.406 ls_scf_post 1 4.0 0.000 0.009 3.480 3.486 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.178 3.405 ls_scf_store_result 1 5.0 0.000 0.000 3.191 3.304 dbcsr_matrix_vector_mult_local 304 10.0 2.782 3.270 2.784 3.272 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 1.150 2.635 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.500 2.587 mp_allgather_i34 111 8.7 0.967 2.412 0.967 2.412 make_images_pack 222 9.7 2.029 2.346 2.031 2.348 dbcsr_sort_data 436 11.2 1.831 2.048 1.831 2.048 dbcsr_data_new 4174 10.1 1.615 1.849 1.615 1.849 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.424000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1704.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d71249dc44ddccc12d8161011fa5955b1ce8fead_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 2.692813E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265470. MP_Allreduce 3058 11181. MP_Sync 4 MP_Alltoall 47 23526250. MP_ISendRecv 186 57600. MP_Wait 732 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.175 0.399 95.731 95.733 qs_energies 1 2.0 0.000 0.000 94.247 94.259 ls_scf 1 3.0 0.000 0.001 92.832 92.844 dbcsr_multiply_generic 111 6.7 0.016 0.017 77.698 78.012 ls_scf_main 1 4.0 0.000 0.000 57.762 57.766 multiply_cannon 111 7.7 0.038 0.102 53.676 57.571 multiply_cannon_loop 111 8.7 0.100 0.106 50.149 54.142 density_matrix_trs4 2 5.0 0.002 0.003 51.954 52.151 mp_waitall_1 7374 11.0 24.469 34.087 24.469 34.087 ls_scf_init_scf 1 4.0 0.000 0.001 31.537 31.539 ls_scf_init_matrix_S 1 5.0 0.001 0.007 30.383 30.483 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.357 27.367 multiply_cannon_multrec 888 9.7 12.560 15.040 21.427 24.820 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.290 23.355 make_m2s 222 7.7 0.006 0.007 17.168 18.380 make_images 222 8.7 1.974 2.293 17.130 18.339 hybrid_alltoall_any 227 10.6 0.620 2.860 9.458 10.929 make_images_data 222 9.7 0.004 0.004 9.769 10.871 dbcsr_mm_accdrv_process 3754 10.4 0.300 0.947 8.374 9.561 mp_sum_l 807 5.4 5.277 9.444 5.277 9.444 dbcsr_mm_accdrv_process_sort 3754 11.4 7.752 8.902 7.752 8.902 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.133 7.558 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 4.032 7.474 multiply_cannon_sync_h2d 888 9.7 6.075 7.213 6.075 7.213 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.457 6.997 mp_irecv_dv 2335 11.1 2.441 6.953 2.441 6.953 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.455 5.656 arnoldi_extremal 4 6.8 0.000 0.000 5.106 5.121 arnoldi_normal_ev 4 7.8 0.001 0.005 5.105 5.121 build_subspace 16 8.4 0.014 0.020 4.776 4.784 calculate_norms 1584 9.8 4.252 4.639 4.252 4.639 mp_allgather_i34 111 8.7 1.370 3.883 1.370 3.883 dbcsr_matrix_vector_mult 304 9.0 0.005 0.016 3.473 3.788 dbcsr_matrix_vector_mult_local 304 10.0 3.064 3.628 3.066 3.630 ls_scf_post 1 4.0 0.000 0.001 3.533 3.546 ls_scf_store_result 1 5.0 0.000 0.000 3.277 3.357 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.725 2.817 make_images_sizes 222 9.7 0.000 0.000 1.119 2.261 mp_alltoall_i44 222 10.7 1.119 2.260 1.119 2.260 dbcsr_sort_data 325 11.1 1.851 2.150 1.851 2.150 make_images_pack 222 9.7 1.811 2.116 1.814 2.119 dbcsr_data_release 9322 10.9 1.336 2.007 1.336 2.007 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=95.733000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2171.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d71249dc44ddccc12d8161011fa5955b1ce8fead_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 3.362664E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 13371. MP_Sync 4 MP_Alltoall 47 30278988. MP_ISendRecv 138 86400. MP_Wait 600 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.065 0.133 97.284 97.285 qs_energies 1 2.0 0.000 0.000 96.538 96.563 ls_scf 1 3.0 0.000 0.001 94.897 94.923 dbcsr_multiply_generic 111 6.7 0.018 0.023 78.685 78.931 ls_scf_main 1 4.0 0.000 0.000 58.919 58.920 multiply_cannon 111 7.7 0.057 0.122 51.699 56.407 density_matrix_trs4 2 5.0 0.002 0.003 52.931 53.051 multiply_cannon_loop 111 8.7 0.114 0.126 46.688 49.414 ls_scf_init_scf 1 4.0 0.001 0.005 32.712 32.715 ls_scf_init_matrix_S 1 5.0 0.000 0.001 31.431 31.506 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.941 28.957 mp_waitall_1 6438 11.0 22.877 28.945 22.877 28.945 multiply_cannon_multrec 1332 9.7 14.181 17.215 22.099 25.622 make_m2s 222 7.7 0.007 0.008 21.213 22.640 make_images 222 8.7 3.140 3.601 21.162 22.591 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.185 17.468 make_images_data 222 9.7 0.004 0.005 11.861 13.490 hybrid_alltoall_any 227 10.6 0.799 3.820 11.224 12.954 dbcsr_mm_accdrv_process 3641 10.4 0.212 0.408 7.551 9.072 dbcsr_mm_accdrv_process_sort 3641 11.4 7.178 8.655 7.178 8.655 mp_sum_l 807 5.4 4.201 7.936 4.201 7.936 multiply_cannon_sync_h2d 1332 9.7 5.552 6.321 5.552 6.321 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.264 6.226 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.657 6.019 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.065 5.892 mp_irecv_dv 3229 10.9 2.042 5.843 2.042 5.843 arnoldi_extremal 4 6.8 0.000 0.000 5.249 5.266 arnoldi_normal_ev 4 7.8 0.001 0.005 5.248 5.266 build_subspace 16 8.4 0.014 0.021 4.897 4.906 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.529 4.689 calculate_norms 2376 9.8 4.179 4.544 4.179 4.544 mp_allgather_i34 111 8.7 2.137 4.417 2.137 4.417 dbcsr_matrix_vector_mult 304 9.0 0.006 0.016 3.602 3.919 dbcsr_matrix_vector_mult_local 304 10.0 3.230 3.745 3.232 3.747 dbcsr_sort_data 658 11.4 3.076 3.440 3.076 3.440 ls_scf_post 1 4.0 0.002 0.008 3.266 3.291 dbcsr_special_finalize 555 9.7 0.006 0.007 2.807 3.190 dbcsr_merge_single_wm 555 10.7 0.529 0.652 2.799 3.182 ls_scf_store_result 1 5.0 0.000 0.000 3.003 3.061 ls_scf_dm_to_ks 2 5.0 0.001 0.016 2.885 2.944 dbcsr_data_release 10477 10.7 1.586 2.499 1.586 2.499 dbcsr_finalize 304 7.8 0.050 0.062 1.811 2.001 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.285000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2716.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d71249dc44ddccc12d8161011fa5955b1ce8fead_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 4.630671E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265558. MP_Allreduce 3049 15663. MP_Sync 4 MP_Alltoall 47 46208988. MP_ISendRecv 90 115200. MP_Wait 573 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.047 0.081 98.970 98.971 qs_energies 1 2.0 0.000 0.000 98.085 98.092 ls_scf 1 3.0 0.000 0.000 96.181 96.186 dbcsr_multiply_generic 111 6.7 0.017 0.018 77.930 78.141 ls_scf_main 1 4.0 0.000 0.000 61.955 61.955 multiply_cannon 111 7.7 0.078 0.134 55.382 60.973 density_matrix_trs4 2 5.0 0.002 0.003 55.034 55.109 multiply_cannon_loop 111 8.7 0.069 0.078 50.983 52.653 mp_waitall_1 5481 11.0 26.609 32.004 26.609 32.004 ls_scf_init_scf 1 4.0 0.000 0.000 30.634 30.639 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.444 29.501 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.284 27.297 multiply_cannon_multrec 444 9.7 13.953 16.249 21.064 23.390 make_m2s 222 7.7 0.004 0.005 17.756 20.144 make_images 222 8.7 3.723 4.419 17.695 20.085 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.113 16.629 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.356 14.528 make_images_data 222 9.7 0.003 0.004 9.944 12.339 hybrid_alltoall_any 227 10.6 0.790 3.754 9.771 12.203 dbcsr_mm_accdrv_process 3003 10.4 0.186 0.375 6.810 8.010 dbcsr_mm_accdrv_process_sort 3003 11.4 6.486 7.660 6.486 7.660 multiply_cannon_sync_h2d 444 9.7 6.563 7.644 6.563 7.644 mp_allgather_i34 111 8.7 2.640 6.987 2.640 6.987 arnoldi_extremal 4 6.8 0.000 0.000 5.893 5.911 arnoldi_normal_ev 4 7.8 0.002 0.005 5.893 5.910 build_subspace 16 8.4 0.015 0.020 5.478 5.489 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.543 4.675 mp_sum_l 807 5.4 2.854 4.636 2.854 4.636 dbcsr_matrix_vector_mult 304 9.0 0.007 0.016 4.199 4.403 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.708 4.228 dbcsr_matrix_vector_mult_local 304 10.0 3.747 4.222 3.750 4.224 mp_irecv_dv 1241 11.2 1.689 4.200 1.689 4.200 calculate_norms 792 9.8 3.541 3.672 3.541 3.672 ls_scf_post 1 4.0 0.000 0.000 3.592 3.598 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.346 3.437 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.931 3.432 ls_scf_store_result 1 5.0 0.000 0.000 3.368 3.422 make_images_sizes 222 9.7 0.000 0.000 1.169 3.376 mp_alltoall_i44 222 10.7 1.168 3.375 1.168 3.375 dbcsr_finalize 304 7.8 0.062 0.078 2.198 2.316 dbcsr_merge_all 275 8.9 0.474 0.538 2.043 2.148 dbcsr_data_release 10123 10.8 1.325 1.987 1.325 1.987 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=98.971000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3608.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d71249dc44ddccc12d8161011fa5955b1ce8fead_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 8.722936E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284111. MP_Allreduce 3043 21950. MP_Sync 4 MP_Alltoall 47 88727262. MP_ISendRecv 84 732600. MP_Wait 309 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.081 0.096 107.484 107.484 qs_energies 1 2.0 0.000 0.000 106.111 106.126 ls_scf 1 3.0 0.000 0.000 103.255 103.270 dbcsr_multiply_generic 111 6.7 0.023 0.027 77.481 77.627 ls_scf_main 1 4.0 0.000 0.000 65.769 65.770 density_matrix_trs4 2 5.0 0.002 0.003 56.987 57.038 multiply_cannon 111 7.7 0.151 0.307 49.985 52.029 multiply_cannon_loop 111 8.7 0.067 0.069 46.469 47.320 ls_scf_init_scf 1 4.0 0.000 0.000 33.813 33.813 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.493 32.512 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.673 29.681 mp_waitall_1 4569 11.1 22.110 25.864 22.110 25.864 make_m2s 222 7.7 0.005 0.005 23.825 24.864 make_images 222 8.7 4.603 5.011 23.717 24.748 multiply_cannon_multrec 444 9.7 17.906 18.765 22.541 23.206 make_images_data 222 9.7 0.003 0.004 13.051 15.516 hybrid_alltoall_any 227 10.6 1.665 3.640 12.873 15.489 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.480 10.871 multiply_cannon_sync_h2d 444 9.7 8.820 8.874 8.820 8.874 arnoldi_extremal 4 6.8 0.000 0.000 7.428 7.441 arnoldi_normal_ev 4 7.8 0.003 0.009 7.428 7.441 build_subspace 16 8.4 0.026 0.036 6.877 6.889 dbcsr_matrix_vector_mult 304 9.0 0.009 0.025 5.556 5.706 dbcsr_matrix_vector_mult_local 304 10.0 5.160 5.453 5.163 5.456 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.995 5.258 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.763 4.856 dbcsr_mm_accdrv_process 1814 10.4 0.230 0.307 4.469 4.598 dbcsr_mm_accdrv_process_sort 1814 11.4 4.171 4.301 4.171 4.301 ls_scf_post 1 4.0 0.000 0.000 3.673 3.687 mp_allgather_i34 111 8.7 1.166 3.631 1.166 3.631 make_images_sizes 222 9.7 0.000 0.000 1.413 3.435 mp_alltoall_i44 222 10.7 1.413 3.435 1.413 3.435 ls_scf_store_result 1 5.0 0.000 0.000 3.423 3.435 calculate_norms 792 9.8 3.268 3.319 3.268 3.319 dbcsr_finalize 304 7.8 0.082 0.089 3.100 3.183 dbcsr_merge_all 275 8.9 0.894 0.918 2.884 2.961 dbcsr_complete_redistribute 5 7.6 1.432 1.473 2.768 2.862 dbcsr_data_release 12724 10.6 2.336 2.844 2.336 2.844 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.826 2.826 matrix_ls_to_qs 2 6.0 0.000 0.000 2.424 2.528 dbcsr_sort_data 325 11.1 2.447 2.516 2.447 2.516 dbcsr_new_transposed 4 7.5 0.243 0.251 2.292 2.315 dbcsr_frobenius_norm 74 6.6 2.056 2.128 2.191 2.237 dbcsr_add_d 103 6.2 0.000 0.000 2.134 2.202 dbcsr_add_anytype 103 7.2 0.859 0.891 2.134 2.202 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.178 2.179 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=107.484000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6839.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: d71249dc44ddccc12d8161011fa5955b1ce8fead Summary: empty Status: OK